Home > Columns > CRM Columns

CCW Nashville 2020 - Solution Providers and End Users Harmonize in Music City

CCW Winter, the first event on the CX/contact center calendar, took place in Nashville, TN from January 28-31, 2020, attended by professionals representing a broad spectrum of organizations of all sizes and business categories. It provided a memorable week for networking with industry professionals and learning from 75+ speakers -- including an inspirational address by Ben& Jerry’s Founder, Jerry Greenfield, who shared his thoughts on the spiritual side of business and a keynote by Kelley Kurtzman, VP, Global Consumer Sales & Service Center at Verizon--as well as the opportunity to attend a plethora of special events. 

The Expo Hall was open the same days as the main Conference, featuring more than 50 suppliers covering a variety of solution areas. CrmXchange had conversations with selected vendors to discuss their specific applications.

Click on each company name to read and download the eBook:

Caller ID	Drips	LumenVox	Tethr	Trendzact	Unymira	VoiceOps
Caller ID mp3	Drips mp3	LumenVox mp3	Tethr mp3	Trendzact mp3	Unymira mp3	VoiceOps mp3

 

Read the entire column below:  

Caller ID Reputation

Over the past few years, carriers have been crowdsourcing spam call data from third-party call blocking apps. This has had serious consequences for many reputable businesses such as healthcare professionals which are being flagged erroneously. Caller ID Reputation enables companies to receive a notification when their phone numbers are flagged on all major call blocking apps or networks. Caller ID Reputation CEO Joseph Alcaraz provided additional details.

We talked about the many companies you have been involved in running. Can you tell me what motivated you to start Caller ID Reputation?

Over my 20 years in the industry, I’ve seen the alarming growth in robocalling and spam calling, along with the nuisance behavior of telemarketers pushing illegitimate products and running phishing scams. In observing this phenomenon, I’ve been paying attention to not only the major networks but also focusing on the major call blocking applications and the way that they identify calls for businesses. As an entrepreneur in the B-to-B space, I instantly noticed when some of the methods of preventing such calls was starting to go in the wrong direction for legitimate businesses. With many numbers being inaccurately flagged, I knew that what needed to be done was to aggregate the information from all the major data generators and aggregates. In taking this action, I could give the companies that I worked with one place to go to identify what their customers see across every major network and call blocking app.

One personal experience that amplified my focus on the problem took place when my mother was in the hospital two years ago. While she was shuttling between the emergency room and though the ICU, as the individual with power of attorney, I was not receiving important calls that should have rang through to my phone. These calls either went straight to voice mail or were picked up by call blocking apps that I was testing. One of the hospitals she was in was a customer of mine and I was handling all their content management and HIPAA compliant do- not-originate inbound numbers.  I instantly knew that something was amiss when I saw the group being flagged. I was able to pull a full profile of every complaint and all the details tied to it. It became immediately apparent that the complaints being filed were due to someone using their phone numbers in spoof attempts to pose as the hospital to fish out information and generate leads for unscrupulous health insurance providers. The light went off in my mind that there was business value here.  I have a company called Doctor Genius, a growth platform for medical and dental practices, and decided to run a scan on all my do-not-originate numbers and see what was happening. It quickly became evident that many healthcare organizations were having the same problem with phone numbers being incorrectly blocked. This set us in motion to find additional partners that have more than a million subscribers.  Our clients include solar energy companies, satellite providers, public utilities…it’s a problem that is ubiquitous across all business models. When we do scans, we see that approximately 20% of all legitimate businesses have numbers that are inaccurately flagged. All carriers have different standards, and many people only see that they have received a spam call, but when you look further into it, it is often a legitimate call that has been blocked.

How does your program work and how do companies who think it’s a good idea get started?

We give potential clients an opportunity to see what our product can do by running a free 15-day scan where they can monitor their numbers daily. This contrasts to others in our space who provide monthly reports. Having been in the contact center field, we are aware of the value of companies having immediate knowledge that their number has been flagged. Given the increasing difficulty of building trust over the last decade, this is particularly critical. Just about every consumer already has their guard up in taking calls and having them come from a number designated as fraudulent makes it exponentially more difficult.

