Home > Columns > CRM Columns

Customer Contact Week 2019 Celebrates its 20th Year


Presented By: CrmXchange

CRMXchange Onsite Review

Customer Contact Week 2019 Celebrates its 20th Year 

In 2019, CCW reached a significant milestone. The event marked 20 years as a leading industry forum that has helped to define, shape, and continuously deliver relevant information on groundbreaking industry growth and innovation for customer contact specialists. Started in 1999 as Call Center Week, CCW has evolved into perhaps one of the world’s largest customer contact event series. CCW 2019 was held June 24-28 at the Mirage in Las Vegas. CCW is produced by the Customer Management Practice - the Analyst, Advisor, and Industry Network for all things Customer Management, under the multinational conference organizer IQPC.

The educational program at CCW incorporated 7 tracks of topical content, 80 interactive discussion groups, and an expo hall with over 200 customer-centric solution providers, perhaps the largest and most robust CCW ever. Highlights included a lock note address by Lakers great and entrepreneur Earvin “Magic” Johnson, as well as high-profile headliners such as Nancy Brinker, Founder of the Susan G. Komen Foundation and Goodwill Ambassador for Cancer Control for the UN’s World Health Organization, and the return of Tony Hsieh, CEO of customer service icon Zappo’s. Networking opportunities included: • Book Signings • “Brilliance” Clinics • Customer Contact Women High-Tea Luncheon • ‘5:00 o’clock somewhere’ Happy Hour • After Parties • and the annual awards gala, featuring entertainment by impressionist Frank Caliendo. 

CRMXchange was on the scene both on the exhibit floor and in the press room to gain insight about a variety of innovative solutions and services available to CX and contact center professionals. Suppliers of AI and machine learning solutions, chatbots, predictive analytics solutions, training, and messaging for the contact center were in abundance. Our coverage includes transcriptions of live onsite interviews with our managing partner, Sheri Greenhaus as well as answers to advance questions submitted.   

Click the link below to read the ebook                   Click the link below to listen to the mp3 recording
Ada Ada Audio
AT&T AT&T Audio
Avaya Avaya Audio
Chatbox Chatbox Audio
Cloud Call Center Search Cloud Call Center Search Audio
eClerx eClerx Audio
Helpshift Helpshift Audio
ServiceNow ServiceNow Audio
Semafone Semafone Audio
ProcedureFlow ProcedureFlow Audio
Observe.ai Observe.ai Audio


Scroll down to read the entire column.

 ada logo


Ruth Zive, Head of Marketing, Ada, discussed the benefits of the company’s easy-to-implement   chatbot technology. The company focuses on providing businesses with a platform that allows non-technical CX professionals to build, manage, and track the automated experience.

How did you come up with the company name?

We're named after Ada Lovelace who is widely regarded as the person who wrote the first line of code. We’re very proud of that and love the idea of aligning ourselves with a strong female leader in the software space.

What do CX professionals need to know about the solution?

At the highest level, Ada is an AI-powered chatbot solution. The word ‘chatbot’ often evokes a lot of feelings in people but we're not the early generation chatbots that were straightforward decision tree, FAQ response bots. Rather, we see our platform, as do our clients, as the underpinning of a company’s customer service strategy. We always recommend an automation-first strategy so that customers can self-serve in a truly meaningful way. Our platform drives efficiency and increased customer satisfaction. Ultimately, our vision for our clients is that they transform their customer support organization from a cost center to a profit center.

When you say “automation-first”, are you referring to the fact that the customer can help themselves? How smart can Ada be?

Very smart, indeed. Not only do we enable baseline table stakes automation where we can respond to frequently asked questions, Ada also allows for sophisticated, highly personalized interaction with the users. For instance, for clients in the travel industry, we can enable their customers to upgrade reservations, purchase tickets, change seats, check on baggage, etc. We can get very highly personalized with the ability to plug into backend systems and Ada can understand, from the customer's past behavior and past history, what their needs and interests might be.

You are presenting here with one of your customers. Can you tell us a little about what they're doing?

We’re proud to be working with Air Asia, which has experienced great success with their automated virtual assistant, which ironically, they refer to “AVA’ (AirAsia Virtual All-Star). Inside of 30 days, they were able to bring down customer wait time from over an hour to under a minute. They shared today that Ada is automating somewhere in the area of millions of conversations.

The way it was characterized was that through Ada, they are now able to do the work of 600 support agents with one sophisticated chatbot. Beyond the amazing improvement in efficiency they've seen their customer satisfaction scores just soar as a result of providing this self-service access on demand, round the clock in the language of their customer’s choice.

Talking about rising customer satisfaction scores, do you think the ability to have the self-service customer get all the information right away makes a significant difference?

Yes, for two reasons. One is because we know that customers prefer to self-serve when it's straightforward and accessible. Perhaps the greater value of Ada is that it frees up humans to deal with more mission critical inquiries. We can escalate to a live agent very seamlessly. You don't have to wait in long queues. Ada can detect when that escalation is necessary using natural language processing and intent to determine when an escalation is warranted.

So how are businesses able to marry the two modalities? How can businesses gauge how many agents they need when they have an efficient self-service option?

While we recommend an automation-first strategy, the reality is that most companies have call centers in place and are dealing with legacy support systems. We integrate with those systems. Whether they use live agent support, call center, or email ticketing, we provide a seamless, easy to integrate backend. The integration can be done immediately right out of the box. The starting point is to work very closely with our clients to automate those inquiries that are responsible for the highest volume of interactions.

What type of people pull Ada together?  What is that team like?

That's a great question. One of the reasons that our customers choose us is because we relieve their customer service organization of its dependency on IT. At the end of the day, it's the CX organization that best understands the needs of customers. We have a very strong bias that they're the ones that should be owning and managing the chatbot platform and automation strategy. This is exactly what happens with Ada. It’s the CX reps who are building the flows that are evolving and are evaluating how it's being used.

When customers are calling in to self-serve and there is an issue requiring something to be changed, how easy is that for the teams to accomplish?

While transactional inquiries and activities inside of Ada are super straightforward. It depends on the nature of what the inquiry is, what tasks need to be automated, and what the backend systems look like. It is very easy to build those flows and our backend is a user friendly drag-and-drop.

Does Ada work in multiple languages? For example, is it able to switch from English to Spanish when talking to a US customer?

Another great question. A lot of our clients are global and so being able to provide multilingual support is a top-of-mind priority. It’s one more reason why we get chosen, because companies can build in one language and deploy in over a hundred. We accommodate most existing languages.

As to the second part of your question, I don't have a technical background, but my understanding is that it detects the region where the question is coming from and then deploys in the language of that region.

How is Ada adaptable to different channels?

The bots can reside on a company’s website, on its mobile app, on Facebook messenger, and on SMS to allow for a complete omnichannel experience. We know that customers expect to be met where they prefer and we want to make sure that they have support on the channel of their choice. It’s easy to deploy on all of those channels.

What makes Ada stand out among the many chatbots vendors on the floor?

With so many vendors it is now a noisy space. There are a few things that make us special. One is ease of use and we've proven that over time. We're intuitive and diminish dependency on IT to make it work. Secondly, we support a highly personalized experience. It's more than just a decision tree and FAQ responses. We do that of course, but that's table stakes. We want to offer our customers a delightful, meaningful experience, thus we put that power in the hands of the CX team. The third way is that we think about customer experience in a different way than most other chatbot vendors. Of course, we're looking to diminish costs, but can truly get our customers to zero cost support – an opportunity that businesses are overlooking. Businesses can drive targeted upsell and promotion and discounting through their chatbots. Our clients are automatically generating revenue through those flows by offsetting the cost of their support organization and this puts them on a direct path to becoming a profit center.

att logo

AT&T Contact Center Solutions

Mike Rajich, Director, AT&T, Voice and Collaboration, talked about how his division of the telecommunications giant works with companies to improve ways on how to collaborate and drive business outcomes in the process.

Please tell our audience about AT&T’s…and your own… objectives in the contact center marketplace?

My specific focus is in helping businesses build and nurture their customer relationships by maximizing each of their touch points to increase revenue, reduce costs and increase customer satisfaction, which are the key goals in a contact center business. We've been very successful on building these relationships with many of our leading technology providers. This has allowed us to build a depth and breadth of options for our customers to choose from. We create our content strategies from this broad array of services with the awareness that a lot of customer engagement models are driven around performance and management and capturing analytics.

Your website cites that 78% of board rooms recognize the customer experience as the most important strategic performance measure. What do you believe to be the essential element of an empowered context center that helps businesses set themselves apart?

That is an important question. We believe nearly every facet of CX is fluid. Customer needs and their expectations are constantly changing and we see technology changes as well. These businesses are rapidly evolving and going through key transformations which is also driving new requirements and changes within their businesses. For businesses to truly succeed in delivering the positive experiences they desire, we approach it with agility in mind and recognize the need to be ready to enhance or even alter the course when it's necessary.  One key is understanding how customers want to communicate with the business - whether that is via voice, chat, social, or video, all the different mediums of communication. We build out these capabilities to be designed to meet the specific customer needs. We also emphasize creating efficient experiences that allow customer data and the contact information to follow the customer regardless of the channel or where there are in the actual journey.

With so many different available technologies, what do you see as the requisite technology considerations that companies should adopt to offer a consistently superior customer experience?

For specifically helping our customers, we're driving a lot of innovation around AI. If AI is currently not part of a company’s strategy going forward, we believe that company is going to be way behind. While the technology is still improving, its ability and potential to deliver a consistent, positive experience is not just future-talk. It’s happening now. It can make an amazing difference to have that predictive element in place to provide your agents with information at their fingertips to understand the customer's needs across all touchpoints.

Key considerations include making sure the required data is not siloed and follows the customer throughout their journey. Further, companies need to consider how the customer information is accessed and presented to both virtual and human agents in a method framework like a single pane of glass.

How can organizations meet the growing demand of customers for real-time information sharing on any touch point, any device or any channel?

We're seeing all channels extremely fragmented. I wake up every day and there seems to be another startup in the space that says it can instantly improve the CX experience, but building a true omnichannel experience is hard. With the bevy of technology options, deployment options, and vendor options, it can be like feeding into a slot machine that doesn’t necessarily guarantee a payout. That’s why it’s important to work with experts with the knowledge and experience of seeing how winning combinations are put together. In the end, omnichannel must be efficient for both the customer around their experience and the agent around their ability to service the customer.

Will you share with our audience some of the capabilities that your division brings to the table that enable its clients to make those improvements in their CX programs?

