The Futureof ContactCenters

Leveraging People,Processes & Artificial

Intelligence

Matt Rocco

Matt Rocco is the Presidentand Chief Operating Officerfor Etech Global Services.

Matt is a 32-year veteran ofthe call center / BPO industry.He has held key leadershippositions within Dun &Bradstreet, The Berry

Company (a subsidiary of Bell South), Etech, Inc.and Etech Global Services. Matt has spoken atmany industry events and has been featured inarticles in numerous periodicals including CIOReview, Call Center Magazine, Call Center Times,Connections Magazines and others.

Matt started as a telephone rep for D&B in 1984and now serves as President of Etech GlobalServices. In the past 32 years, he has spent time inevery facet of call center operations and outsourcingprocesses.

Since 2003, he has been the President and ChiefOperating Officer of Etech’s operating companies, amulti-national outsourcing organization with over3,000 employees that provides world class solutionsto Fortune 500 companies from its locations in theUS, India and Jamaica. Under Matt’s leadership,Etech has received numerous industry and

community awards including: Top 20 MostPromising Quality Monitoring Solutions Provider –CIO Review, Professional Association of CustomerEngagement (PACE) – Technovation Award,PACE – Corporate Citizen Award (4 times),GEISA – Best Call Center / BPO Award, Top 50Minority Owned Businesses – Diversity.com, USChamber of Commerce – Corporate CitizenshipAward and many others.

Email: Matt.Rocco@etechgs.com

Jim Iyoob

Jim is the EVP of CustomerExperience for Etech GlobalServices. Jim hasresponsibility for Etech’sstrategy, marketing, businessdevelopment, programimplementation and productdevelopment across all ofEtech’s existing lines ofbusiness – Etech, Etech

Monitoring Solutions (EMS), EnterpriseTechnology Solutions (ETS), and Etech SocialMedia Solutions.

He is passionate, driven, and an energetic businessleader with a strong desire to remain ahead of thecurve in outsourcing solutions and service delivery.

Jim has 25+ years of contact center outsourcingexperience in inbound, outbound, chat operations,and social media management spanning across theglobe; domestic, nearshore, and offshore. Hisexpertise includes operational performancemanagement, e-commerce sales and service with animpeccable track record of innovation and advancedbusiness intelligence. One of his strongest suits isbeing able to assist customers in launching and

developing their live chat operations into a world-class customer experience solution.

Jim has an exceptional understanding of thecustomers’ products, their requirements & processes,and a complete analysis in order to provide atailored solution. This has helped him develop,implement, and retain some of the most successful,award winning outsourcing solutions that deliverconsistent ROI. Due to his level of expertise, Jimhas been featured in numerous publicationsincluding CIO Review, Contact Center Pipeline,Connections Magazine, and was also listed as oneof the ‘Top 20 Influential People to Follow’ onTwitter by “Iwantitnow”.

The gained experience over the past 25+ years hasenabled him to be a subject matter expert for callcenter solutions.

Email: Jim.Iyoob@etechgs.com

Foreword, Mike Mings Co-Founder, Tethr

Speech analytics solutions must move beyondlegacy simple transcript search and sentiment. Thenext generation call center will leverage AI to mapword and concept level relationships within humanconversation and then leverage these insights toproduce business relevant intelligence or insights.Further, these models must be fluidity dynamic toevolve or continuously learn. This is where machinelearning contributes to an evolving customerspecific domain of intelligence. The resultingsemantic intelligence and based on deep learningwill produce continuously improving predictions.At Tethr, we see current industry solutions lacking asufficient level of intelligence depth. Currentlyindustry solutions simply produce superficialreporting based on statistical analysis. Customersare asking for and need better understanding thatcan be applied to improve their business. Theindustry will not only simply detect information, butwhat we are saying what it means to your business.The market is telling us that traditional speechanalytics and sentiment value is not realized inbusiness metrics and performance. Because of alack of market and product maturity, traditionalspeech analytics tools tell customers a lot aboutsingle words, but the organization loses contextualunderstanding when a tool focuses on single words.The conversation has to be at a point of contextual

understanding and assisting the leadership andmanagement team of call centers with a means ofpredicting and assisting them with helpingunderstand how the customer experience impactsfinancial outcomes. Then tailor an engine to find theright context that matters to an organization. Whenthis happens, you can impact sales conversion,churn, call volumes, compliance, QA. AI in thecontact center must move beyond traditional legacysolutions. Beyond sentiment. Users need the abilityto map relationships used in human language topredict outcomes based on those relationships.

Table of Contents1. Artificial Intelligence Trends You Need to

Know

2. Artificial Intelligence’s Impact on the ContactCenter Industry

3. How Artificial Intelligence Can Improve CallCenter Performance

4. How Contact Centers Can Put ArtificialIntelligence to Work for Them

5. Speech Analytics — An Emerging Player in theContact Center Industry

6. The Artificial Intelligence – CustomerExperience Relationship

7. Listening to Every Customer With the Help ofArtificial Intelligence

8. Artificial Intelligence as a Driver of BetterBusiness Decisions

9. Flexibility and Analytics — Your Keys toCustomer Engagement

10. Transforming Your Quality Analysts IntoBusiness Analysts

11. Better KPIs Through Combining People and AI

12. The Noble Blessings of Contact CenterEvolution

13. Why Technology is Nothing WithoutPeople…and Neither Is Your Business

14. Embracing Technological Development

15. Your Future Growth is Technology-Driven andHuman-Centered

16. An Important Reminder: Artificial IntelligenceShould Enhance, Never Supplant

17. Why Artificial Intelligence is Your Gateway toSuccess

1

1.

Artificial IntelligenceTrends You Need toKnow

Today’s consumers want more than good productsand good prices. They want good service and aneffortless experience as well. In fact, they demand it.They want it fast, they want it good and they want ittheir way. Part of the challenge for contact centersis that “their way” varies from customer to customer.While one person’s way is via phone, another’s isvia an app and yet another’s is via social media. Tobe successful in today’s environment, you’ve got tobe everywhere all the time.

What happens if your company isn’t everywhere allthe time? Your customers will leave you. And, as ifthat’s not bad enough, you can expect up to 95% ofthem to tell their friends, family and coworkersabout their bad experience. Believe it or not, that isexactly what is happening in American commercetoday. How much of an impact can this have? $41

2

billion. That’s how much sub-par service is said tobe costing U.S. businesses every year. 1

But, before you rush out to get expand your contactcenter’s voice capabilities to every channel such associal media or email stop! You’ve got to approachthis strategically and with an awareness of thetechnology you can use—and how and why to use it.In this e-book, we’ll examine several topics aroundhow you can put artificial intelligence to use in yourcontact center.