We get inquiries from all spaces: we focus on companies that have 100 or more reps on the phone because we see the inaccuracies there. But the more we track this and the more aggregators we bring on, we realize that even smaller businesses are being affected. Our top priority is to identify everything going on in the network.  We also have an enhanced caller ID product. Testing is one of my passions, so we meticulously evaluate everything. We test 18 phones on every network to determine any gaps.

We look forward to seeing every cell phone tower and every carrier have the right equipment to have the product be effective 100% of the time and allow incoming calls to have the company’s logo and the reason for calling. Right now, every Fortune 1000 company has what we call a Spam Department.  That may not be how they choose to think of it, but it is how it’s categorized by the networks. Companies might be calling 30-, 60-, 90-day aged leads or making calls on accounts receivable. When these businesses call someone who frequently changes his phone to escape, he is going to flag the caller, especially if he has a call blocking app. These providers want to make their app valuable by enabling people to block anyone. If a handful of people have done this and it becomes a threat, the company will be put on a higher alert and monitored more closely. If complaints go above a certain threshold, it will be flagged. We have a hygiene process that helps clean the numbers.  After that, we have a strategy that we walk all our customers through. As things tighten up, spammers will be spoofing more private cell phone numbers.

Is there any way to prevent spoofing?

We opened perhaps the biggest can of worms ever when Voice over IP (VoIP) was brought into the space. It’s going to come down to the validation process on the front end from the carriers who allow these VoIP companies to operate. I know how it easy it is to set up an out-of-the box 3rd party solution, get IPs and start operating, often without ever even communicating with a carrier. That is the most pressing problem.  There needs to be a true validation for every business, which is what we are doing with the companies with which we work. We give them a Level 2 verification. Not only do we do a Mail PIN verification; we check to ensure the email is at their business and validate every phone number to make sure it rings to the business.

What type of results have companies that work with you experienced and what are their key needs?

Companies that work with us have seen improvements in their Answer Seizure Rates (ASRs), enabling them to not only connect with more people, but increase sales. Before they partnered with us, half of the databases that had opted in with them were never receiving their calls but were being sent to voicemail. This stimulates a cadence that make the calls more of a true nuisance where companies become more aggressive in trying to reach them, such as in the case of aged leads or collection calls. Even if a company has a contact center platform that enables them to implement event-based rules, such as sending a text after voice mail, there are regulations involved in testing and the numbers will still be flagged.

Companies need to have full awareness of what is going on both with their current staff and current reputation on the phone networks. The immediate need is to find out what numbers have been flagged and why, then learn what departments may have been causing these issues After 90 days of working with us, they can get a feel for where they are and build a strategy for moving forward.

Drips is the first conversational texting® company of its kind, founding a new category and leading the way for some of the biggest brands in the world to use automated, humanized conversations at scale. Drips engages in tens of millions of completely humanized conversations with zero client-side human resource operators. The result is improved lead-to-call conversions, a better customer experience and reduced overall lead generation cost without having to hire or fire agents. Jonathan Pogact, VP of Marketing for Drips, elaborated on the benefits. 

How does your solution make it easier for businesses to manage leads?

We exist to bring brands, agencies (their clients), service providers and resellers back into contact with prospects who have raised their hands about a product or service. We live in a world where consumers have more choices than ever and rather than merely interacting with by the old way of ongoing calls from a contact center, we enable them to communicate in a way that the prospect prefers, such as through conversational texting. 

We use context to determine how to establish communication: where did the lead come from, what have they opted in to receive, what are their expectations. Did they connect through social media, such as Facebook or on the client’s website, which we might acknowledge when getting back to them. We want to make sure that the message coming from the business is familiar to the consumer and puts them at ease. We only work with businesses that obtain consent, so the prospect is aware that theirs is a request that is being followed up on. This is done in a variety of ways. We could be the first touchpoint after a person has submitted an online request for an insurance quote. Sometimes we act to support or augment contact center operations in cases where they have unsuccessfully tried to reach the individual on several occasions. This is important because sometimes prospects are either too busy to pick up a call or are unavailable because they are at work. Texting allows the prospect to be empowered to respond when it is most convenient for them as opposed to when it is best for a contact center agent.