AT&T does business with tens of millions of customers around the globe and we continue to learn and implement best practices at scale. This combined with the longstanding, strategic relationships with industry leaders enables us to have a holistic view and perspectives that few can match. We provide contact center solutions for 7 of the Forbes 50 and several of the largest federal and state government agencies. AT&T has experience selling, managing, and integrating our award-winning network with all the leading contact center providers.

In addition, AT&T itself has some of the largest contact centers for its customers and has tested, vetted, and used many of the leading technologies to serve its customers. That experience is leveraged to both understand customer issues and to better serve them in their contact centers.

avaya logo


Paul Lang, Head of Contact Center Marketing, Avaya, discussed the company’s Avaya IX (Intelligent Experience) Workspaces which use advanced desktop analytics across the entire customer journey to help improve agent productivity, customer responsiveness, and matching customers to the right agent both inbound and outbound. The conversation also covered Avaya’s expanding AI solutions.

How does Avaya see Workspaces and what it can do in the contact center?

Workspaces describe an agent environment that is really linked into a customer journey or customer experience. The agent has a single pane of glass to look at…one where they're not scrambling across various screens or devices to see what's going on as part of that customer journey, regardless of media or channel or how they're operating. Agents can clearly see the customer journey and they can spend more time listening and working with the customer as opposed to having to toggle back and forth. They're able to drill down and to see where that customer's been, how they're touching the contact center. They can tell whether they've used self-service, the web, or engaged in chat. This enables a truly great customer experience to happen, which is going to be where more and more differentiation is going to happen as we move forward.

What does an agent's screen look like and what would they be seeing?

They'd be able to see what the company wanted them to see. It’s very much widget based. The screen could be shared, showing their own agents statistics or the customer journey. It could also show third-party statistics or something as simple as a Google map. In short, the screen would show anything the agent needs to be successful in delivering that great customer experience, ensuring customer loyalty.

Would Avaya customers that have Workspace be able to set up the screen that best fits their needs?

Yes, absolutely. That’s the whole point. It's the notion of providing a widget or optimizing the screen real estate. What we've done recently is expanded the reach of Workspaces as a desktop across our solutions starting off with our digital products, then added live chat, web chat, web collaboration, adding voice to enterprise solutions. This month we’re introducing it on our outbound solution, offering our clients the consistency of having the same desktop for inbound or outbound communications. We are able to set up the system in days and weeks rather than months. Businesses can communicate within the agent community, with the supervisors, and most importantly with the customers.

Agents looking at a screen now have all the information they need to best serve the customer in front of them, but if something happens in mid-conversation, does this information now go easily into a CRM?

Yes, that’s its most important benefit. It’s bringing that holistic view of what's going on with the customer or customers with whom they interact. The agents have all that information at their fingertips. Whatever becomes the system of record, the Workspace plays a key role in bringing that information to the agent.

As we talk about what the agent can do in conversations or other communications, there are processes going on in the background. Can the agent just push a button or perhaps an automatic process is triggered as a result of previous actions. This brings us to a discussion of AI and robotic process automation and how that all fits into that whole communication with the customer.

Great segue. Last year, we announced a partnership with a company called Afiniti, a specialist in AI-based behavioral pairing solutions. The purpose was to take the routing capability beyond where it has been to a brighter evolutionary step. Now when you’re calling into a contact center, it knows your records, your purchases, perhaps your complaints. We know all about the customer and then we have an agent who is aware of the context. Our joining forces with Afiniti will enable a business to route by attributes to pair the right combinations of customers with agents rather than doing the generic and less effective first in, first out.

We’ve also extended that capability to outbound. While there are many outbound use cases, they primarily fall into three categories, some form of sales call, a debt collection call, and calls that are directly related to customer care. Now we've got an available agent with these various skillsets so when the sales department reaches out to the customer, they are best equipped to sell a product or service in which we think the customer is interested  Thus, why not reserve this agent specifically to make that call, to set up right paring for successful outbound campaign? Now extending what we did with inbound on using intelligence, we also do on an outbound basis. I think the two together are a really great match.

What attributes do you measure and what do you base your pairings on? Is it product knowledge or demographics? Will you walk us through the process?

When a business has a customer’s profile, knows their buying patterns and preferences, it can use the attributes that have been have assigned to set up a productive pairing. The business can assign an agent who is a both a knowledge match and experienced in selling through contextual communication.  As companies test pairings, it becomes a learning process. Artificial intelligence is about incremental machine learning to make it work. Even if the optimal person is not available, the system can identify someone else who has similar attributes and make that pairing on either an inbound or outbound basis. We also recently introduced a conversational intelligence product that is now in use at numerous sites which enables businesses to drive desired business outcomes in real time. It goes beyond the capabilities of traditional speech analytics that can take two or three days to provide the insights that conversational intelligence can deliver in two to three seconds. If the call is going badly (as it is being consistently scored on various parameters), an alert is sent to the supervisor and who can listen-in in real time to avert a customer experience disaster. The information can also be leveraged to assist in customer journey mapping.

We've been talking about AI for a long time and it seems to be finally coming into its own. On a scale of one to 10, where do you think it is now as far as implementation and where do you see it going?

A recent survey that we commissioned revealed that nearly 99% of companies are experimenting with or putting into production some form of AI, whether for automation or better self- service. Over the next three or four years the number of businesses who are doing it in some way, shape or form is going to grow even more quickly. I’d say right now, those in the process of implementing is 9 out of 10, but those who have it in production at this point, maybe a 5. The technology is readily available. I think the barrier is use cases. How can my business benefit? Companies will use it for cost reduction, to improve customer service, and find ways to use it to build revenues. Right now, I envision AI as having just graduated from high school as an honor student. But to speed through college and get its PhD, it needs to be both relevant to specific business needs and serve customer experience goals set up across the entire business. The contact center has its role as part of the customer journey but is just part of the delivery mechanism.

What can we expect from Avaya in the coming year?

You’re going to see an ongoing acceleration in the customer experience solutions. As digital channels become more mainstream, how do we take that more to the cloud? We believe there's no one-size-fits-all solution. If a company wants the off-the-shelf, readily available option, it’s probably going to go to a public cloud. On the other end of the spectrum, if a business wants something heavily customized, perhaps in the government space with its various required certifications, it’s probably going to go private cloud. There's room in the middle for a hybrid cloud, complementing on-prem contact center solution. It could also be routing or digital in the cloud, Avaya is certainly cloud-first, but cloud-first with a choice.

 chatbox logo


Phil Gordon, CEO and Laura Ashley, VP Marketing, Chatbox. CRMXchange had an engaging conversation with these two multi-talented leaders about the company’s enterprise messaging platform which enables businesses to create and automate personalized, conversations across texting, chat, and social channels. Sheri also attended the evening poker clinic Phil conducted at the Mirage and found it enlightening even though she conceded that it won’t make her a tournament player anytime soon.

With texting having become such an integral part of most people’s daily life, why has it been an uphill battle to get more organizations to adopt omnichannel messaging to communicate with customers?

We think it comes down to three things: logistics, regulation and integration.

Logistics: Adopting a new communications channel means new technology, training, cost, integration, management, controls, permissions and security, to name a few. This can be a lot for organizations to wrap their heads around, convince decision makers of the ROI and then generate the momentum needed to change existing workflows. 

Regulation: Brand to consumer texting is regulated in the US (and elsewhere) yet the technology behind it is still widely misunderstood. This has led to a lot of market (and legal) inconsistency in terms of acceptable use and compliant behavior. Policy makers and market participants struggle between not wanting texting to be intrusive, while still providing consumers the convenience they demand.

Integration:  For an omnichannel strategy to make sense, your channels must be fully integrated with your various data sources so you can deliver a consistent brand experience in real-time. Integrating channels with CRMs, back-end solutions or other data silos can be tough. If you don’t do it, you will frustrate your customers and you might have been better off never deploying the channel in the first place.

Please tell us about the benefits of the Chatbox platform?

We help businesses deliver amazing text message-based experiences to their customers for service, sales, and marketing. Some of the largest businesses in the world are utilizing the Chatbox platform for those purposes. We're helping those businesses communicate better using text messaging that today’s customers demand.

Some call it texting while others refer to it as SMS messaging. Which do you prefer?

The terms are used interchangeably depending on the audience. We even sometimes revert and just call it messaging. We think of them all as the same.

For years at CCW, we’ve seen numerous messaging platforms, and it seems like a great idea, but we still don't see a lot of companies using them. What do you think is a hold up?

There are several obstacles that are holding back mainstream adoption. First, it's a complicated space to deliver outstanding customer experiences via text message. It requires a lot of work for companies to stitch these experiences together themselves. They need to worry about channel management and whether they operate on short code, long code, 800 numbers or more. They have to be concerned about automation. Many big businesses operate at scale and unless you have some sort of automation in play, it's impossible to reply to all the inbound traffic. Until recently, this has been a difficult challenge. The advent of Amazon Lex and Google Dialogueflow and other natural language processing engines have made that problem more tractable. The third problem that is holding companies back from deploying messaging is that while a text message can provide intent, it can also set off a chain of complex communications. Think about a simple item return. Suddenly a company is faced with having to text message back and forth six or seven times. Which item are you trying to return and what address should we send the box etc.? It can become very problematic. I think integration with incumbent systems has been an issue, all those things combined have made it a little bit slower than potential customers expect it to be. We’ve been trying to address those core problems and more with the Chatbox platform.

What level person would be pulling this whole program together?

Some of our biggest customers have appointed a Chatbox administrator. Think about it like a Salesforce administrator.  These administrators are responsible for monitoring inbound traffic, provisioning agents in the system, setting up the automation, tracking the analytics and more. We've had enterprise grade customers come to us and in under a week we have them fully up and operational with integration to their Salesforce system. It can be a very quick and efficient process, but it can also take longer depending on the level of integration that a business need with its backend system.

What makes your solution different?

We're the only platform on the market that seamlessly integrates the six key technologies that are required to operate these experiences at scale. The seamless orchestration of these six technologies is what makes Chatbox a platform versus just a point solution. First, we offer omnichannel capabilities, so that brands can communicate consistently across social messaging apps, text, and live chat options. Within each of those channel categories our platform can effortlessly manage multiple channels - a company may have hundreds of agents, each with varying permissions, all texting with customers via hundreds of separate numbers that may even be sourced from multiple SMS providers. Second, we allow companies to leverage automation for scale – think chatbots and natural language processing – so agents can focus their “human touch” on the most important customer interactions while letting bots handle the repetitive calls. Third, our Instant Apps provide an app-like experience to exchange rich, structured data on demand, all without a cumbersome download and install process. Fourth, our APIs integrate with backend and CRM systems to continually personalize the customer experience and ensure there is one source of truth. Rounding out the platform, our agent tools help service and call center reps manage their communications and permissions, while our analytics provide a configurable view of metrics across the entire platform.