Four key trends guiding our discussion involve:

The rise of the cloud contact center. Machine learning and data mining. Cross-functional responsibility

for outstanding customer experiences.

The Rise of the Cloud Contact Center

The cloud is no longer only an option if you want toeffectively compete on customer experience. It is atthe heart and soul of the new standard. Only cloud-based systems give you the bandwidth andflexibility you need to power systems that can trulywow your customers.

A move to the cloud may well necessitate aretooling of your existing infrastructure, but whencompared to the potential for 95% of dissatisfiedcustomers speaking poorly of you to others, thedecision should be clear.

3

Through the cloud, you can deliver the omnichannelexperience your customers want, need and deserve.It is also through the cloud that you can obtainmassive amounts of high-quality data that can directyour operations. In fact, it is this data that canchange your customer experience strategy fromreactive to proactive.

Machine Learning and Data Mining

As a form of cognitive technology, machinelearning takes artificial intelligence to a whole newlevel. Machine learning has always been asignificant component of AI and is a subfield;allows a computer to mine large quantities of dataand find the golden nuggets you need to increasecall center productivity and effectiveness, decreasecustomer effort, reduce channel switching andimprove processes. It can classify, predict andpersonalize the customer experience and make itmore intuitive. It can also help directly withidentifying issues and the most effective solutions.AI’s role in finding real solutions to customerproblems or identifying sales opportunities is moresubstantial than simply personalizing the customerexperience.

Cross-functional Responsibility for OutstandingCustomer Experiences

Call centers no longer shoulder 100% of the burdenfor making customers happy. This responsibility is ashared one across all departments in an organization.

4

But, in order for other groups to participate in this,they need access to information. Once again, thecloud will be instrumental here as it is the gatewayvia which contact center software can be connectedto CRM and ERP tools, for example. Some softwarethat provides access for all departments is evenavailable today. Marketing, technology and morecan share knowledge and make business decisionsusing the same information.

Armed with contact center data, management andexecutive leadership can make decisions aboutproduct development and marketing messaging thatproactively address the customer experience.Suddenly, the customer experience is an ongoingcycle that starts even before a product is developedand continues all the way through to its use by thecustomer.

5

2.ArtificialIntelligence’s Impacton the ContactCenter Industry

There is no doubt that artificial intelligence has hadand will continue to have a dramatic impact on thecontact center industry. Automation of routine tasks,enhanced service delivery, targeted marketing andimproved rep responses are just some of the benefitsour industry has enjoyed, thanks to artificialintelligence.

Automation of Routine Tasks

By automating routine tasks, reps are freed up forother, more valuable work. A great example of thiscan be seen in the banking industry. Very fewpeople today actually go to a bank to deposit acheck or transfer money between accounts. Thesetypes of things can be much more easily done via amobile banking app. Bank employees can then be

6

more available for customers who have greaterneeds that may result in more business for the bank.The customer enjoys the expediency andconvenience of not having to fit in a trip to the bankand a wait in line.

When those in-person interactions do take place, AIallows them to be supported by much smarterinformation that helps reps better satisfy—and evenexceed—customer expectations.

Enhanced Joined-up Service Delivery

Providing a fully functional omnichannel platformvia which customers are served is no easy task. But,thanks to artificial intelligence, it is possible. Thisapproach can transform the customer experienceand our industry by allowing customers the optionto interact in the ways that they want to, not theways we tell them they have to.

It can be easy to get caught up in all that technologyallows us to do and put us on the treadmill to makecustomers keep up with us. But, it is important thatwe use technology to serve customers, not forcecustomers to use technology to serve us.

For this reason, it is essential that we offerinteraction options for people who may be lesscomfortable with “talking” to us via social media oreven a chat feature. Older consumers especiallymay wish or even need to have the ability to talkwith a live person on the phone. Even a younger

7

person with a disability may have this type of need.Once again, it is the customer that dictates theomnichannel experience.

Targeted Marketing

Knowing our audiences allows us to effectivelytarget not only our presales marketing but also ourcontact center marketing and delivery. AI is yourgateway to better understanding your customers —and then taking what you know and applying it tomeet their needs. Big data information likebrowsing history or location can let you personalizemore of your customers’ experiences than everbefore.

Improved Real-time Responses

The concept of skills-based routing is a good one,and today’s AI can amplify your ability to put thisinto practice. Many times, the customer will noteven know that a computer is behind the scenesarchitecting their interaction with you—and that’sokay. In fact, it is almost preferable. The moreseamlessly you weave technology into your contactcenter delivery, the better. Your customers shouldfocus on their experience, not the technology. It’sup to you to use technology to make that experiencea good one. The goal should be to provide thegreatest benefit to the customer with the leastamount of effort. Different channels for customercontact should facilitate customer interactions rather

8

than asking the customer to adapt to reduce thecontact center agent’s efforts.

3.How ArtificialIntelligence CanImprove Call CenterPerformance

Some of the biggest drivers of artificial intelligencethat call centers can benefit from are speechanalysis, predictive analytics and optimization ofcurrent contact center approach.

Speech Analysis and Natural LanguageProcessing

Today’s software has the ability to listen to callsand deduce what a customer needs. Voice analyticssystems are smart enough to differentiate betweenthe customer and the rep so as to provide accurateinformation to the company and integrate into yourCRM tool.

9

Most of us have probably been in phone trees wherewe had to say key words to get to the right person orplace. Now, however, solutions enable us to saywhole sentences, not just specific words like “rep”or “billing”. We can say “I need help understandingmy bill” and the system will understand us. This canbe a huge assist to traditional skill-based routingtechniques.

Predictive Analytics

In addition to voice data, predictive analytics takesother information known about a customer toidentify those people who are most receptive toconversion. That conversion can take many forms—a sale, payment of a debt or some other desiredoutcome. Even better, identifying these patterns thatmay take a human being weeks if not months tocomplete can be done in a matter of hours by theright software.