Does the Drips team send the texts themselves or advise the client on when to send them?

We have our own AI that has overseen hundreds of millions of conversations to date, so we understand when someone says something like ‘I’m in a meeting at work right now.’ Since we have the necessary data, we can reach out to them to see what the right time for them might be to communicate via text.  Sometimes, it takes hours, days or even weeks to set a time when it’s convenient to have a two-sided conversation over the phone.

Is there any reason that texting is more effective than calling? Is it different for specific industries or the type of conversation that happened previously?

The context does have a great influence. It’s important to know that we support contact centers and are not limited to text only. We do implement IVR phone calls in our text messaging cadence.  The first message a consumer might get in our Drips campaign would be an IVR phone call. If they don’t agree to take the call (press 2 if this is not a great time for you) then they will be placed in the text cadence that is tailored for the campaign based on where the individual opted in from. Some people who have taken proactive action such as requesting a quote might pick up the first call, but that’s it. Most people do not respond to calls. This is not a trend that is going away anytime soon.  In many cases, my phone will block unrecognized calls   through my carrier. We’re at a point in time where we want to communicate via our own terms.

Sending SMS at scale can be a challenge, particularly if the company is not experienced and doesn’t know how to deal with what we call ‘throughput’ which is the delivery of an actual message. When a business is sending SMS at high volume, it needs to come up with a strategy or several different strategies. It’s part of the reason why our messages are what we call ‘humanized;’ we don’t just send one canned message to hundreds or thousands of people: we hand-tailor our messages using nuanced contextual personalization’s and employing multiple methodologies so that each one looks like --and should feel like-- it’s coming from a real person. 

What’s the Perfect Outreach Cadence? How much is too much?

It depends on what the person has indicated that they are interested in learning more about. We only communicate with prospects who have responded to a company’s offer or have expressed written consent for that company to follow up with them. If an individual filled out a quote request, they are expecting a call or other message in return. Similarly, if they have completed an application or set up appointment, they are expecting some level of follow-up. We’re just providing a better way to respond: one where 98% of people will see it within the first two minutes. We see our approach as “polite persistence.” We’re not trying to hammer anyone over the head, just following up on requests. We are constantly split testing different cadences and contact points to find the perfect balance. 

How does your AI contribute to your effectiveness?

Our AI understands many things. It references previous conversations to make these determinations. Our advantage is the vast amount of data we’ve seen and the corresponding trove of information we can use to train the AI. In a way, it’s similar to Tesla who has a head start in programming their cars because they’ve seen millions of deer, stoplights, people, other cars, or other objects that help train its system.  When their Model 3 drives down the road, it knows what one is and what actions need to be taken. Our AI knows the consumers situation from these previous conversations and can use the data to move conversations along. On the consent side, which is equally important, most SMS systems are short code, offer-based SMS. We’ve all gotten this kind of message with a link to some offer which also gives the customer the option to use some method to stop receiving them. Most systems have four to six ways to opt out. Our system offers hundreds of thousands of opt-out requests, including a middle finger emoji. This lowers costs, increases conversions and helps keep companies complaint. 

When you make that first outgoing call, how do consumers know it’s from the company as opposed to a spam call?

That can be a real challenge. They don’t and that’s why texting is so important. As far as “spam likely” calls, that’s comes from reputation management lists that telephony providers monitor actively. We provision our phone numbers as opposed to recycling them to ensure that they haven’t been on any spam lists. When we’re working with a regional provider, we will make it a local number to give it a local presence. We have found that the combination of an initial phone call followed by text works best. Depending on the company we can reach them even faster than a contact center.

Can you tell our subscribers how your performance-based pricing model works?

The cost has a lot to do with the volume of consumers to be reached.  The amount of leads or contacts coming into our system and the kind of environment that the company is working in. For example, an insurance company is very different than a window manufacturer. We customize each campaign and look at multiple ways to price them, typically on a per-contact basis.  There is a high level of consultation for each. We want to learn our clients’ business.  We need to understand what their lead generation process looks like, how the contact center is structured and how your operations side functions. Each Drips’ client gets a team, not a tool. Our staff of over 70 includes experts in scripting, industries, quality assurance, subject matter experts, client success and account managers. We want to provide the maximum amount of value. We’re a managed service, not a SaaS solution. Companies pay once for the full scale of activities for each individual lead. We’ve opened up tens of millions of dollars for our clients.