We have robust channel management that enables us to deploy to any channel, to any SMS phone number, in a scalable way. For instance, one of our big customers has 900 phone numbers that we manage on their behalf. We help with automation as well, with keywords or natural language processing, and do it in a way that's straightforward, scalable, trackable, and moves the ball forward. A lot of companies get hung up right now. They think that AI is some magic bullet that's going solve all their problems and that no one's ever going to need an escalation to a human. That's simply not the case. The AI is not good enough yet, but it’s crucial to get some form of automation going and our platform makes it easy to automate the things that can be automated and hand off to humans when needed. As we alluded to earlier, we can handle the collection of structured data inside of those conversations. In our platform you can build checklists, wizards, forms, collect e-signatures, and deliver rich media content all within the context of a conversation. These experiences can be built without an engineer involved. When business requirements change on the fly, it's just drag and drop.

One of the largest travel agencies in the United States is our customer. There's a huge integration with their flight information service.  When you book a trip with them, you'll receive text messages with updates regarding your trip. Now you no longer have to call customer service to find out if your flight is late or the gate has changed. We can also process the data that's required to change that flight right within the text message conversation.

Why is it critical to maintain a consistent brand identity and customer experience across all channels?

Many companies have multiple ways of communicating with their customers, using a variety of channels which are often run by different departments. This can lead to a disjointed experience for the customer. If a customer has a bad experience, they don’t blame the team (or technology) that handled the communication, they blame the brand. Customers expect brands to have their data on hand (irrespective of channel) and they don’t want to have to repeat themselves or be transferred multiple times to solve a simple request.

Consumers do not have the time nor the inclination to figure out how to best communicate with brands. They want brands to meet them where they are, on demand. Consumers effortlessly switch channels as they go about their daily lives (from email to text-to-chat to voice to social) and they expect brands to be able to do so as well. If your customer experience is inconsistent across channels you degrade the customer experience and risk diluting your brand. The stats around the benefits of a consistent and integrated multichannel strategy are staggering. If you believe the adage that time is your most valuable resource, then a brand that doesn’t respect the value of their customers’ time does so at its own peril.

If there needs to be an escalation, does the agent receive the information from the texts?

Yeah, absolutely, and not just the text message transcript, it's also all the data that was collected inside of those app-like experiences that are built inside the platform. The agent actually has a real-time view of every bit of data that has been kept, that has been sent to and from that customer right inside the conversation. It's an astounding experience both for the customer and for the service agent to have all of that data right there. As an agent, I have full visibility into every button you clicked, every input field that you modified and all that data lives right inside your customer record.

What if a customer needs to make a return and requires a return label? If they're texting through some sort of robotic processing automation can the workflow create a label and send it to their mobile phone without any human intervention?

The key thing is for businesses to automate where it makes sense and escalate where it makes sense. There will always be a requirement for human-based interaction for the most complicated of cases, but that shouldn’t be an insoluble problem. Look for opportunities to automate the most common use cases, the simplest use cases so that your agents can concentrate on the issues that require both a brain to solve and the ability to express empathy.

What do you think the average consumer is looking for in text interactions?

Convenience. Texting is the only thing that every single mobile phone on the planet does and does well. If you look at the average US adult, they spend many hours a week text messaging. Since we’ve all just become accustomed to text first for our personal interactions, it should be the same way for businesses that care about creating meaningful conversations with their customers. Three years ago, only 4% of calls to mobile phones were spam. Last year it was 40% and this year it's going to go over 50%.  Fifty percent of phone calls to a mobile phone today are spam. Texting first should be one of the ways that you contact your customer: “Hi, this is Phil from Chatbox, I need to speak with you about your account. What might be a good time for a call?” That's a respectful way to interact with your customers, by contacting them in the fashion they are already interacting with other individuals in their sphere of influence.

We started experimenting with texting for our webinars. When somebody registers, we ask for permission to text and we promise we will discard their numbers. For those who agree, 75% show up.

That’s yet another benefit of SMS, nearly every text message is read within three minutes of delivery. The most recent stats show 95% of text messages are read within three minutes of delivery. Compare that to email of which only 20% are ever read.  Our big travel customer reports that approximately 90% of their travelers prefer getting notifications via text and changing their flight without having to call customer service or having to download and install another app.

Do you have any final thoughts for our audience that you would like to discuss?

One thing that's important to understand is that the text messaging world can still be complicated. Consumers that want to text message can sometimes be a little wary about being scammed or phished.  There’s a relatively new term called smishing. It is a combination of SMS and phishing. Being able to authenticate that you are who you say you are is a really important piece of the puzzle that we've just recently patented and solved.

We have authenticated SMS. If you use the keyword “ID” you'll get an automated reply that says, “Yes, this number is owned by [company name]. Visit our website and enter this code.” If you follow the instructions, you’ll receive a confirmation with the date and time of contact. Now you have confirmed this was an authenticated conversation and are assured that you’re talking to a legitimate company. We believe the space growing like crazy and ultimately you will see just about every major consumer brand in the country light up text messaging for sales and support.

cloud call center search logo

 Cloud Call Center Search

Fred Stacey, Founder. With over 500 cloud software providers in the call center market, choosing the right cloud call center software solution is no simple task. A 25-year veteran of the industry, Fred put together a team of experts to help companies make better informed decisions.

With all the suppliers in the marketplace who claim that they make it so easy to implement cloud contact center technology, why do businesses need assistance in selecting the right one?

The reality of our technology landscape is that there are far too many contact center technology providers for any IT or operations professional to possibly know. I am always learning of a new technology company launching in our space, and I live in the technology world because it is my business to know. It is impossible to expect anyone whose primary job function is running IT or operations to know all the right companies to consider.

When I started in this industry over 25 years ago, there were 5 major players and they all circled around the telephony side of the business. Today we have technology companies from AI to workforce optimization and everything in between.

What is your mission and how can you make it easier for companies to make the right choices?

We are a master agency for contact center technology. Most of the industry is not aware of the master agency model on the technology acquisition side, but it's been here since the early nineties when master agents first started to help on the telecom side. We partner with hundreds of different technology companies, evaluating them for their infrastructure integration points, functionalities, what verticals and industries they specialize in, get input from their clients, etc. We also have our clients and BPO partners that come to us and say, “We need a new omnichannel platform” or “We’re thinking about bringing in AI.” How can you make it easier for us to make the best investment?” We assist those individual buyers by starting a conversation about what the business needs first, what it’s trying to accomplish, what is its current infrastructure? We go through a discovery process, understand the client's needs and then bring vendors for them to review. With the CPaaS players coming into the marketplace, there are multiple technologies entering daily built on top of those platforms. I spend half my time evaluating technology companies in our industry and every week I still hear of a brand new contact center technology. I can't imagine as a buyer who has a day job, anyone can possibly understand the landscape of our marketplace. 

How do you see the cloud contact center technology landscape changing and what do companies need to know about the evolving options before they start the search process?

First, we are seeing both an expansion and consolidation in the technology players in our market. Acquisitions are hot and heavy as many companies in the tech space are trying to leverage the momentum of the growth in valuations and the “land grab” of the premise-based customers being forced to migrate.

With CPaaS platforms we will continue to see a rise in new entries into the market, which is only going to make the process more confusing for the end users/buyers. I expect to see this trend continue for the foreseeable future as no one has really established themselves as the clear industry leader in any segment of the market.

Companies considering their move to a new contact center technology need to really understand that there is a ridiculous number of options that can solve all kinds of problems. There is no one-size-fits-all solution in any segment of the contact center technology market. You need to prepare your internal organization by taking the appropriate steps to get a plan together for the jump to cloud. Journey mapping and digital transformation are not just buzz words. These processes make the understanding of what you need from a vendor significantly easier to define and implement.

In putting together a plan, do you create a complete timeline and roadmap for implementation, or do you find businesses simply want to fill an immediate need?

It’s a mix of both. In our industry, we encounter people who just need to implement something different. In those situations, we help them identify the right solution for their current infrastructure. We take pains to ensure what they implement today will help them move forward, but I prefer to be able to talk about a plan. My services are free to the end user and I can leverage my nearly 25 years in the industry to help them identify the first things they need to do, figure out the platforms they need to take a serious look at and show them how to map it out so that they use that information to put their own plan together.

There certainly are a lot of startups. What if you analyzed one and found they had an amazing solution, one that shows the type of difference-making potential that you had not previously seen. Would you advise a customer it might be a risk worth taking?

Absolutely. There are a few providers in my portfolio that know the needs of the contact center market and have built something that is truly unique; a solution which could revolutionize how a company operates and possibly change the landscape for my clients. I will bring such a company on as a partner, knowing full well that there is a risk involved. I'll tell my clients that and not every client with want to take that risk.  Often though, because it’s a startup with a smaller customer base, they have an opportunity to get in at a lower pricing structure but like any early adoption there is inherent risk.

If you walk through the exhibit hall at CCW, many vendors seem to be saying the same thing. If you were going to advise a company trying to find the right solution, what are some of the questions they should be asking?

Fundamentally, any business should start with understanding its own infrastructure. I always began with what I call the “system of record”.  I consider it to be one of the most critical pieces as we progress in the marketplace to digital processes and automations. If a company doesn't have its data normalized, it doesn’t have everything prepared for business intelligence, analytics, and AI for the future. Even if a business is not ready to implement these technologies, it’s still imperative to start at the system of record and understand where it stands. Then they can intelligently start looking for the most effective solution for their company in that space.

What are the technologies in which companies are now expressing the greatest interest?

We get a lot of “I need AI tomorrow” calls, but the reality is that many businesses are still kicking the tires. That’s what we saw with omnichannel and chatbots as companies got excited by the industry buzz without knowing what went into the adoption process. The tide is turning and AI is being implemented in many places. I'm seeing true progress in moving from POC (proof of concept) to full implementation.

Voice analytics is also a big talking point right now. Natural language processing or NLP engines are becoming more accurate. With the addition of AI and machine learning, a company can do some pretty impactful things with voice analytics and do it quickly. The one that has surprised me is CRMs. While everybody knows Salesforce and a lot of people bought into it, they’re tired of paying the license fees and issuing developer checks every month. Salesforce is the 800-pound gorilla and offers some tools that are good for the contact center, but now many of my customers are looking for contact center focused technologies that are much more flexible and cater towards the contact center model.