Predictive analytics is an important aspect of AI inthe contact center space, but it is less aboutidentifying patterns in the interactions of a singlecustomer and more about identifying recurringpatterns across customers. If calls regarding aspecific product are resolved in a similar way 90%of the time, it is not necessary to know a lot aboutthe current customer, if they have the same interestor reason for calling, you already know with 90%confidence what the resolution is going to be. Thisis the whole point of “predictive” it allows agents to

10

predict a customer’s needs without knowing a lotabout a given customer.

Optimization of Current Contact Center Approach

Between voice analysis and predictive modeling,contact centers can learn much about theircustomers in order to serve them better goingforward. However, the benefits do not stop there; itcan be used for customer effort analysis (CEB), riskmanagement and post call automation. Becauseinformation is also gathered about the reps, thesetechnologies offer tremendous potential when itcomes to rep feedback and training. Somecompanies are even using it to help recruit top repsbased upon specific needs.

11

4.How Contact CentersCan Put ArtificialIntelligence to Workfor Them

When learning about the various capabilities ofartificial intelligence, it can be hard to know whereto start in putting it to use. Perhaps one of the mostimportant points to remember when deciding howto implement an AI solution for your contact centeris that it can do many things but cannot doeverything.

Computers are smart and can even be programmedto be emotionally intelligent but should not belooked upon as complete replacements for yourhuman reps. Using technology for the right tasks—like identifying customer preferences, alerting whenthere is a need for a supervisor on a call or evenautomating outreach—is essential in order to getmaximum value out of AI in a contact center.

12

A few specific areas where artificial intelligencecan benefit contact centers are for responseknowledge recommendations, running AIalgorithms against the right data sets, and improvingrep performance and customer satisfaction withpredictive analytics.

Response Knowledge Recommendations

By leveraging voice analysis and natural languageprocessing, artificial intelligence can guide reps asto what they should do, say or ask when interactingwith a customer. For example, if a caller to aninsurance company mentions that she recently gotdivorced, the system could display prompts to therep that relate to the keyword “divorce.” The repmay be prompted to ask the customer if she wouldlike to update her address, name or evenbeneficiaries. This can be especially helpful to newreps, allowing for reduced training time and a betterlearning curve, allowing new reps to reach theperformance of your tenured team much sooner.

Machine learning leading to AI

Dark data (voice) converted to normalized data andthen processed with a Machine Learning engine toproduce BI.

When looking to make the most out of artificialintelligence, the “garbage in-garbage out”philosophy should be considered. What we mean isthat choosing what data sets to use for building

13

algorithms matters—a lot. One of the most commonproblems companies make is choosing massquantities of data as the primary selection criteriafor building algorithms. Instead, the quality of thedata should take precedence here. It is equallyimportant for whole solutions to be properly brokendown into individual tasks requiring processingbefore the rational algorithm is designed. AIsystems can do many things, but those things areultimately done based upon the instructions given tothem, so starting with good set of initial instructionsand building from there is a must.

Improving Rep Performance and CustomerSatisfaction With Predictive Analytics

Preventing problems before they occur is always anoptimal route to good outcomes. When looking atthe performance of contact center reps and thesatisfaction of customers, predictive analytics canbe a company’s best friend. Artificial intelligenceexcels with this type of non-customer-facinganalytical need. Reviewing everything fromkeywords in different channels to voice pitch fromcalls, the ability of reps to effectively addresscustomer needs can be accurately assessed. Trainingcan be developed based upon this information thatproactively arms reps with the right things to do orsay so they are not just solving customer problemsbut delighting them in the process.

When you employ AI to understand customerinteractions, you enable your teams to provide top-

14

notch service and manage any issues that arise withsurgical precision. This not only helps thecustomers, but also helps in coaching and trainingthe reps to exhibit the behaviors you expect andprovide customized services based on customerpreference. Most importantly role of AI isidentifying the most common customer issues andthe most effective solutions to those issues. Or, inthe case of sales, identifying what most customersare looking for and what the most effective salespitches are.

The Need for a Quality Monitoring System

If asked right now how well your contact center isperforming currently, what would you say? If youranswer would be any form of “I don’t know,”you’re not alone. If your answer is a resounding“fabulous,” how can you prove that assertion? Ifyou can’t, you’re also not alone.

Relying on hunches or anecdotal data to gauge theeffectiveness of your contact center opens yourbusiness up for potentially serious problems downthe road. These problems can quickly extend farbeyond an individual customer’s experience andmove into the realm of reputation management. Allthat needs to happen is for one disgruntled customerto get onto social media to vent and, well, thereputation slander snowball begins.

Did you know that it takes 12 positive customerexperiences to negate one negative customer

15

experience? That is why it is so important that everyinteraction with a customer be of the highest quality!

Many things work together to make up the overallquality of a customer experience. At a minimum,these include:

The politeness and professionalism of therep.

The resolution achieved. The time required to reach a resolution

(including time waiting for a call to beanswered as well as time required tocomplete a call).

The simplicity of getting help.

When even just one of these things does not meet acustomer’s expectations, you could be in trouble.This trouble does not always mean your staff didsomething wrong, either. It may mean thatsomething in your process or system isn’t workingwell. You won’t know where the problem lies oreven know that you have a problem unless you havean effective quality monitoring system in place.

Understanding clearly the experiences that yourcustomers have when engaging with your contactcenter is essential. These experiences can directlyplay into the marketing and branding efforts of yourbusiness.

Your contact center does more than simply funnelcalls and process tasks. Consider the following

16

results that a high performing contact center canachieve or contribute to:

Improved customer satisfaction andretention.

Cost savings from more efficient callhandling.

Increased number of cross-sells and upsells. Reduced employee turnover due to

improved training and overall jobsatisfaction.

Better adherence to company policies.

Quality assurance programs for contact centersdirectly impact your ultimate bottom line. Equipyour staff with the tools and support they need andthey will be happier and more productive. And,having happier employees generally translates intohaving happier customers.

Developing a Customer Driven Monitoring System

Before going any further with a contact centerquality monitoring system, a discussion about theimportance of your customers and their experiencesis warranted. Certainly you want to reap somebusiness benefits from your monitoring program,but those benefits, quite frankly, should besecondary to improving the overall customerexperience. In fact, it could be said that anybusiness benefits should really and simply benatural by-products of an improved customerexperience.

17

When not at work, you yourself are a consumer.Think about your experiences as such. Whatmatters to you? Chances are, what matters to you asa customer matters to your customers as well. Yourquality monitoring system could be thought of like asolar system with your customers in the middle asthe sun and everything your business does revolvingaround them. The number one reason to monitorand improve your contact center experiences is forthe benefit of your customers.