LumenVox transforms customer communication. They provide a complete suite of speech and multifactor authentication technology to make call center customer service faster, stronger and safer than ever before. They also support a multitude of applications for voice and facial biometrics, inclusive of passive and active authentication for fraud detection. And they’re compatible with major platforms including Asterisk, Avaya, Cisco (including BroadSoft), Dialogic, Enghouse, five9, Genesys and many others. Tom Farquhar, Regional Sales Manager at LumenVox, answers our questions.  

What are the key areas of focus for LumenVox?

Good question. We have zero-ed in on the following:

Speech Recognition: This includes conversational IVR with text-to-speech capabilities that create a very functional, conversational IVR self-service.

Voice Biometrics: We get really excited about this, as it addresses a huge threat to businesses right now—fraud. Voice biometrics is used to keep customers safe, to secure authentication and keep the customer experience painless and easy. Businesses can implement voice biometrics using IVR, mobile applications or in the contact center.

In 2019 we hit over a million voiceprints. For 2020, our development team combined customer feedback with their biometric expertise and created a robust passive engine, which allows for machine learning to enhance features and customer benefits. With the inclusion of Deep Neural Network technology, LumenVox has positioned itself to provide higher accuracy, more fraud prevention tools and increased customer satisfaction/ service.

There are numerous companies in your space…how do your solutions stand out from your competitors?

We’re easier to work with in every way. Our company is very focused on our channel partners. And our customers rave about how flexible the architecture and capabilities are. Our consistently high NPS scores (currently 89) reflect how much they appreciate our responsiveness. All speech recognizers and text-to-speech engines perform the same tasks, but they really don’t offer the higher value that comes from our simple install-configure-use approach. Things like our built-in diagnostics make it quick and easy to set up or troubleshoot things, and the LumenVox Speech Tuner is the easiest product on the market to implement and improve the performance of your speech applications, based on real customer usage.

On the voice biometrics side, while we do the installation and training, we’re also very open and flexible: We have APIs that are easily integrated into a contact center agent’s desktop or CRM applications. We see this as an important differentiator because managers really want applications consolidated. So the capability to take our biometrics results and integrate that information smoothly makes it easier for everyone.

From an implementation services standpoint, we don’t lock our customers into having to use our professional services. We enable both our partners and IVR developers to work from various platforms, using their own services while deploying our speech recognition technology. Our holistic approach results in us being more competitive, or and cost-effective. We don’t think cutting-edge technology needs to come at a premium price. We want this technology to be available to everyone.

Can you define how biometrics work in the contact center?

All biometrics measure something you’re made of. We’re used to smartphones using fingerprints and facial recognition/ faceprints to authenticate us now. Makes you feel more secure, right? The call center is evolving, too, and with our technology can create the same level of protection using the human voice. An enterprise contact center obtains a sample of your voice and converts it into a secure file called a voiceprint. Once there’s a voiceprint on file, the next time you call into the center, you don’t have to answer those painful security questions (which, by the way, are often vulnerable to theft). Using a voiceprint means that as a consumer you have a better calling experience. And as a business, you get more security.

Is it possible to fraudulently manipulate a voice biometric? Can someone pick up a customer’s voice pattern or convincingly imitate them?

Well, even mothers can be fooled by twins. We don’t want to be so hubristic as to claim that we somehow have something over Mother Nature. But voice biometrics takes major precautions: We use multiple factors when creating the voiceprint. The human ear might not be able to detect an impersonator, but our solution will notice hundreds of subtle differences. A company can also use multiple phrases as an identifier or ask varying questions to prevent a breach.