When a client comes to you for guidance, do you give them individual or multiple recommendations?

I always try to find at least two, if not three vendors, that fit all their needs. There are specific requirements in every deal. There's always going to be some requirements on integration and budget, but then you find out unique ones such as the ability to do routing based on certain data points, that not every system has available. I’ll put the solutions in front of them and they pick which vendors they want to talk to and can then skip the entire RFI/RFP process and go right to the short list of providers.

Does that make you essentially the Human RFI?

I had a partner tell me that I was the ‘Match.com of contact center technology’. We're trying to find the right match. We are more human RFI, RFP machine which takes the discovery process back to the technology companies, verifies that they're willing to compete for this deal and they have all the correct requirements.

Can you provide a few brief examples of how your services have enabled organizations to find the right cloud contact center solutions to meet their needs?

In one example, we had a client who needed to move very quickly to a new vendor for their CCaaS platform, but they had some unique requirements that not many vendors could meet. We worked with the client to define those needs and quickly engaged the vendors, bringing them right to the client to move quickly through demo into the selection process. The time from engagement to selection and implementation was a little over a month. This type of timeline is extremely difficult, but because of working together with the client and vendors we were able to accomplish this without any interruption to the brand’s ability to service their clients.

Do you have any final thoughts for our audience?

In this landscape, especially if your’re new to the contact center space and trying to identify the right technology, I would say even if you don’t use my company, find resources in the marketplace to help you. Read as much as you can. There are many great organizations to help you identify the vendors that you need. Just don't do it alone, because no one has do it that way today.

eclerx logo


Pamela McGlone, Marketing Lead, Customer Operations and Dave Goldberg, Head CX Group, eClerx.  Pam and Dave stopped by the CCW press room to talk about how the company was able to elevate the customer experience, drive transformative change, and reduce operating costs.

In what ways do you collaborate with business partners to deliver holistic and tightly focused customer support?

eClerx is a full-service business process outsourcing company that provides traditional live agent support via voice, chat, email, social, and other channels. We have a rich array of technology available to augment our live agent support channels which include Robotic Process Automation, Artificial Intelligence and Machine learning that we deploy not only in direct service of customers but also to support our analysts as they strive to deliver a superior customer experience. Our Data & Advanced Analytics team is interwoven throughout our business model so that we can pinpoint areas of focus and ensure our delivery tactics are aligned.

When you say you have analytics to help your customers make decisions, isn’t it also your customers’ customers who are making the decisions?

Pam: Both. We’re primarily focused on delivering for our clients and their customers to do digital analytics off the website based on the tagging and finding out where people are getting hung up on websites, whether trying to complete transactions or make purchases or whatever they're doing. We deploy analytics in the customer experience space to understand what's going on in the interactions, how the agents are performing against what we expect from them. It’s also to learn what's driving some of those customer contacts in the first place so that we can be proactive in deploying better digital solutions and better channel support for customers.

Can you explain the monitoring process?

Dave: We monitor across customer service channels. It could be voice, chat, email, or social. We have analysts with domain expertise in certain industries who have listened to 2 million customer interactions a year across the client base we serve. We have a lot of data and information that's not about the data that we capture. It's how we create the output and then provide business recommendations to our clients that say, “We know your agents are struggling in these areas. Here are our recommendations to help drive improvement. Here are the gaps in your business processes today.” We know their domain because the people supporting them either come from that industry or have come up through different channels in that same support group and understand where the challenges are. When we call to talk about analytics, it's more about the insight and the access to data that we can offer. We're entrusted by our clients to understand their processes and systems and present opportunities to drive improvement.

Is it system or human monitoring?

Dave: We are monitoring our clients’ customer care agents or their technicians in the field. It would be a recorded conversation or a recorded chat. Some of what we do is through human listening as well. Much of it is through our analytics engines. We have speech to text analytics that drive insight as well. It’s a combination of automation and then manual because auto QA is a goal for just about every company that wants to know what's going on in every part of their business. While technology is getting there, we're still a few years away from truly having auto QA.

At this point, a company can auto QA about 80% of its business. It's that other 20% that's really tied to customer experience, agent behavior, customer engagement that it can't get out of a tool or a system because it can’t pick up all the nuances. That requires analysts that are trained in that customer's domain.

Many companies have their own QA teams. When would they go to you?

Dave: Sometimes we find there's an internal bias around processes and people. Since we are independent, we have no stake in the agent or the supervisor. We're a third party and can be audited but are also auditing what's going on internally. A lot of companies today outsource much of that work even if they have internal QA teams. We have a robust QC process built in that ensures that we are 99% accurate on how we score things. The challenge in some cases is that companies are now moving toward behavior-based programs which often have a degree of subjectivity that's built in as well. That's where calibrations are required. I'm getting alignment on how those behaviors are defined. Since 90% of the work we do is offshore, we have ensured that our agents understand not only the behaviors that are being monitored but have culturalization training around that. We invest a lot of time and hours in training our agents to understand the countries that we're monitoring. We account for different cultures and different customers. It’s a constant process of refreshing, relearning, and knowledge management.

How does your solution drive improvements in first contact resolution and customer retention?

Over the past few years, customers have become increasingly intolerant of friction in the customer experience and tend to favor low-effort resolution to their issues. Our Customer Interaction Monitoring program identifies both through live monitoring and speech analytics is fine-tuned to evaluate based on the behaviors most valued by customers in today’s market and to identify barriers to quick resolution. We are then able to provide our partners with solid recommendations to address any gaps in agent behavior as well as suggestions on how to improve processes, product, and other elements impacting overall experience and retention. The use of speech analytics in addition to analyst-centered monitoring allows us to capture a varied and statistically valid sample of contacts to analyze and pinpoints areas causing a disproportionate impact to customer experience so initiatives can be prioritized and addressed effectively.

Why do you consider it critical for companies to optimize and update processes to minimize outdated manual functions to meet market demands and achieve peak performance?

The benefit of minimizing manual functions within customer support operations is twofold. First, manual processes have a higher likelihood of error and can be more cumbersome to complete introducing effort and friction into the customer experience, both of which are strongly correlated with poor customer satisfaction and churn. Secondly, manual processes are often more costly to carry out. Reducing them can save money that can either be returned to the business as profit, reinvested in efforts to improve customer experience in other ways, or used to make products and services more competitively priced in the market place. Therefore, reducing manual processes is a win-win situation for the business and its’ customers.

There are quite a few BPOs on the exhibit floor at CCW. How do you distinguish yourself?

Pam:  What I picked up when joining the organization was a strong passion for customer experience delivery and bringing together whatever components are necessary to do that. Our strength is our people and enabling them through analytics and technology. We don't approach anything in a box. We come in, look at the problem, the client, and the environment to create the perfect mix for this client.

helpshift logo


Linda Crawford, CEO, Helpshift. CRMXchange met with the head of this next- gen digital customer service platform which works with many leading brands to implement messaging, bots, AI and more.

In what ways do you think Helpshift excels?

We see ourselves as a company that's revolutionizing B-to-C customer support by unifying the customer conversations into a message-first sort of approach. We offer a full platform that incorporates everything from support for the agents and supervisors through to the end consumer. It’s a super-modern product, powered by AI and bots providing the full range of capabilities expected of a next-gen customer service platform.

When you say full platform for agents and customers, can you expand on that?

We set out to build a platform that would offer a great experience for agents. One that performs all the back-office functions such as managing workflows, queues, and service tickets. We also thought about the end customer and how consumers now expect to be serviced, which translates to ‘respect my time.’ We do that through messaging, whether in an app or elsewhere. For our customers who offer apps, why not service their customers within the app itself and a friendly message-based experience on a website, either mobile or web, integrating the phone into that messaging experience. We’ve all had situations where you need to talk to somebody at some point even though you're trying to self-serve. Thus, we are bringing the power of automation through bots, then coupling that with AI to improve both the self-service experience and the productivity of agents.

Let’s discuss your recent press release about the Connected Customer Conversation. Can you explain what that entails?

It’s unifying everything we do for customers under one umbrella. By connecting digital channels, phone support, self-service, and bots — and by embedding these within a single messaging thread — this platform enables consumers to have one continuous conversation with brands. That translates into providing customers with an experience that respects their time and allows them to switch channels. It also brings every type of interaction regarding support issues into one home and that home is messaging. In our own personal lives, we are using many message-based apps. People are very comfortable with this on-again, off-again experience. Sometimes we're connected and we're rapid fire messaging back and forth to one another and then we do something else because life takes us away, but we have the expectation that at some point if we come back into that conversation, that history of it endures. Whether I sent you a photo or called you, there's some record of that call through the information that exists in the messaging applications. We’re now are bringing that to customer support and giving consumers that same type of experience. The cool thing is that we can bring in bots powered by artificial intelligence into the conversation itself so that the customer, the agent, and the robotic agent can all work seamlessly to resolve questions much faster to respect the customer's time.

Once a customer has started this conversation but gets interrupted by, say an emergency phone call, what happens when they go back in and want to finish what they started?

In today's live chat applications in that situation you described, you’d have to start over because that window just disappears. Either you terminate or the agent terminates and poof, it's gone. In our world, it's picks up right where you left off. It just resumes when you come back into the conversation. We see that as being pretty magical, and if you've interacted with that brand before, you can just scroll up and see a record of all the other past interactions you've had with that brand.

If the customer has been in an SMS conversation, could you continue in SMS?

SMS has its own limitations. Partly because it can cost the consumer something since they are going across a telecommunication channel rather than an internet protocol.  It also doesn't have the same state, which is understanding the sequencing of how these messages are all interacting with one another in SMS. The technical term is 'bucketed', sometimes the messages can get out of sequence. With the advent of Facebook messenger, WhatsApp and iMessage, people and companies can get around the limitations of SMS while still offering it as a great notification channel.

What happens after an SMS conversation if the customer needs to talk to someone?

Part of the Connected Customer Conversations platform is the ability to move across channels. Imagine if you are placing a call and you are told the estimated wait time is 10 minutes. Some IVRs have the capability to call you back. We can offer another option ­­- would you like to go to messaging? This provides a seamless transition from that phone conversation into a messaging experience where we open a messaging experience on the mobile phone you're using. We bring the context of that conversation from the IVR into the messaging experience. If you've already authenticated yourself by punching in your account number or pressing two for order information or order status, that comes with you into the messaging experience. A human or a bot can pick up right where you left off in that journey on the IVR: If either you or the agent says, ‘I think we should just have a live conversation’ that can escalate into something similar to click to call. The record of that phone conversation can come back into the messaging experience once it’s ended.