Consider these statistics:

70% of consumer buying experiences areinfluenced by the customers’ perceptions ofhow they are being treated by companiesaccording to McKinsey.

In 2010, 36% of companies consciouslycompeted based upon customer experience.The 2015 Gartner Marketing SpendingSurvey indicated that by 2016, that numberwill rise to 89%.

Forrester Research reported that a pricebased value proposition is today far lessimpactful today than a positive customerexperience.

If all of the above is not enough, Accenture explainsthat a poor customer experience is far and away the#1 reason that makes customers change companiestoday. The ease of shopping online makes finding aparticular product or price from another company

18

easier than ever. The only real way for companies todifferentiate themselves, then, becomes customerexperience.

So, how do you make sure that your qualitymonitoring system is centered around yourcustomers? You must make sure that your questionson evaluation forms focus on the areas that mostdirectly impact the customer experience. At aminimum, you must track the following:

Hold time by channel. Abandonment rates by channel. Number of interactions to resolution. Speed to resolution. Total volume of interactions by channel.

In addition to these very objective measures,questions that point to any potential extraordinarynature of an interaction should be included. Thegoal of every customer interaction should always beto create a “WOW” experience. Your monitoringprogram should determine whether or not this isbeing done.

19

5.Speech Analytics –An Emerging Playerin the Contact CenterIndustry

While an omnichannel approach to customer serviceis important, the reality is that the vast majority ofAmerican consumers—90% of them in fact—stillprefer the telephone as their mode of resolvingsupport issues.1 This makes speech analyticsperhaps one of the most essential forms of artificialintelligence that any contact center could employ.

Armed with the information available from speechanalytics, you can improve the customer experience,monitor contact center quality, reduce operationalcosts and gain key business insights.

20

Continually Improve Your Customer ExperienceStrategy

With the rapid advance of modern technology, itseems that nothing in our world stays the same forlong. That certainly is the case for your customerexperience strategy, which should be regularlyevolving to best meet the needs of your customers.By using today’s data, you can better informtomorrow’s business decisions and set yourself upfor success. More importantly, you can set yourcustomers up for delighted experiences becausetheir needs were put first.

Improve the Customer Experience

Speech analytics can tell you everything from thereason a person called to what their emotional statewas at any stage throughout the call. Byunderstanding a person’s needs and wants, you arebetter poised to meet customer expectations—and tohave a happy customer. AI can not only tell youwhat the customer’s emotional state was during thecourse of a call, (which is found to be almostuseless,) AI can tell you how likely they are to buy,how much effort they are experiencing, and avariety of other specific and important metrics

Monitor Contact Center Quality

Speech analytics programs do not only trackcustomer voices but rep voices as well. This holisticapproach to data gives you a window into thequality of your reps’ work. From customer

21

satisfaction to regulatory compliance, you can stayon top of how well your contact center is meetingits objectives.

Reduce Operational Costs

By learning more about what customers need, youcan put things into place that may even pre-empt theneed for some calls to your center. This keeps repsfreed up for more essential calls. You can also learnwhat things your reps may need in order to moreeffectively serve those customers who do call in.The more efficiently service can be deployed, thehappier your customers will be and the morecustomers your reps can serve.

Gain Key Business Insights

By sharing information from your speech analyticswith other departments in your organization, youcan use the data to make important decisions aboutproduct roadmap and development, marketing andmore. Your reps may also be able to learn how tobetter identify cross-sell and upsell opportunities.

22

6.The ArtificialIntelligence –Customer ExperienceRelationship

Make no bones about it—artificial intelligence isintimately linked to customer experience. To deliverthe best customer experience possible frombeginning to end in today’s world, you shouldemploy some form of artificial intelligence. Yourcompetitors are doing so and therefore your lack ofintegration will quickly become obvious to yourcustomers and impact your results.

You should look to AI to help you deliverinformation to the customer, obtain speechrecognition data, pinpoint needs with exceptionalaccuracy and continually improve your customerexperience strategy.

23

Deliver Information to the Customer

Much of the time when a customer reaches out forhelp, what they really need is information. Moreand more, that information can easily, quickly andaffordably be delivered via artificial intelligence.This makes customers happy because they get whatthey want faster and with less hassle. It makes youhappy because you don’t have to allocate yourvaluable rep time to issues easily addressed bytechnology.

Speech Recognition Data

The information that can be obtained by speechrecognition software is nothing short of aninformation gold mine for your business. Scalingspeech recognition software allows for AI andmachine learning to sit on top of the data andbecome insightful. It can empower your reps withgreat information to facilitate better call handling onthe spot. It can give you insights into how yourcustomers view you and your competitors and evenwhat things your competitors may be up to that youmay not have otherwise known. Speech recognitionsoftware lets you test out different call scripts andlet AI help you quickly determine which one ismost effective. Patterns across multiple calls can bepromptly identified, giving management theopportunity to get in front of some issues andproactively give customers the information theyneed without having to wait to get to a live rep

24

while simultaneously reducing the call volume loadon reps.

One of our clients, a leading Wireless Internetservice provider, previously utilized traditional QAmethods, which helped them, understand the reasonof the customers call and what the consumers weresaying. The feedback was interesting, however,when compared with the amount of calls they werereceiving, the sample size was simply toosmall. After a month of utilizing ArtificialIntelligence to help with this issue, they were ableto process all of the calls for two specific skill setsthat provided them the below key business insights:

Major Call Reasons: Customers called because:o They were moving and wanted to cancel

the accountso They had issues connecting to the

services or getting logged ino They cancelled previously but now

wanted to reactivate Negative Customer Sentiments: Customers were

unhappy because:o They were charged on closed accountso They had speed and connection issueso They were calling to check refund status

multiple times Agent Performance: The agents failed to:

o Provide effective rebuttals to retain thecustomers

25

o Probe and effectively troubleshoot thereasons for customers not able toconnect

o Provide complete details on refund thatincrease repeat calling

While the traditional QA would sporadically pickup on these issues, the AI provided proof that theseissues were much more common and happeningquite frequently.