The truth is customers understand that the threat is real and want an added layer of unique security. A recent study noted that 74% of Americans believe that biometrics is a more secure method of verifying accounts than traditional PINs and passwords. Hacks and data breaches are commonplace occurrences now. And there are long, long gaps in notification—customers may not know they were exposed for nine months sometimes. Since many people tend to use the same PINs and passwords for multiple accounts, they are vulnerable. Voice biometrics protects and defends privacy. It’s that simple.

In what industries are recognition technologies becoming prevalent?

Any and all, as the desire for a seamless customer-agent interaction increases. People want to reach a customer representative to solve the bigger, more complex problems and spend their valuable time self-servicing/ solving the easier ones on their own. We’re working more and more with financial institutions and healthcare providers, as we have the capabilities to not only enhance their IVR experience, but also provide stronger security. As businesses grow, we see these two factors go hand-in-hand. People want security, but they don’t want to compromise speed and efficiency to get it. Our —speech solutions provide the best of both worlds.

What are some of the tangible benefits of Natural Language Processing IVR applications?

NLU gives self-service that human touch that people really seek in the customer journey. It’s the best of all worlds: Customers can help themselves quickly, but can also feel as if they’re doing it effortlessly--with a fellow human mind at the helm. As for LumenVox, our new Conversational ASR combines 20 years of experience in Speech Recognition with the latest in Artificial Intelligence & Machine Learning, allowing any business to build new AI-based IVR applications which support natural language processing and intent determination from an existing voice application platform (IVR). The best part is that as a business you really don’t have to start from scratch to do this. Text-based AI tools can be given a voice with LumenVox ASR. You can leverage your existing infrastructure and preferred tools to provide rich, voice based self-service that exceeds expectations. 

Tethr enables organizations to unlock the insights provided in voice interactions as well as those on other channels. While 70% customer communication still take place on the phone, few businesses tap the valuable information contained in these direct interactions to improve customer experience, sales conversions, compliance, or operating margins. Tethr’s Conversation Intelligence platform employs AI and machine learning in listening to, accurately transcribing, and analyzing conversations in real-time, giving you searchable, actionable insights.  Dean Cruse, VP of Marketing for Tethr and Matt Dixon, Chief Product and Research Officer, gave us a closer look at how it works.

What are some of the ways Tethr empowers enterprises to better understand their customers?

Our voice analytics platform enables these companies and their leaders to access the data in conversations, whether phone recordings, chat interactions, case management records from organizations such as Salesforce Service Cloud or Zendesk, SMS exchanges…anywhere where there is dialogue between a company and its customers. Our most compelling use case is to get companies out of the business of sending out post-call surveys. Since our technology can capture the entirety of the conversation with the customer, the algorithms we’ve developed are capable of effectively predicting scores that a customer might have given, skipping the tedious process of filling out and evaluating a survey.

It’s proven quite transformative: businesses get a much bigger sample size, lots of actual feedback, the ability to close the loop with customers, fix problems at scale, and target coaching for agents in ways that were not previously possible.

How does your solution help companies know what customers are thinking when sometimes even surveys don’t ask the right questions?

Research has been done in providing the ‘Effortless Experience’ to measure customer effort scores. We’ve taken it a step further by building an algorithm we call the Tethr Effort Index, which is based on more than 200 separate variables derived from conversations. It runs the gamut from elements such as the duration of the call to silence time, overtalk and more.  But it also includes sentiments such as customer frustration and confusion which we’re able to pick up via machine learning and AI. It also incorporates both good and bad agent behaviors: positive language, the rep demonstrating advocacy versus hiding behind policy. It also includes things said by customers before the conversation even starts: “your website or app was extremely frustrating and now I have to call,” or “the field technician never showed up to fix my problem.” The Index picks up product issues and competitive mentions. We built it by taking tens of thousands of completed surveys from a host of companies across a broad spectrum of industries where the customer had given an effort score. We then trained an algorithm to predict what that score might be from raw audio on the back end.

Does the platform include looking at tone and keywords?