Are there any best practices that you can share about how companies can start to think about how to best use this technology? Are there certain flows that you suggest?

We’re big fans of the 80-20 rule. Whenever a business is seeking to push automation, it should start with something where the workflow is very well known to the company and where it already has a lot of data. If it’s going to use artificial intelligence, a fair amount of data is necessary to train the model. Companies need to find a couple of very common use cases where they have the data and clearly defined workflows which are right for fully or partially automating. Our customers are seeing incredible amounts of situations for automation which create opportunity for either further reinvesting in their contact centers or simply eliminating costs. I've been in this industry for about 30 years, but this is the first time I’m seeing that not only the agents have a  better chance to work on more interesting things, customers can have a better experience, and the brands have the opportunity for exponential savings. When businesses hone in on these opportunities, it's amazing to see how few issue types represent the largest chunk of the inquiries that they get.

Is that where a lot of the savings comes in?

Some businesses are now saying, “Wow, we previously only could afford to staff our contact center 10 hours a day and now we can do it 24 hours a day.” It’s a combination of the technology with the asynchronous nature of messaging, meaning both parties can pause the conversation. When the brands issue the pause they can say, ‘We've taken this case so far, we're going to get back to you tomorrow when our agents are back.” Customers acknowledge this process and understand when the workflow has started and perhaps progressed three or four steps into an eight-step process. That is one reason why we see customer satisfaction results going up in some cases up by 30-40%. It’s also the intersection of these bots in human workflows operating together. Across some channels such as email, bots may improve it to a degree, but not by as high a magnitude as messaging. When going to a live agent or into a human-oriented workflow, there’s also only a minor improvement. It’s that blend of humans plus bots where the CSAT goes way up. We're encouraging our customers to adopt messaging as fast as they can and start the process of eliminating email. In addition, a good amount of voice can be deflected into the messaging channel.

Do you hear anything back from the agents now that they're working side by side with bots?

Agents doing very rote and routine types of work will be worried about their jobs. If a business has agents reading call scripts, a bot can take on that task. The questions agents may have are “how can I be retrained? What opportunities are available?” This can be somewhat of a scary time for part of the workforce. Most of what we see here in the United States are Tier 3 type of agents, these jobs are likely to remain intact because they're working on the long tail. The issue is more about Tier 1 or Tier 2 agents. What percentage of them can now do more of the pre-sales activities or expand the hours that a brand can offer service? I think there's a lot of opportunity in proactive support if you know specific customer behaviors, not just marketing to them, but anticipating service issues.

lumenore logo


Vanessa William, Marketing and Branding Manager for Lumenore, answered pre-submitted questions about the company’s suite of advanced products focused on performance management, designed specifically for B2C companies.

Can you describe how your Contact Center Solution’s in-depth modules consistently refine and elevate the performance of agent teams?

The Lumenore platform provides a complete view of all business performance, covering important KPI’s, while allowing your team to focus on identifying areas for improvements and determining decisions that benefit your company’s business goals.

Lumenore utilizes a common dataset to feed its powerful platform to visualize, analyze, actionize, as well as monetize business performance.

In what ways does your Business Intelligence Plus platform deliver actionable insights from a variety of sources while leveraging the capabilities of Text, Email, and Voice Analytics Plus?

By unlocking the important business value contained in an email, discussion threads and documents, as most valuable customer interactions, are via calls, chat and social media.

How can businesses benefit from your Training Audit Plus (TAP) classes?

TAP provides remote monitoring/auditing of training programs using a proprietary quantitative analysis of training programs, facilitators and trainees. The Lumenore platform tracks the operational performance of the trainees post-training for a 90-day period to evaluate positive and negative performance/origin via an interactive Agent Performance Dashboard, with all deviations from your curriculum highlighted and video timestamps for ease of viewing.

Can you elaborate on the technology that drives the success of your contact center solution?

Lumenore is a full-stack technology platform and leverages the following technologies - Robotic Process Automation, Artificial Intelligence, Business Intelligence.

nGUVU logo


Pascal LeClerc, VP Marketing & Product Strategy, nGUVU supplied answers to pre-submitted questions about the company’s employee engagement platform, which enables businesses to gamify their contact center KPIs.

How does your NGagement agent journey platform influence agent behaviors by their natural desire for competition, collaboration, achievement, status and recognition?

nGAGEMENT leverages people’s natural desires through Games and Social Media concepts. Intuitively, we all know that games work, adding a game or social media concept to a business process will create that personal connection and build intrinsic motivation and engagement.

Why is it important to create an immersive engagement experience to empower agents to manager themselves?

Offer a completely new way of connecting with their results, with an easy to use UI, near real time results, focusing on objectives and scoring points.

Adding intrinsic motivation to our customers’ existing processes of managing performance, coaching, training, etc. That are considered like “external” motivators. Adding the Intrinsic piece really becomes very powerful.

In what ways does your platform enable managers to identify risks such as employee departures and frequent absences and take proactive steps to prevent them?

Blending performance scores and our engagement indicator allows supervisors to track engagement levels and proactively open discussions with the concerned agents. Future features will use AI to identify trends and predict risks up to 2 weeks in advance.

How does your platform go beyond the boundaries of traditional gamification solutions to prepare contact centers for the realities of operating under today’s changing standards?

Gamification is more than games, it also includes the social media perspective. We are also building AI and ML features that will recognize performance and engagement levels and automatically trigger activities in our platform.



Anand Janefalkar, Founder and CEO, Jason Sophian, Head of Communications, UJET visited with CRMXchange at CCW to shed light on how the company’s transformative cloud contact center software enables businesses to optimize customer communications at every touch point: phone, web and app.

Please tell us why progressive companies should be looking at UJET?

We offer a modern, cloud-native customer communication platform. Without going into details on the technology stack, I think the differentiator is that we understand and meticulously focus on user experience. What we’ve observed is that people now communicate visually and contextually from smartphones. Most services that people use and products they purchase are helping to make their lives simpler. These are accessed via smartphone apps or on mobile sites. Banking's a great example. Many people don’t go into a bank anymore. They can log in with their fingerprint and take a photo of a check. But, I think some support tabs are fully evolved. When you go to the support tab, there's often a slew of phone numbers. If you select the wrong one, it can throw you outside the brand experience. We believe in meeting the customer where they want to be met, whether they're accessing the product or service through the app, through Web, mobile, or desktop. We make sure customers can access a chat through desktop and mobile. Our unique architecture provides best in class, uptime, and SLA. A company should look at us when they're thinking of modernizing their contact center because lot of customers are more loyal or have higher long- term value when being engaged with a product, brand or service through a smartphone.

What new announcements can you share with our readers?

Throughout the four years since our company was founded, we have consistently adhered to one simple premise: that the customer support experience should be exceptional for everyone involved. Over the first two years, we looked at human interaction and focused on three and a half basic personas. There's the end user who is accessing the product or service either through the web, phone, or through an app. The second persona is the agent or customer service representative that is either looking at data CTI integration or simply answering the phone or responding to a chat. The third persona is the contact center manager who is setting up this journey and the interaction points. Then there’s the half part we usually add in. Many times it could be the IT team or the developers or anyone else behind the scenes taking important steps to make sure that this experience is up and running.

We built this experience from the ground up from scratch from 2015 to 2017. We launched in May, 2017. Today we count an incredible array of big brands throughout the world as our customers. These businesses are accessing our solutions to help their customers reach contact centers by phone, web, mobile app and SMS using the power of their smartphones to help resolve their issues. We were also selected as a Gartner Cool Vendor in CRM Customer Service and Support. We just announced our integration with Microsoft Dynamics to go along with Salesforce and Zendesk integrations. We’re bringing Microsoft Dynamics into the mix because we’re seeing more and more companies that are making that transition, especially the ones that are going toward either Microsoft cloud or international. We also announced as our integration with Verint Monet WFM and QM.

What are some of the cool applications you’ve been involved with?

We're seeing great success with on-demand delivery. We're showcasing our solution integrations with our customer Instacart. In IoT on-demand delivery, we’re going be making some other big announcements later this year. We’re focusing on the financial sector as well as the healthcare sector.

How easy is it for a company to get started with working with UJET?

We live in a world where customers and companies alike value instant gratification. A contact center deployment should not take months. We've been known to do contact center deployments for 200-300 seats in under a week. We’ve also been able to integrate with the platforms or their underlying CRMs in less than a day and we have had companies integrate our mobile SDKs and web STKs and under two days. That’s why we include the developer persona as a part of our focus. It’s very easy to get started with our contact center in a box. Businesses can enable various channels. Some companies love our end-call SMS feature. When people are engaged in a call, they can communicate visually and contextually without an app via SMS. There are companies like Wag! which has nearly eliminated SMS because they see a great interaction model with the UCF SDKs in their app. Whether they’re dealing with a walker or a dog owner, we have a metric that the engagement and delivery team can focus on. We’re seeing and being told by our customers is that our platform is best in class and they've never seen the ROI or time-to- value realized as rapidly as they have with our solution.

What kind of team do you see on the customer side? Is it people from IT, marketing, or the contact center?  How do they put together an effective group?

I think the budget planners and the stakeholders are from the contact center side. They’re the ones that are feeling the pain daily and everyone might have a different KPI or metric that they want to improve. Some might say ‘my handle times are too long’ or ‘CSAT scores are too low’ which might be because of transfers or dealing with repeat callers. Another issue might be that agents are not armed with the right information. Other people from the product side might get involved, especially if it has to do with the mobile app or the website. There might be people from the IT team, whether it’s an internal team or an IT team at the BPO or the in-house or outsourced contact center. We’ve also seen cases where execs who are invested in the image and perception of their product, brand or service not only get involved but are excited about how it can help realize their vision.

Where do you think it’s all going?  Do you see contact center operations continuing to gravitate to self-service, with human agents being less and less involved?

The shift to automation has been a consistent trend. We see a percentage of queries that can be elegantly solved with either structured FAQs or provided to a customer through an AI or a chatbot. We believe that controls in assisted conversations to resolve complex issues will continue to reside with agents. The agents will work on interactions that either require empathy or human decision making, while also focusing on retention, loyalty or upsell. Having an assisted AI platform will be a game changer as opposed to agents still having to perform rudimentary tasks. With our platform, the ability exists to obtain data in real time from an unlimited number of systems without needing to keep the customer engaged while agents locate the information. Using record locators in the CRM to thank a customer for their loyalty and their purchases is a far better way to start a conversation than needing to ask for personal information.

Is there anything more that you would like our audience to hear?