By understanding the issues, the client was givenrecommendations that included:

Steps on website for reactivation (This willreduce the call flow by 16% when customers areeducated vs. calling the contact center)

Notification and automated system to avoidcharges on closed accounts (This will reduce thecall flow by 5%, and also increase customersatisfaction as this was one of the top threenegative sentiment drivers)

Agents were coached to save, but with AI andlistening to 100% calls, it helped in ensuringthat the coaching was implemented on the call,and if customers were retained

Handle time was reduced with automated IVRcustomer verification VS verification on the call

Benefits:

Increased customer satisfaction

26

Increased Agent morale and performance withtargeted coaching

Reduced Cost through:o Reduction in call volumeo Reduced handle time

Pinpoint Needs with Exceptional Accuracy

Not only can speech recognition tell you what acustomer wants or needs, what makes themunhappy (wait times, pricing, etc.) and what it willtake to make them happy, and you can get all of thisinformation faster than you could ever get it withoutAI. Remember: you can’t expect to fill a need youdon’t know about or truly understand.

Script Testing

A/B testing is not just for marketing anymore. Yourcompany might be rolling out a new product andyou may therefore logically expect some increase incalls as customers seek assistance getting their newproducts properly setup. But, just how much help isenough and how much is too much? Perhaps youhave two different call scripts, one of which goesinto far more detail than the other. You might beconcerned about using the longer one because youfeel people would be put off by spending that muchtime on the phone with a rep, but if you use theshorter script, are you leaving out too manyimportant steps that in the end compromise thecustomer experience? You don’t have to choose

27

because AI lets you test both scripts and see whichis most effective.

Event Response

Sometimes calls into your contact center aregenerated by events external to your company. Anew governmental regulation and a competitor’sannouncement are two examples of things that mayraise questions from customers. For example,imagine that you are in the insurance industry andnews just broke out that a competitor is underinvestigation for allegedly raising premiumswithout telling customers. Even though yourcompany is not named in any way, it is logical foryour customers to wonder about their ownpremiums. Suddenly, you notice an increase inrequests to review accounts as people want to checkon their premiums with you. AI doesn’t just have tolet you know that you’re getting more calls but itcan actually tell you why and give you the chance tocreate new scripts to most effectively address thesecalls. This can result in less time needed to handlethese calls and, ideally, more satisfied customers.

Product Enhancements

It’s important to remember that a customer’sexperience extends far beyond their contact centerinteractions with you. It also includes theirexperiences actually using your products or services.Let’s go back to the first scenario above in whichyou launched a new product and experienced a

28

jump in calls from customers needing help to getthings properly set up. Even if you have the bestcall script or automated assistance in whichcustomers can get the help they need in minimaltime that may not be getting the most out of your AI.AI can raise the flag for a broader need. If a highvolume of calls indicate that customers arecontinually unable to understand how to completesome very basic tasks because your product UI isnot intuitive, that information should be funneled tothe product teams. They can then use your data tohelp improve their products and eventually reduceor eliminate the need for calls on the subjectaltogether.

29

7.Listening to EveryCustomer With theHelp of ArtificialIntelligence

Customers are Everything

Your customers are the lifeline of your business andmust be at the center of everything you do. Whenyou listen to your customers, you are able tounderstand their needs, customize products andservices accordingly and create a great customerexperience. A great customer experience not onlytranslates into a successful brand and business, butalso helps to increase market share. The best wayto make a customer feel important is to ensure theyknow that you are listening to them. If you arelistening to only few of your customers, you maynot be able to keep them all satisfied, which willeventually lead to high customer churn.

Listen to the Voice of Customer (VoC)

30

Since Customers are everything, businesses mustensure that they are listening to the Voice ofCustomers (VoC). This not only helps inidentifying issues, but also helps in findingadditional opportunities for your product offerings.When you ask for feedback, customers feel valued,but just filling out surveys on a fast food receipt toget a free sandwich or a discount is not enough.Customers truly feel valued when companies act ontheir feedback and then communicate back to them.When you listen to your customers and respondaccordingly, it shows that you really care, helpskeep business, generates repeat business/referralsand also increase brand loyalty.

Analyze the Voice of Customer (VoC)

There are multiple opinions in the industry aboutwhat the best tool in the industry is for gaugingVoC, and while the tools may differ, the “voice”can be categorized into two forms based on the datathat is captured, i.e. structured and unstructuredVoC.

Structured VoC: This is the data that is capturedthrough surveys, polls, questionnaires,cancellations and website experience; and iseasily processed and measured withconventional systems.

Unstructured VoC: In today’s times, customersinitiate contact via multiple channels such as

31

phone, email, chat and social media. These callrecordings, email/chat transcripts and socialmedia mentions are a rich source of data forwhat customers are saying but need help fromanalysts to make feedback actionable.

61% of customers call a contact center to talk tosomeone via phone either because they prefer totalk to a real person, or because they were nothelped through other channels. While theseinteractions are rich in data, only 2% of them areactually listened to, and only some of the feedbackis retained and then acted upon. The AI SpeechAnalytics engine transforms this unstructured datainto a structured indexed data that ensures you arelistening to all your customer interactions,analyzing them and have a broader view andinsights into the “real” VOC versus the sampledversion that is obtained through surveys, polls andtraditional quality assurance methods.

As a summary, when you utilize ArtificialIntelligence to listen to your customers, analyzetheir feedback and act upon it, you have all theingredients that are needed to help you makedecisions that will have the greatest positive impacton your business and customers.

32

8.Artificial Intelligenceas a Driver of BetterBusiness Decisions

The manner of business decisions you make isdependent on your experience, your goals, and yourunderstanding of the current needs of your business.To make these decisions, you need access to thebest data possible to correctly inform your decisions.If you are given incomplete or incorrect information,the decisions that will be made may not help youreach your goals and put you at a disadvantageversus your industry competition.

When booking a hotel reservation in Austin, TX,one team looked for the best prices and found someunbelievable deals. However, before booking, theycalled these hotels and checked their amenities suchas WiFi, parking and breakfast, all of which areimportant for a business traveler, and all of whichwere a-la-carte and not included in the reservation

33

prices. What the team thought were great dealswere now no longer even a viable option.

Are you reviewing only the data you receive tomake your decisions, or also determining thesources, the legitimacy and if the right questionswere asked to gather that data?

Let’s take a moment to look at the impact artificialintelligence is having on the business world outsideof the contact center arena. AI is essentially leadingthe charge to make scientists out of today’s leaders,facilitating smarter decisions and outcomes. It alsoallows businesses and consumers to make smarterdecisions thanks to the ability to make predictionsmore quickly.