Speech analytics is a technology that has been around for 15 years or longer and many of the products are based on keyword spotting. This could be a negative word, a competitive name, or product name. But it also could be tone such as audio characteristics such as pitch or inflection, elements that help us determine if customers are amped up, angry or confused. Compared to these keyword spotting tools, Tethr is focused on is the syntax of what is really being said in the conversation. In phone calls, we transcribe the audio and train the machine to spot more than just keywords, but broad concepts.  It is much more challenging to have the machine discern when a customer is expressing frustration in order to recognize all the approximately 350 utterances and phrases that constitute the concept of frustration. But, we’ve conclusively taught the Tethr platform to accurately recognize these concepts so now we can dashboard them and track them and count them as variables in our model. Similarly, advocacy in agent behavior is a complex concept. We’ve learned the words and phrases that demonstrate that an agent is communicating that he or she is on the customer’s side. Along the same lines, Tethr can identify language that indicates when agents are either hiding behind policy or shirking responsibility.

While we do translations into other languages, we are mostly North American-based and the machine is trained in English. However, we can train the machine to spot phenomena such as frustration even when the cultural ways it is expressed are different: we can also train it to spot customized characteristics to meet the needs of specific clients. 

What level of person is required to effectively use your platform within a company?

It can be administered by someone at a relatively junior level. To effectively use many other platforms on the market, a company would need a data scientist to do the coding. Our “coding” is as simple as thumbs up or thumbs down: sort of like what people do in listening to Pandora.

If someone want to track a problem that is specific within an organization, they come up with a few sample phrases and feed it to the machine which comes back and basically says “how’d I do in identifying this type of behavior?” When it provides examples from different calls, the person who made the request can communicate which ones are relevant and which ones are wrong. When it reaches the point where there are no more mistakes in how the issue is interpreted, the person making the request can simply hit ‘publish’ and everything is good to go. Appropriate titles to set such parameters can include Quality Assurance Supervisors, Contact Center Managers, Product Specialists, and others. We send them through a training program that takes just hours to teach them how to do this on their own and they are excited to be able to control AI and machine learning. We find that is more important that the people know the company’s business problems and how they are articulated by customers as opposed to technical knowledge. When businesses delegate the task of creating a machine learning category to a data science team, they can get the procedures wrong because they don’t understand the problems.

When the company has the data available, what do they do with it?

In cases where businesses can use the Tethr Effort Index, for example, they can use the score assigned to each individual interaction. As someone said to me at the most recent Las Vegas CCW, ‘Now I don’t have to worry about sample size or whether a survey was only filled out by angry customers. How do I use the information?’ I laid out three different use cases. In our experience, about 15% of calls are going to be in the highly negative range where a lot of effort was required. In those difficult situations, the customer is more likely to churn out and less likely to make any further purchases, often becoming a detractor on social media. So, if a company knew who that 15% were, it presents a powerful opportunity to close the loop with them. This does not necessarily mean making special offers but simply acknowledging that the company has not lived up to expectations and then ask the customer how it can do better.

The second use case is using the Tethr Effort Index as a replacement for someone sitting with headphones evaluating calls in the traditional QA process.  A business can be aware where both good and bad calls are happening, tracked by call reason and agent or by team and supervisor. Management can go deeper in finding out which reps are responsible for increasing effort by the customer and delivering more impactful coaching to modify these behaviors.  The final one would be spotting issues that cannot be fixed in the contact center such as product deficiencies or website problems. Once these insights are uncovered, the company can work to fix these issues to prevent the flow of calls concerning them.

You mentioned coaching. Does your platform enable companies to be more objective than subjective in their approach?

One of the biggest problems of QA is that it always has been a manual labor-intensive process that consumes a lot of time and can be quite costly. This has led to companies sampling no more than 1% of call volume, if that much. In addition, in many environments, the QA process runs side-by-side with an appeal process where reps can complain that they are graded only on their worst calls. This means more manual labor and becomes very cumbersome. More to the point, QA doesn’t always work the way it is intended to which is of course improving quality. It often becomes a compliance-oriented, check-the-box activity. We have found that it is a lot more fair, objective and effective when a machine does the listening as opposed to a person. It globally addresses identifying the reps who don’t know how to deal with specific product issues or are hiding behind ‘powerless-to-help’ rhetoric which allows the company to address the important work of helping them perform better. One insurance firm we worked in was able to redeploy the 100+ people they had doing QA to coaching,

Do companies keep their data proprietary, or can it be shared in some fashion?