We are grateful for having had a banner year in 2018 when we grew our business threefold and expect even greater growth this year. We hope businesses that are just starting to think about digitization explore the opportunity for greater engagement through digital channels. We urge companies thinking of modernizing their contact center to look at our customer case stories, which show a 25 to 30% improvement in TCO, as well as significant NPS increases.



Holly Simmons, Global Product Marketing, Customer Service Management at ServiceNow offered her input on how the company’s “Now” platform transforms outdated manual ways of working into modern digital workflows to ensure that employees and customers get the information they need, when they need it.

Can you tell us a little bit about ServiceNow and why companies should be interested in your offerings?

We focus on making the world of work better for people. That applies to every part of the business -- helping to connect employees together, to connect employees with customers, as well as to manage all the work and what we refer to as ‘digital workflows’ throughout a company. We started in IT, extended into HR and customer service.

We are now hearing a great deal about customer retention management. Is this a trend?

Everyone is familiar with traditional CRM systems in the customer service space and the intent has always been to capture customer information, develop a view of the customer, and track cases. While that is important, at the end of the day, it doesn't fully resolve all a customer's issues, nor does it necessarily support making people want to continue to do business with an organization… that requires an effective retention piece.

In our conversations, CRM is really ‘call reaction management,’ focused on agents waiting for customers to call with their issues.  What’s really needed is an approach that is end-to-end, focused on the entire customer journey with a company. Helping companies retain customers requires being more proactive, often finding ways to solve issues before customers start making inquiries. Every interaction is no longer an isolated engagement.  It is about more than capturing a single issue. It's part of an ongoing relationship with the customer. Everything a business does to build that journey and connect all the pieces together makes for a seamless experience every time the customer's interacting with a company.

How important is exploring the addition of artificial intelligence bots and self-service?

A lot of the press coverage over the last few years goes down a negative path, portraying AI as some evil technological weapon designed to take jobs away and eliminate the human element. The truth is, it can be a very positive force not only for customers, but also for the agents and others within a company who are trying to deliver the best possible service and support for customers. If you look at machine learning, we can automate many common requests allowing agents to focus on more interesting and meaningful work. They can become advisors to customers and use their time to help customers resolve more complex issues. It’s not about eliminating jobs but changing and improving what agents can do, which creates a higher level of job satisfaction and makes their work a far more engaging activity. Automating mundane tasks ultimately translates into positives, not only for agents, supervisors, and customer service leaders, but also for the consumer.

With rote and repetitive tasks deleted, do you envision any change in the type of person that becomes an agent?

We hear the term “super-agent’ being used more frequently. That is an outgrowth of the growing opportunity to do work differently. There probably will be more problem-solving required, but there's also unprecedented potential for agents to learn new skills. It’s important for leaders to put the right training and support mechanisms in place to ensure that as many people as possible can make that type of transition. And I think most agents would be excited about the prospect. They would prefer to see repetitive tasks automated and use their love for the product that they’re supporting to help customers get more out of it. 

Where would ServiceNow fit in to help the agent?

We provide many capabilities to empower agents.  This includes an agent workspace for omnichannel engagement and case management that takes advantage of AI for routing requests to the best agents and exposing related cases and solutions to problems.  It eliminates the typical ‘swivel chair’ activities for agents.

We have extremely robust workflow that creates a connection between customer service and other departments. A customer service agent can engage someone in billing, legal, or another part of the organization, and can even work with an external partner to help them resolve a specific problem. They can assign specific tasks that foster increased visibility. The workflow can not only manage all these functions, but also automate a lot of processes. 

We also support service management and service operations, which greatly helps companies selling and supporting technology, digital products, and services. We help them monitor customers’ products and services, as well as identify, diagnose, and resolve issues.  Incorporating necessary changes based on customer input results in continual improvement over time of the product. It also reduces the number of issues that arise for other customers. Agents don’t get the same inquiries about a bug in the software because the organization is able to resolve it. We help companies avoid outages and when a problem is identified, agents can proactively notify customers who might be affected. There's no greater feeling of accomplishment than being able to proactively fix an issue before a customer contacts you.

We offer many opportunities backed by technology and intelligence to help the agents. A company can layer on machine learning to both prioritize requests coming in and identify resolutions for one specific issue or something affecting thousands of customers. Businesses can add chatbots to help customers without human involvement, but when necessary, route the case to a live agent. These types of efficiencies are valuable, but they also lend greater visibility to customer service teams. Through their expanded role, they are engaging the entire organization to make it everyone's responsibility to deliver a great customer experience.

If the customer needs to talk to an agent after they’ve received a text or self-served, is there a flag created for the agent that gives them all the information?

Sometimes the company can take the customer directly to a personalized portal which has all the detailed information specific to that customer.  It may recommend knowledge articles because it knows who the customer is, what products or services they have, and what cases were opened. The system could even identify postings from the community that are related to the customer’s products or issues reported. The business can bring all that information together for the customer.

There's seems to be a great deal of flexibility as to what part is machine learning, what part is automated, and when a customer needs an agent. Is that determined by the specific situation?

Yes there is a great amount of flexibility because every business and industry is different. For example, I was talking with a customer service leader who runs an emergency response center. She laughed when we talked about self-service for her business because, in an emergency situation, the callers wants to speak to a human. Alternatively, you have technology companies selling services like clothing or shaving subscriptions.  These are digital services and customers don't usually need to speak with anyone.  Self-service is the desired engagement model.

What can we expect from ServiceNow in the near future?

We’ve made several investments in AI and mobile and will continue along that path. We’ll continue to build out an amazing agent workspace and agent experience, which a major focus for us. This goes back to the overall theme of making the world work better for people - agents, and other departments in the organization. We're integrating more third-party data and applications in our agent workspace because a productive customer service agent shouldn’t have to have two monitors and 72 applications with which they have to engage. Our goal is simplifying their job and cutting down the time needed to help them achieve resolutions by giving them access to information that they might not find easily on their own.

semafone logo


Gary Barnett, CEO, Semafone, sat down with CRMXchange to talk about how the company provides simple, fast, cost-effective PCI DSS compliance and data security for contact center environments.

Gary, would you like to tell us how you got involved with Semafone?

It started when I did some work for Semafone’s board of directors late last year. I've been in the contact center business for a long time, but one of my greatest passions is bringing the great add-on capabilities to the market. Data security is one of those great new disciplines to which every company must adhere. I simply fell in love with what the company is now accomplishing and how well it fits the direction of the marketplace. It's also unusual to have a company that started out internationally and is expanding and moving to the U S.  After I began working with Semafone in October, 2018, I wound up joining the company as the chief executive officer this past February. It's been a fantastic experience to work with the team. 

Can you elaborate on what the solution is designed to accomplish and why businesses should be interested?

It’s quite easy to explain. Most consumers know what it's like to pay through a contact center. We can call an airline and when they respond that they can book a reservation, the next question is ‘what is your credit card number?’ The issue with that process is that it’s not compliant with PCI regulations. It leaves a person’s credit card number exposed. The agent or someone sitting next to them in the contact center hears it and could easily copy the numbers. Companies also store multitudes of credit card numbers which can and have been breached. We’ve all read articles in the Wall Street Journal about millions of credit card numbers escaping to the open world. What Semafone does is enable contact centers -whether on premise or cloud-based – to reduce risk and take secure credit card payments on the phone and through other channels as well. One way to protect sensitive information is to ensure none of it is stored. You can't steal what you don't store. Even though no information is kept as available records, the consumers’ payment is processed.

Here’s where our difference comes in. The agent confirms they can take an order, but instead of asking the customer to read credit card numbers, they instruct them to enter it on their phone’s touch tone pad. While the voice interaction can continue as the customer enters the information, the agent never hears the DTMF tones and has no way of finding out what credit card number the customer is using and the sensitive information is never stored. Our solution enables the number to be processed to the payment provider, at which point the agent can verbally confirm that it went through. As such, an agent can thus provide exceptional customer service with no immediate or long-term security risk.

While we’ve positioned ourselves as an industry leader in secure voice transactions, we've also moved into the digital world. When a business asks a customer in a chat conversation to make a payment to place an order, it’s understandable that the customer would be concerned about entering sensitive information.  Our solution enables the agent to reassure customers that they will be put into a secure transaction where no one will be able to see what credit card number is being entered. Just as in our solution for protecting voice payments, the agent will be aware of the customer’s progress and if the payment has been accepted and nothing more. Customers have the same level of security in SMS or email or even social media.

What does the consumer who is told they are in a secure payment environment see during the transaction?

We employ a very clever user interface which we put a great deal of thought into creating. First and foremost, it must be secure. Second, we want to make it an enjoyable experience with a certain ‘wow’ factor involved. The agent remains with the customer but not in the same chat window while the information is being entered. The customer sees a picture of a credit card and can fill their numbers in on the surface. When it's time to type in the CVC, the credit card visual rotates to where they enter it. Simultaneously, the agent on the other end can verify that the customer has successfully entered the credit card number. None of the credit card information is ever visible in the agent’s chat window.

Of course, our customers still get audited and are required to get their PCI DSS compliance certified. However, it's much easier for them to receive because we’ve separated all secure information. Therefore, when the QSA examines the auditing, they’re not concerned with agent accountability or possible breaches. Our customers like it because we centralize the security as opposed to it having to touch 20 different systems and our clients don’t miss a second of the recording – as it’s the same in chat transactions.

Does using your solution make information less hackable and more secure than if the company was taking the information?

In terms of securing credit card data, the saying we like to use is “They can’t hack what you don’t hold!” Semafone’s Cardprotect solution prevents cardholder data from ever entering the business infrastructure, so organizations do not have to worry about hackers accessing it—as there’s nothing for hackers to get ahold of in the first place.

Once a transaction has been done, it’s gone?

We ensure there is no long-term storage, and in the fleeting moment when a number exists as the system processes payment, we are absolutely protecting that transaction. Once it's over, our customers don’t need to think about it again. 

What kind of companies are best served using your solution?

Excellent question. We see conservative businesses such as insurance companies that take payments on the phone or through digital transactions that want to make sure they are fully compliant and not just PCI DSS compliant. We work with other companies such as retailers which have a very high volume of payments on the phone or through the contact center. We also work with healthcare organizations which need to do more than simply protect credit card information but also secure member IDs or social security numbers. There’s one additional interesting category; companies which have been hacked and having learned their lesson, want to ensure that it never happens to them again.

When I’m shopping online, even if it's a site I use frequently, I always decline when asked if I’d like them to store my credit card information. Do you see business ending that practice to limit their liability?