Turning Leaders into Scientists

It’s a fact that executives are people like the rest ofus. Like all of us, they approach any major businessdecision with their own sets of biases and emotions.Smart executives, however, know that theirdecisions must be supported by facts when makingfar-reaching decisions. Enter AI. More and moreleaders are looking to the big data that only AI canprovide to help them better direct theirorganizations. In fact, in a recent PwC survey,executives indicated that they would rely onalgorithmic data for at least 41% of the weighting intheir next strategic decision.1

34

The accessibility of data allows leaders to testhypotheses, much as a scientist would. The datareturned can either confirm or deny the originalthoughts but either way provides invaluableguidance forward.

Predicting to Win

Predictive analytics have thrust tremendous powerinto the hands of many industries. In health care, forexample, it is used to design treatment plans thatbetter address patient needs while saving money atthe same time. In the financial services industry,investors can look to data for bid strategy guidance.

Knowing when to let the machine do what it isprogrammed to and when to let the human do whatthey have been trained to do is an important elementof making AI work for business. Doctors andfinancial planners use AI as tools to help designeffective treatment plans and investment strategies.The whole process is interactive and not a secondguessing of what the software comes up with.

35

9.Flexibility andAnalytics – Your Keysto CustomerEngagement

Getting customers to engage with you is vital. It’sanother way for you to stay connected with themand at the top of their mind for not only today’spurchasing decisions but tomorrow’s as well.Artificial intelligence can play a large role inengaging customers but should not be relied upon todo this on its own. The combination of people andAI working seamlessly together, is what will reallydifferentiate you from your competitors and achievethe engagement you need to be successful.

Deliver the Human Touch

AI should never be thought of as a completereplacement for humans. Instead it is a way ofallowing humans to do what really only they can dobest, in part by allowing technology to do what it

36

can do best. One of the things that technology cando for you is to identify if you are giving yourcustomers the right level of human interaction. Arehuman reps being used at the right touchpoints?Does your system know when to bring in a human?Does this happen too soon? Too late? Let youranalytics guide you to the right balance to maintainsatisfaction and engagement.

Engage Across Multiple Channels

Every one of your customers is a unique individualand they are not all going to want to engage withyou in the same way. Similarly, one single customermay want to engage with you in multiple ways evenfor a single issue. AI is the way that you canaccommodate these demands.

While phone is still a highly preferred method ofinteracting with contact centers, it is predicted thatby 2018, 49% of consumers will interact withcontact centers via mobile apps.1 That said, it is alsonot just enough to offer multiple ways for customersto engage; you must also allow customers the abilityto switch between communication methodsseamlessly. Don’t make them go through anautomated system where they input all of theirinformation only to have to repeat it again whenthey talk to a live rep.

Connect to the Bigger Picture

37

As discussed earlier in this book, AI can collect datathat may be useful to many parts of yourorganization, product development being just oneexample. However, this takes human eyes as well astechnology to know what data should be sharedwith what departments. Bringing people and AItogether in this way leverages your contact centerdata far beyond the contact center.

38

10.Transforming YourQuality Analysts IntoBusiness AnalystsWhile most organizations have some manner ofQuality Assurance methodology in place to ensurethat their customers’ key compliance and qualitycriteria are met, all of us will agree that QualityAnalysts are bound by check boxes in a QA form orsaying something was done or not done throughoutthe interaction. With the help of artificialintelligence, you can transform them into somethingmuch more powerful. Let us show you how!

One of the primary benefits AI brings to yourcontact center is in the form of data, and what youdo with that data is what really matters in the end.You need to be prepared to take the information anduse it to empower your reps. How do you do that?

Start at the Top

Before you identify what items you want to track ormeasure with AI, you should identify the businessgoals you believe your contact center can support.For example, your organization might have a statedgoal to improve the customer experience. This is

39

clearly something to which your contact center cancontribute in the areas of QA, 1st contact resolutionand many others.

Look at the Whole

Your contact center’s AI may often provide insightsthat other parts of your organization may find useful.Understanding the larger business goals can helpyou determine what nuggets are worth sharing withwhat departments in order to facilitate meetinglarger company goals. Training your reps to lookfor this information is also important and helpsthem to grow and understand the bigger pictureabout how their role really matters to the company.

Combine Evaluation Methods for SpeechAnalytics

When using AI, you will certainly want to rely onauto-scoring of speech analytics, but you shouldalso use manual reviews for a small portion of callsto ensure your AI remains properly calibrated withyour expectations. In addition, it is important toproperly determine what phrases or elements of thecall should be mined for speech analytics. Thingsthat tend to be most effectively measured in thisway include greetings and closings, the collection ofinformation to verify identity and the generalempathy level of a rep.

A super analyst tactic with speech analytics is tocombine reviewing this data with a look at overall

40

customer satisfaction data. This might be fromthings like the average time to answer a call or evencustomer surveys. The bottom line is that you mustremember speech analytics should always connectback to the overall customer experience. It is highlypossible that a rep properly uses all key phrases butthat a customer is still left unsatisfied. It is equallypossible that the root of the customer’s lack ofsatisfaction is unable to be addressed by the repthrough no fault of their own, but is somethinganother member of your team could resolve.

Get Reps Involved in the Evaluation Process

From the very beginning, your reps should knowthat call monitoring is simply part and parcel ofyour operation’s processes – not because you don’ttrust them, but because your customers’ experiencesmatter that much. Let your reps be part of theevaluation process from identifying things to trackas well as using calls as real-life learning examples.This may be in the form of formal training sessionsor weekly mini-reviews.

Speaking of weekly reviews, it is best to have aweekly cadence for call reviews rather than monthlyso as to make this a very normal part of your work.Reps should also be able to ask to review specificcalls with managers or even with peers in order toget feedback and learn. When they do this, the callsused should not be included in rep scoring becauseyou do not want to hamper a rep’s proactive requestfor learning.

41

Calibration Across the Organization

When measuring a call, against what do youmeasure it? This question must be answered fromthe outset, and the answer should be agreed on byreps, contact center leads and upper managementalike. This cross-organization alignment is anotherway of ensuring that your contact center is workingto meet larger company goals.

42

11.Better KPIs ThroughCombining Peopleand AIIdentifying the right key performance indicators isan important part of making sure you accuratelymeasure the effectiveness of your contact center andyour AI software, but that’s not the only part ofgetting the most out of your AI capabilities. Thereare things that technology can do better than peoplebut there remain some things still require a humantouch. Putting the best of these two resourcestogether is what ultimately can deliver you the bestKPIs possible.