We have numerous clients who each have practitioners creating their own machine learning categories and teaching the platform to do new tricks. While we respect that many companies want their data to be private, we can in some instances use leverage the code some clients have developed to establish best practices, standardize it and share it across the breadth of our customer base. It’s almost like crowdsourcing machine learning.

Trendzact operates on the principle that the reputation of a brand is largely determined by the quality of its customer engagement. TrendzAct offers a cloud-based omnichannel CRM platform that incorporates routing optimization, workflow automation and business analytics, enabling contact centers to build more enduring customer relationships and meet rising customer expectations.  Their method contributes to rapid results improving CSAT, NPS and customer retention rates. Matt Gabrielson, President and Product Visionary at TrendzAct, gave us an in-depth look at the benefits of the contact-center centric platform.  

You characterize your product as a CRM system: why does the market need another one now?

CRMs have been around for a long time and most people don’t see them as being all that interesting. But because we are highly selective and don’t try doing everything for all companies, such as ERP and point-of-sale functions, we are able to focus on the unique needs of contact centers. I was an agent when I was in college, I’ve owned a BPO contact center, and both of my sons have done QA while they were growing up.  So contact centers are in my blood and so is emerging technology. We decided to build a CRM that concentrates on the agent experience. The agent is the front line of any company and if you have a CRM that gives them the tools, then customers will soon know that they are in for a good experience and where they get reliable information. Our focus is strictly on mid-size to enterprise contact centers with complex integrations. We designed the platform specifically to be on Amazon Web Service (AWS) and we natively utilize the Amazon machine learning tools to create and deploy AI right into the CRM without add on products from an app store or relying on a 3rd party provider integrators to make it work.

Most CRM systems already incorporate purchase data, ticket history and other information needed by contact center agents. How does yours go beyond that?

We look for complex integrations. Some contact centers need to have access to external legacy systems, such as order history that must be retrieved from another system or data lake. Our platform brings in the external data or accesses the external integration, so the agent doesn’t need to do any swivel-chairing or double entry copy and pasting.  The agent UI also provides automatic recommendations for the interaction. Let’s say that a customer initiates a web chat.  The agent starts a ticket and the web chat is analyzed real-time so when our system sees certain keywords in the conversation, it steps in to recommend what should be asked next. It provides an optimal path so the rep can guide the customer to issue resolution. That’s a true differentiator: it doesn’t assume that the agent is autonomous in a silo and hope that they learned everything they should have in training. This hybrid approach guides the agent along the resolution path or to take advantage of an upsell opportunity. This in turn reduces agent training time. Instead of having to learn all the nuances of a business, the system, based upon the inputs of the customer in a call or webchat or SMS exchange, the system prompts the agent ask certain questions and it will provide ongoing prompts based on the customer’s responses. For instance, in cases where the customer wants to update a reservation, the agent can go down to the reservation section where all the information is pre-filled and relevant options are automatically displayed. The capability of our native AI to effectively guide the agent is the true differentiator.

The scripting is highly visible and uses the Amazon Personalize engine based on the same technology used at Amazon.. Over time, it uses machine learning to automatically update guidance information to indicate the best source to make the responses even more precise. It is of course not directed at one specific agent but distributed throughout the business to all front-line personnel. Agent skill is also accounted for: the information is customized to be routed to specific queues for different skills or specific products.

Is it possible to populate to the agent view as well as provide an integration to other self-service channels people are using and the company website?

Our Dynamic Lookup, which is the module we use for the machine learning, also ties into an existing company knowledge bases. One of our clients has a very complex internal financial knowledgebase. So, as agents type in specific parameters such as credit card type or error codes, not only does it populate the script and identify the root cause analysis, but we also pings back and present information from the company’s community forum and internal knowledgebase. This makes for a truly holistic approach where it’s not just one piece of data but multiple sources providing options and conversational texture for the agent. And over time, when the agent acknowledges that a knowledgebase article is helpful, the machine learns that and will rank that at a higher rate.

How does someone start the process?