While some people see that as a convenience, enabling a company to keep your credit card on file would only be acceptable under very limited circumstances. We would certainly highly recommend that those companies use a process called tokenization. In that scenario, the card information goes to the payment provider and the payment provider comes back with a token. If someone were to steal them, the tokens would almost always be useless. No one should even consider allowing a business to store their credit card numbers in the clear. The next worst thing is the storage of credit card numbers that are encrypted. Our advice is don't store at all or at the very least ensure that information is tokenized by the payment providers.

When I’m shopping online, even if it's a site I use frequently, I always decline when asked if I’d like them to store my credit card information. Do you see business ending that practice to limit their liability?

While some people see that as a convenience, enabling a company to keep your credit card on file would only be acceptable under very limited circumstances. We would certainly highly recommend that those companies use a process called tokenization. In that scenario, the card information goes to the payment provider and the payment provider comes back with a token. If someone were to steal them, the tokens would almost always be useless. No one should even consider allowing a business to store their credit card numbers in the clear. The next worst thing is the storage of credit card numbers that are encrypted. Our advice is don't store at all or at the very least ensure that information is tokenized by the payment providers.

kryon logo


Richard French, Chief Revenue Officer, Kryon answered our advance questions on how the company’s RPA Solutions can help businesses provide faster, more personalized, and more effective service

In what ways is Kryon using Robotic Process Automation to help contact centers meet constantly rising customer expectations for personal service that rapidly resolves their issues?

Robotic process automation (RPA) can play a pivotal role in driving transformation within call centers, bringing greater efficiency to their operations while improving customer satisfaction. The impact of RPA on call centers can be seen through KPIs such as first call resolution, average handling time, customer satisfaction score, employee error rate, average revenue per user, customer retention rate, employee attrition rate, and time to proficiency. Customers expect to be serviced efficiently and promptly, and RPA not only reduces the time a customer is on hold, but also allows for fast data collection and the ability to bring up previous call history and account information.  

How does RPA empower agents to learn more quickly, access the most relevant information in real time, navigate often complex applications and deliver consistently reliable service?

Kryon’s uniquely business-friendly RPA solutions make it easy for contact centers to identify and leverage their most promising opportunities for automation. Our intelligent Process Discovery solution enables them to quickly, reliably, and efficiently find the processes best suited for RPA, while our variety of automation solutions means that each business can choose the right tool for any process.

For example, Kryon attended robots guide novice agents through proper handling of tedious tasks. Kryon sensors accelerate order entry and provide a script for agents to confirm before completion – saving time and improving customer service.

Education is a key component for agents and through Kryon’s ‘Guide Me Mode,’ agents are able to learn quickly and efficiently, freeing up other employee times, reducing the onboarding process. In addition, RPA assists with the wrap of calls, pulling all information together onto a system so that other employees can follow up and understand exactly what the customer needs. This automatically happens in the background remotely, so an agent is able to move on to the next call more quickly.

How does your solution help contact center operations cope with high employee turnover rates while also reducing the need for hiring more reps for short-term increases in call volume?

Instead of requiring employees to perform tedious and burdensome processes, today’s contact centers are increasingly turning to robotic process automation. With RPA, a software robot can either automatically perform a process assigned to it or provide a CSR with helpful guidance and assistance in real time. This way, CSRs can offer faster, more personal service with fewer distractions – and they can boost sales as robots alert them to promising opportunities for upselling and cross-selling. RPA is key to reducing employee error, ensuring that the wrong information isn’t entered incorrectly. In addition, employees are able to achieve higher sales quotas, as screens can pop up and identify a customer who might be a great fit for another product so that the agent can pull information quickly and sell it in as a recommendation.

Can you tell our audience about some of the tangible benefits that businesses can derive by implementing your solution in their contact center?

Kryon’s Robotic Process Automation has had a tremendous response from businesses who have implemented RPA, including higher sales, reduced handle time, heightened productivity and morale, increased accuracy, enhanced customer experience, and lower operational costs.

Some use case highlights: 

  • A leading telecommunications provider cut lost revenue by 14% for new hires and 5% for tenured agents – while reducing new hires’ time to proficiency by one week, improving service delivery time, and preventing incorrectly entered work orders.
  • A major insurance company cut average call times by 70%, reduced average handling times from 10 minutes to 3 minutes per call, and decreased operating expenses by 20%.
  • A large bank saved over $600,000 in employees’ annual work hours, while significantly shortening wait times to offer an improved customer experience.
procedure flow logo


Micah Peterson, VP of Product Management, and Amanda Verner, Marketing Manager, ProcedureFlow offered details about their innovative training solution designed to visualize the way subject matter experts think to help guide new agents through the process of learning complex procedures.

Please tell us about the inspiration behind ProcedureFlow and what your solution can do to help successfully compress training time.

Micah: I'm from eastern Canada which has been a hot spot for call centers since back in the 90s. There was a company doing tech support outsourcing work there, which had an eight-week agent training program and a 900-page standard operating procedure binder. When agents finally completed that ordeal, they'd sit in their cubicle and after taking the first couple of calls, many would start crying. It was horrible. The company was missing SLAs (service level agreements) and not meeting expectations. They knew they needed to radically change how things were being done, so they built their own highly effective performance support tool. This took training from eight weeks to a two-day orientation. The agents were a hundred percent confident to start on that second day.

When was that?

That was around 2001. The company kept the solution for themselves as a competitive advantage for about 10 years.  In 2011, I joined the company and we turned it into a cloud-based product, then spun it out as its own company in 2015. We’ve experienced lot of growth in the last few years. 

It seems like there would be more involved than simply taking all the information in a manual and putting it online. Can you walk us through how it works for the agent?

First, there is the challenge of trying to train people. The situation in most call centers is that they have about 12 months before agents either get promoted, are managed out, or leave on their own.  It's a very short window for the call center but many apply a traditional training model that takes six to 12 months for agents to become proficient. That kind of timeframe is a luxury because of the high turnover rate. When it takes 11 months for someone to become fully productive, who’s then gone a month later, that’s a broken model. At ProcedureFlow we're totally changing the way businesses onboard agents. We offer a program that enables new hires to self-serve through the training process. The only way that can be done is through better documentation. If you look at the way most documentation is set up now, it’s in paragraph format, like a book. When an agent is on the phone with a customer, they don't have time to read paragraphs of text.  The other alternative is often a six-month traditional training model to create visual cues that the agents can use while they're on the phone with the customer. Accomplishing that in a shorter timeframe is what ProcedureFlow is about. Our program takes written paragraphs and turns the information into small hyperlinked flow charts or visual colorful guides where the colors, shapes, and arrows have meaning. The agent can easily use it while on the phone with the customer. 

If the customer has a question and the agent's not sure of the answer, how do they pull up that information?  Do they find it through keywords?

That’s a great question because efficient searching has always presented a major challenge in knowledge management. There are two key questions; how do agents find what they're looking for and can they read and interpret it? The first one is difficult, especially if they’re new and don't really know what to look for. The second is an ongoing challenge, can they consume and communicate information on the fly? ProcedureFlow eliminates both these issues. Agents won’t need to search for anything. We help them from the moment they initially take the call; guide them through the authentication; through their issue; through probing questions; all the way to what standard operating procedures they should execute. This is all represented in the flow. The agents can self-serve all the way through the call. Since they never have to search for anything specific, agents can just flow right through it. 

With agents following a complete Procedure Flow, what it is the impact on training? 

From a non-trainer’s perspective, it would be a way to cut the duration of the program. We can cut a business’s training time by 50 to 90%.  If the existing training model is not changed, training will be around 50% shorter. To diminish the length of the other 50%--that lecture style piece-- a business needs to change the way it does training to be more scenario-based. Once agents are brought into the classroom and immersed in ProcedureFlow, trainers need to ask “what is the best way to provide agents with workable scenarios involving customers, their common issues and the problems they wish to solve?” Once the agents understand the flow, new agents can just jump out and start doing the work. We can cut training time by up to 100% if the organization is willing to take these steps.

When you say up to 100%, you mean there's no training needed?

That's right. If you go back to the original company that built this for themselves, they went from eight weeks of classroom training to a two-day orientation discussing simple things like here's your badge, here's your cubicle, the rest room is down the hall. Then they brought the trainees into Y-Jack calls that agents were taking. With no effort or expense on the organization's part, beyond the Y-Jack set, they were able to determine the system's working, hear conversations and have ProcedureFlow on their own iPad or laptop.  When new agents feel confident that they’re ready, they'll jump into the cubicle and get going. That is a full 100% reduction in training time. 

Do you have any feedback from agents who’ve been trained in other environments vs being onboarded with ProcedureFlow? Have there been any surveys to learn about   differences in the process and how it affects their motivation and preparedness? 

There are several different types of agents who could make comparisons. Agents who worked in call centers with traditional training and later experienced ProcedureFlow. What we’ve heard anecdotally is seasoned agents see an incredible difference and can barely believe how much easier it is with our solution. Those who have no call center experience can’t really conceive of how training was previously done. We’ve also heard stories about agents who’ve been through ProcedureFlow actually building flowcharts on their own time when they come to other companies that only offer traditional training.

Would it be accurate to say, agents trained with ProcedureFlow don't need to be product experts because the solution takes them through everything and gives them 90% - 95% of what they need to know?

I like that question. A few years ago, I was thinking about why contact centers exist and had a revelation about the human brain, which is the most amazing machine in existence. If you think about the contact center as being analogous to a human brain, it exists because processes are always changing. If they were stagnant, you could automate everything and not need humans. We put humans in the mix because they're so adaptable to change, and these human brains drive the contact center. 

With Knowledge Management, you're always trying to capture a flow of what's happening in the brain of your subject matter experts. We did a pilot program that involved mapping out the knowledge of someone with extensive experience handling complex telecom issues. We had a new person working with us who commented “it was a privilege to be a part of this pilot because I felt like I was staring into the mind of somebody who had been doing this for years.” Using small hyperlink flowcharts is almost like diving into an expert’s head and navigating their brain. When new agents deal with exception handling, this knowledge enables them to surf through the mind of a subject matter expert who can deliver the highest call quality, lowest average handle time, and greatest customer satisfaction via ProcedureFlow. The business is thus enabling 100, 300 or 1000 agents to operate as efficiently as its best person.

It seems almost like cloning new agents to succeed. Do you find when they’re more comfortable and able to help customers that they can be retained longer?

Aside from providing a high salary or unionized benefits, I’m not sure any business can keep a high percentage of the call center agents longer than two years because people get bored and want a new challenge. Most attrition is generated in the first 30 -90 days, and that is a result of people just being overwhelmed. We’ve seen situations with our customers where people left after a few weeks because they thought they could work at a coffee shop and make the same money without constantly being attacked by QA people. If you give them confidence and can impart the knowledge of the best agents, it can be a game changer for early attrition. 