Pulling the right data from calls is step one. Steptwo is putting that data into an understandableformat. Step three is then having the information bereviewed by contact center management. It is herethat trends can be seen and decisions made. Thesedecisions are then able to be used to further directfuture AI system activity and the cycle will becomeone of ongoing process improvement.

When done correctly, you will be able to:

43

Pin down the exact reason for a customercall.

Offer enhanced solutions based upon whathas worked for other customers with thesame or similar problems.

Provide tips at critical points in a call,reducing the length of time needed toresolve an issue.

Identify customer stumbling blocks tostreamline call resolution.

You will also have the ability to know if a call is arepeat attempt to resolve an issue or when issues areable to be effectively addressed in only one call.

At the end of the day, reducing time spent per callcan save money, improve productivity and boostcustomer satisfaction. But, because working withhuman beings is never black-and-white, the targetedinvolvement of personal review of AI data is arecommended step in identifying the KPIs youshould track to achieve these goals.

Customer Satisfaction

Call analytics can help you get a window into howwell your contact center performs from the eyes ofthe people who matter most—your customers. Youshould track the following here:

Customer Effort —a score that represents anoverall sentiment and thought on the part ofthe customer.

44

Customer satisfaction index—a compilationindex that indicates general satisfactionlevels.

Customer efficiency score—a scoreindicating how efficient customers viewyour contact center.

Customer retention rate is another factor you canmeasure that may indicate overall satisfaction.Greater satisfaction will lead to less churn andstrengthening of your brand.

45

12.The Noble Blessingsof Contact CenterEvolution

One way to understand the evolution of the contactcenter is to take a quick look at some of theautomation technology that has graced this industryover the years.

Two of the most influential technologies in theworld of contact centers have been automatic calldistribution (ACD) and interactive voice response(IVR) systems. However, these once industry-leading technologies are fast going by the waysideas new players take center stage. More and more,these systems are associated with an increase incustomer frustration, not the decrease that a contactcenter wants and needs to achieve. What isreplacing ACD and IVR systems?

Virtual reps have proven successful and emailinterestingly continues to be popular and has grownin its use and effectiveness. Web chat with an

46

intelligent interface between customer and agentinteractions is found to beneficial in improvingcustomer experience. Web chat with AI intelligentlyrepresents a specific set of information capturedduring a chat session. Sensitive data such as creditcard details, personal phone numbers, certifyinginformation, validation data, in-chat survey data andmuch more, can now be handled “behind thescenes” by the AI while the agent is effectivelyengaging the customer. The result is highercustomer satisfaction scores, lower handle times,completely secure transactions and increased sales.Skills-based routing, callback features andestimated wait times have all emerged as useful intoday’s contact centers. Of course, the ability forreps to receive accurate information about theircustomers is of special benefit to them.

Humans or Technology?

Seeing this recap of some of the technology that hasfueled the modern contact center, one may wonderif humans are even needed anymore. The answer is:absolutely. At least today, there is no technologythat is able to effectively replace a human. Whattoday’s technology does is allow humans to dodifferent things by freeing them up from other tasks.The most successful contact centers will be the onesthat know when to use technology and when to usehumans. For example, transactional-based needsmay be more easily fulfilled by a machine whereas

47

detailed procedural issues may require a humanbeing.

Choosing the Right Technology

When technology is identified as a reasonablealternative, the next step to success is to identify theright technology. It is equally important to avoid the“technology for technology’s sake” pitfall. Anyartificial intelligence or other function that is putinto use must be done so with a specific purpose inmind. It must solve a direct problem and thatproblem should never be first and foremost thecontact center’s costs; rather, the problem shouldalways start with improving the customerexperience.

Choosing the right technology solution can also beabout choosing the right time to use that technology.Some experts indicate that systems like callbacksoftware should be kept and used only when callvolume is extremely high. Using these features on aregular basis can actually end up becoming a crutchand adding to the level of complexity for a customerand a rep.

Reducing Customer Frustration

As mentioned above, a contact center essentiallyexists to eliminate customer frustration. Therefore,it is important that the use of any technologysupports this goal. That may mean knowing when tobring a human into the picture even if a machine has

48

initiated the contact. It may also mean keepingthings simple and reasonable. A menu that requiresa person to select through seven options is toocomplicated. Every customer should also alwayshave the option of reaching a live service rep ifdesired.

The best way to know if your contact center plan isa solid one is to test it out yourself as a customer.Every single decision made should be made fromthe customer experience perspective.

49

13.Why Technology IsNothing withoutPeople…and NeitherIs Your Business

There was time when, for many people, driving acar was a dream. Now technology has progressed tothe point that driverless cars are being tested inmany metropolitan areas. But a driverless car is stillbuilt for a passenger to reach their destination, ormay be to deliver a cargo from point X to point Yfor use by customers. Technology is nothing morethan a means to provide convenience to humans.

As we have said, the whole point of your contactcenter is to help your customers—people. They arecontacting you because they have a problem andthey need it solved. It is your job to figure out howto do that in the best way possible. It is your job tomake people happy. It’s about people. And, at theend of the day, only people can really make otherpeople happy.

50

Yes, the contact center of today and certainlytomorrow must rely on technology. But, that doesnot mean that your contact center does not needyou—or other people. In fact, one could argue thatyour contact center needs people more today thanever before.

The demands on your contact center are greater thanever. From an omnichannel presence to a customerexpectation of virtually instant assistance, you arepummeled on a regular basis with needs from allangles. Technology can be your best friend inanswering these calls and turning disgruntledcustomers into happy people, but it cannot replaceyou.

Perhaps one of the best ways to think about this isto look at how top recruiters today consciouslyevaluate the “how” as well as the “what” inpotential candidates. There may be two peoplevying for a job and one may have superior skills orknowledge, yet may display a serious lack ofinterpersonal skills. The less qualified person,however, may end up becoming the company’schoice for his or her strong interpersonal skills.Knowledge can be gained. It is far harder to learnhow to behave well at work, especially as peopleget older. This emphasis on soft skills cannot beoveremphasized, especially for the contact centerwhich deals exclusively with people.

51

14.EmbracingTechnologicalDevelopment

We all know that change is inevitable, constant andsomething that most of us resist, but the only wayfor businesses to succeed and stay competitive is tostay updated on new innovations, understand theirbenefits and get on the “change train” before othersdo. You can choose to play the “wait and watch”game, but your competitors and customers willlikely leave you behind. Remember the fates ofBlockbuster, Nokia, Kodak, Motorola and manyothers that lost their edge just because they waitedtoo long to adapt and evolve.

As it should be clear to you by now, contact centerseither need to get in on the AI game or, well, let’sjust say they need to get on board. But, what exactlydoes that look like? Who owns the responsibility forembracing technology in the contact center? Wewould suggest that the responsibility is actually ashared one.

52

The Rep’s Responsibility

Individual contact center reps must remain open-minded and accepting of changes that come theirway. Instead of feeling threatened by technologicaladvances, they should embrace them asopportunities to improve their own performance.

The Organization’s Responsibility

Companies should never think that the technologythey employ today might be “the” thing becausetomorrow will certainly bring about a new “thing.”Choosing which technology to implement when -and why - is where the art of advancing contactcenter capabilities comes into play. Leadershipshould always know that change is an ongoingnecessity to remain competitive in today’smarketplace.

The Academic Responsibility

Whether in undergraduate environments orprofessional development settings, the continuedfocus on training and evolving to meet new needsand capabilities should never be lost. This canactually be an integral part of instilling a mindset ofcontinual learning and improvement intoindividuals and organizations alike.

It is essential when adopting new technologies thatbuy-in at both the rep and the company level occurs.Reps need to know that the company supports themand that an implementation will not be done

53

halfway. They also need to be encouraged toembrace their new and growing roles,understanding that technology is not there to replacethem but to allow them to expand what they can do.

For some companies, this may represent somethingof a culture change which does not happenovernight, which goes back to the importance ofsupport from all levels of the organization from thetop down.

54

15.Your Future GrowthIs Technology-Drivenand Human-Centered

For decades now, doom-and-gloom reporters havepredicted mass job loss at the hands oftechnological advances. In fact, some have gone sofar as to suggest that technology could actuallyreplace humans. But, the time has finally come inour current phase of industrial revolution where weare maturing to the point of understanding theunique place for both technology and humans. Infact, done right, there can be an otherwiseunattainable synergy when both are used together.That, by definition, precludes the mass job loss andespecially the full-scale replacement of humanswith technology.

55

Perhaps one of the best examples of this shift can beseen in the technology industry itself. Companieslike Google and Facebook have for a long time hidbehind the guise of technology as the force behindwhat content is served up. By pointing to theiralgorithms, they could maintain a neutrality andavoid responsibility for anything that may in anyway not met the mark.

Now, however, these companies are looking tointerject humans into the content mix. Editors willbe involved along with algorithms to makedecisions about what information and content isdisplayed when, where and why. The limits of thealgorithm, it seems, have been acknowledged loudand clear.

There is no doubt that the growth of technologicaladvances will drive the growth of business solutions,customer experiences and revenue streams; but, atthe center of all these areas, it will be humans thatwill dictate what will stay and evolve, and what willget scrapped. New innovations are driven byhuman needs. Technology should be looked uponas an essential and valuable tool but one that is usedto serve human needs, and one that is never meantto replace the human element.

56

16.An ImportantReminder: ArtificialIntelligence ShouldEnhance, NeverSupplant

In addition to the limitations of technology as areason for the continued involvement of humans incontact center work, there are also financial realitiesthat make this essential. Clearly we have seentechnologies promised to be the next best thing failto be that (IVRs, for example). Those technologiesthat are actually capable of making substantialpositive changes in the customer experience can becostly, and these costs are not just borne up front.The ongoing expenditures and infrastructure

57

changes that may be required are simply out ofreach for some companies.

When you stop to consider that one of the benefitsof automation in the contact center is an ultimatereduction in costs to the business, it quicklybecomes evident that an overly expensive solutionsimply won’t make the cut. The completelyautomated contact center, it seems, is nothing but afigment of someone’s imagination.

58

17.Why ArtificialIntelligence Is YourGateway to Success

All in all, it’s all but impossible to ignore the realityof artificial intelligence as your path to creating asuccessful contact center, and why would you wantto ignore this reality anyway? It’s a bright shininglight for the future of our industry because itleverages not just data but also the machine learningthat allows systems to deliver actionable insights.These insights can drive business decisions,empower reps and ultimately delight customers.

Customer-focusedFirst and foremost, everything you do inyour contact center should be for yourcustomer. This is true whether it is delivered

59

by technology or by a person. When itcomes to AI, before you go down a path, askyourself if that path will improve thecustomer experience. Will it give thecustomer what they need and will it give itto them their way? If yes, you’re on the righttrack.

OrganizedOne thing that machines can always dofaster and better than humans is sort throughand organize data. This gives your contactcenter the ability to help customers havetheir problems solved faster, often withoutthe need for human intervention. Your repsare therefore allowed to grow and work onthings that they couldn’t before—and solvethose customer problems that a machinesimply cannot.

IntegratedKnowing when to use technology as yourvirtual assistant and when to let the humanbe the expert is a critical component ofmaking AI work for you.

AdaptableThe best AI tools will not only be able toabsorb and spit out data but actually learnand adapt to that learning. It’s like thedifference between Google’s once-employedreliance on literal keywords to return searchresults to its now ability to understand thepoint behind a long-tail keyword phrase.

60

The goal is always to give the customerwhat they want with less hassle and in lesstime.

AI in the contact center:

Puts customers’ problems first by goingbeyond basic facts to understand theemotion and need behind them.

Allows your reps to grow and work onthings they couldn’t before.

Adapts and learns so as to fuel your abilityto give customers what they want with lesshassle and in less time.

So, are you ready to put artificial intelligence towork for your contact center—and for yourcustomers?

61

GlossarySources:

TCN - https://www.tcnp3.com/home/ 1to1 Media - http://www.1to1media.com/ Deloitte - https://www.deloitte.com/ Betanews - http://betanews.com/ CallCenterHelper-

https://www.callcentrehelper.com/ Rank Miner - http://rankminer.com/ Xdroid - http://www.xdroid.hu/en/home Forbes - http://www.forbes.com/ PTP - https://www.ptpinc.com/ NTT Technical Review - https://www.ntt-

review.jp/ Clarabridge - http://www.clarabridge.com/ TechTarget - http://www.techtarget.com/