I would imagine that the quality level is consistent across the base of agents in any company using ProcedureFlow because they're following that same flow and using the same information.

There’s a funny story in the book “The E-Myth Revisited” by Michael Gerber about his experience with a barber. On his first visit, they put a coffee in his hand, tell him what they’re planning to do, trim his hair and get everything done quickly. He’s impressed and goes back again. This time, no coffee, but the haircut itself is still good. He thinks perhaps they just forgot. He returns a third time, still no coffee but they suggest shampooing his hair first. In all three experiences, he received a good haircut in a reasonable amount of time. He stopped going because what he ultimately wanted was consistency. This brings up the hypothesis that a business could have 20 experts in a 100-seat contact center and a customer could talk to two different experts in separate interactions and get totally different answers. Even though the people are experts, by giving different answers, they’re giving the customer a weird sense about the brand. Consistency is so important. That’s why we believe it’s important to want to bring your agents into one process that everyone agrees upon and have everyone execute. No matter which agent the customer calls, they're always going to get the same story. 

We have reporting data that shows how the agents are using the flows. Higher level executives can see the start of processes and ask the team to provide the best path. A company can consult its legal team to ensure that the right disclosure is being used and if the direction mapped out is logical. The marketing group can ensure that a program is not too confusing for customers. The business can work with different departments to evaluate what has been visualized. Having a number of eyeballs look at plan being mapped out to do continuous improvement would be common practice for any Lean Six Sigma organization. ProcedureFlow provides a vehicle for continuous improvement right out of the box. 

Is there anything we didn't touch on that you think is important for our audience to know?

I consider myself an industry outsider because I never really envisioned getting into the call center space. I don't know if many people start out thinking as kids of being involved in this industry, but I'm in it now and I am enthusiastic about improving customer service.  From my vantage point, when I come to conferences like CCW and see hundreds of vendors talking about omnichannel, chatbots, AI, and other technologies that are mostly cost reduction tools, I'm at a loss.  I believe the game is expertise. When customers call in, will they be speaking with reliably knowledgeable agents? Does it make sense to spend so long bringing agents to proficiency just to see them leave? That's a big problem which few companies are trying to solve. Instead of trying to better use human capital, they’re looking to deploy expensive technological solutions that aren’t always as effective as properly trained human agents. If a company can onboard a successful agent who feels comfortable in the role after a week or two and remains productive for nearly a year, doesn’t that improve the customer experience for voice, chat, or other omnichannel touchpoints? I believe the biggest challenge for call center operations is developing subject matter expertise in a consumable format.

observe ai logo


Swapnil Jain, CEO and Sharath Keshava Narayana, CRO, Observe.ai discussed how their agent-first voice AI platform differs from traditional speech analytics software which is built on text transcription. The company uses what they call a SpeechNLP approach to perform a deep analysis of the combined audio and text stream to detect patterns in the tone of the speaker.

Let’s talk a bit about Observe.ai, can you tell us how you got started and what you saw as an unfulfilled market need?

Sharath: I had a background in product research and was working in Manila in the Philippines which was perhaps the Mecca of the contact center world. At one point it had more than two million agents in one metropolitan area. Most of the techniques for speech analytics were archaic.  I became aware that there are many operational advances in a contact center that could be augmented or automated with the help of AI. After seeing things in that context, I believed there was a need in the contact center ecosystem for a true speech analytics technology company with a different approach. Suppliers building speech analytics solutions were simply applying text transcription technology, which is the way things have been done for the past 20 years. They would then utilize some keyword spotting algorithm. Some would even do some form of Natural Language Processing, but the program was still primarily text-based. Our hypothesis was that businesses needed to measure customer experience by listening not only more closely to what customers were saying but how they were saying it. Our interest was building a true speech platform to manage customer interactions, one which helps the contact center, the agent, and ultimately the customer as well.

When you say a ‘true’ speech platform, what makes yours more “true” than existing solutions?

Swapnil: When you hear the term ‘speech analytics’ in a conversation, there are so many companies essentially trying to say the same things. The big difference comes into play when you look more closely at the current conversational AI systems employing speech analytics. It seems like these suppliers are trying to build conversational AI for text. Now the industry is trying to take the same model to voice. We only focus on voice, we don’t do multichannel.

What they’re trying to do is use the paradigm of only taking into consideration what people are saying...the content of a conversation. In the world of text-based speech analytics, the meaning of statements made in contact center conversations lacks context. For example, when a customer makes a statement like “Hey...can you please tell me when I might get my refund?” it sounds benign.  But what the customer is saying is really “Hey…can you please tell me when I might get my refund!” While the two statements are essentially the same when the conversation is based on text transcription, where voice intonation is stripped out, and the business completely misses out on the context of what is happening. This is super-critical because the personality integration puts calls into context. Because contact centers are all about the customer experience, understanding intent and customer emotion --which comes from tonality-- needs to be part of the equation. That is first big difference we believe puts us ahead of our competition. The second difference is as we noted, all these conversational AI systems are built on text and they’re not updating their model. If you think about humans communicate with businesses today, we are looking at three primary media: chat, email and speech. In chat, perhaps the most structured form of content, many write properly grammatical sentences, put in full stops, then check what they wrote. I’m sure that chat is weaker than email in that respect, but still fairly structured because it is written in sentence format. But when it comes to speech, it’s a very different story. Just imagine how humans often speak. They don’t always complete sentences and the grammar is wrong.  They cut each other off in conversations, so when a supplier tries to take the same model in interpreting speech that it does for chat and email, in effect, the model isn’t even sure that it’s the same language that is being spoken. A text-based model can’t make sense of conversational data when sentences are being cut off and there’s no full stop. But if you think from an NLP perspective, why should this be?  Now a business can move beyond the existing models with a solution that focuses on the audio signals as opposed to simple text transcripts. When we say true speech, we mean that we are a pure-play speech company who brings proven expertise to giving our customers the full context of conversations. 

Our technology is backed by a chief scientist from Samsung with a strong research background who runs our AI group. He is the founder of a product called Samsung Bixby which is a competitor to Google Home. Our speech-to-text spatial accuracy is at least 20% higher than Amazon. That’s how far our technology is taking us. 

Let’s talk about AI compliance and QA. You state that you can do QA for a 30-minute call in less than a minute: how do you accomplish that?

Sharath: In most cases, looking at current contact center QA processes, there’s one QA person for every 20-30 agents. That specialist’s job is to pick up and analyze the call recordings. When all calls are recorded the QA team might pick two to five calls per agent per month to review. What does that mean? They have a checklist of questions that they need to listen to and determine whether the agent did or did not perform certain required functions. There is no longer a need for businesses to do this manually. Our technology can understand everything that happened in a call and answer such questions automatically. Even if the solution is not able to immediately supply the answers a business seeks, we have an advanced audio player which marks the elements of a call that need to be reviewed. For example, a collections agency can ensure that agents followed the proper protocol when contacting an individual to collect debt. A QA person listening to call and filling in sheets for 30 minutes can instead look at the call player, see the transcript and find the needed information annotated in a mini-window, ready for review. In 30 seconds, the QA person can locate the call elements they need to review and be done, making the process infinitely more efficient. This has been proven time and again with our customer, some of whom have reported improvements of making the process as much as five times faster.  

It would seem to be easy to determine whether a company is in compliance with regulations. The AI can be trained to say yes or no to certain requirements. How does it work when an agent disagrees with the QA’s findings? Can the company listen to the call to prevent a long back-and-forth?

Sharath: The company can use the call player to show the transcript and play both the agent’s and customer’s part of the call separately if necessary. It can quickly go directly to the selected points in question and has the advanced capabilities to review the call much more quickly. It’s an always-on system that listens to every single call, so no company need worry about picking up their sample from a group of randomly selected calls.  Everything is transcribed and annotated and ready to be reviewed. 

The company can select all the statements they expect to be used in every single conversation and sample the calls that do not have these required points covered, which is one of the first ways they can save time. Another time-saver is that it is not necessary to listen to the entire call. They can review the moments within the call that they wanted to track, and do it up to five times faster.  We’ve also built in a feedback loop system. If a supervisor sees something missing in a call, they can immediately forward details on that instance back to the agents in question. The feedback is very qualitative, it can pinpoint exactly where an agent has gone wrong in a specific section of a call. When they receive it, they can clearly see how they might have mishandled the problem analysis or not properly captured the next series of steps. 

That feedback can be given in real time or delivered subsequently depending on the company’s resources. Our technology is not limited by real-time or non-real-time constraints, but it is dependent on the infrastructure of the business to which it is connected.  

Since we’re talking about integration, what platforms are you integrated with?

Swapnil: At this point in time we are integrated with companies such as Talkdesk, Genesys and Mitel, but we could integrate with any telephony system if they have an API. We have APIs with Cisco, Avaya, LiveOps and Five9. But with Talkdesk, Mitel or Genesys it’s a one click download to implement Observe.ai and the next minute, all the company’s calls are being analyzed.

Have you been able to track results yet in the companies that are using your platform? What types of benefits have they been able to realize now that they have been using it for a while?   

Sharath: It primarily depends on the use case. So far, the core use case has been companies telling us that they had a team 50 QA specialists to monitor agent activities and that after six months of using Observe.ai, they have been able to reduce those teams by half. So, the solution is starting to pay off by itself with far lower QA costs. One of the world’s largest auditing firms is using us to help predict customer churn. Since we have an always-on system, they want to use our platform to review every call to help them determine the probability of each customer churning. They use that feedback to find ways to improve the product or call back the customer and see if it can be handled more to their satisfaction. We also work with a major travel firm which uses us to help qualify a lead. If our platform categorizes a call as a hot lead and it hasn’t been closed, then they call back with a discount offer.

As we wrap up, is there anything more you’d like our audience to know?

Swapnil: As an engineer, I hear a lot of AI buzz in the market. But while everyone is talking about it, we’re still far away from total automation. Thousands of blogs warn about how jobs are going away but the people who write these blogs don’t really know anything about their subject matter. We’ve started a company that is succeeding right out of the box by using advanced technology to make measurable improvements. 

In collecting data for call recordings, we help businesses provide the support their customers are looking for. The data is available to be used to build a blueprint of how to improve a customer service system. It's collecting data and then moving on to augmentation. 

Our process is to go from data collection to augmentation. The question is ‘how do you bring a human and an AI entity together to provide amazing customer service?’  We’re not ready to claim total automation because it is still too early in that journey.