Our CRM is a strong replacement for existing CRMs or even homegrown systems.  A company can start by taking what it already has and replacing dated components and enhancing others. We have our own data scientists to serve our clients, so they don’t have to hire a team of learn the AI processes. Our  data analysts will go through what the company is doing and identify the best use cases. We can help the client decide what they want to get out of AI: is it looking for anomaly detection for cases that are outside the norm which could entail 30 different datapoints? Or is it examining metrics like average handle time (AHT)? Or to reduce after call work (ACW) by eliminating the time spent by agents having to type in the root cause? Or let’s transcribe the audio or the webchat and automatically post it. Once we identify a company’s most important priority, we can help it set attainable goals. Then, we build that model and provide a Proof of Concept which we can optimize over time. Trendzact CRM with AI is not just a black box or a slogan, but a concrete plan which produces meaningful, actionable data. Then, we repeat the process to identify another business case.  Our business analysts know the contact center landscape and are fluent in all the metrics and acronyms which shortens the learning curve. TrendzAct helps them determine what is most important to affect significant improvements.

Once the business analysts have identified the use cases, the data scientists will do what I call the geeky work of figuring out how to solve the business opportunity from a 500-word document and write the algorithms to make it work. The teamwork between the analysts and scientists  can be compared to the relationship between a sous chef and a master chef.  The scientists come up with new ways to approach a problem and the analyst refines it for the client. It’s always about actionable data. We never start a project just for the sake of doing it.  We stress giving measurable value to our clients. We also strive for continuous improvement. The theories behind algorithms for artificial intelligence are changing all the time and there are always new tools being made available. We know we must prove our value every day and we tell our clients if we’re not finding new ways to earn your business, then they should fire us.

We know we need to provide more than just linear reports. It falls on us to give meaningful, actionable data to contact center leasers who want to make strategic decisions. They already know their metrics quite well. Our AI can go above beyond and measure 30 or 40 additional datapoints that might not come up:  such as all of calls from Milwaukee to a specific queue and skillset are out of range compared to all of Wisconsin. And that is the kind of information that can make a difference which would never come up in a flat report but can only be surfaced with AI. We also provide real-time alerts that go right through to the supervisor.  We also offer ad hoc reporting and enable companies to build their own dashboard as well as numerous self-service tools.

How easy is your platform to use? What skill level does someone need to build their own reports?

We want to make it as simple as possible. When we built this product, we didn’t want our customers to need a technician or an in-house developer. Many fields like statuses, drop-down boxes and layout …even the ad hoc reporting… are drag and drop. Someone can do their own reports then easily share them. We do offer professional services for those who want to build highly complex reports but most of what we offer is intuitive and self-service.

When we go into for a consultation, we ask three questions about their CRM and reporting: What do you like and don’t want to change? What do you hate and must change? And what is your CRM and reporting wishlist? From there, we can build a good requirement assessment:  here’s how we can help you continue to do what you do well, here’s what we can fix and often, even the wish lists are attainable. Of course, sometimes we do tell companies that some things just can’t be done. Part of being an honest vendor is knowing when you have to say “no”. But with the technology we make available today, we say “yes we can” much more than in the past. 

Unymira is is a division of Aspera Technologies Inc. and part of the USU Group with offices in the United States and Germany. Its Knowledge Center platform provides a comprehensive knowledge base for contact centers that empowers customer facing agents with relevant, easy-to-use to understand knowledge so that they can provide better, more efficient customer service. Chris Rall, Sales Director North America, offered insights on the company’s capabilities.

 How does your knowledge management platform help improve contact center performance?

Our goal is to provide a better customer experience for a company’s customer. We work to centralize all its internal knowledge in one place so we can help them bring this content to different channels. One of the challenges is that businesses quite often have knowledge siloes which is confusing from an overall customer journey perspective. A website visitor often gets different information than they would from speaking with a company rep. This can be a huge problem because then the information provided is not transparent and it’s unclear what is actually correct.

Can Knowledge Center be accessed at any point across channels?

Usually, the first step we do is called knowledge mapping. We identify where a company’s knowledge is currently stored, which is often in multiple places. Then we centralize this information in one single place. The good thing is businesses can segregate content within it to more easily administer data and distribute it to different channels. When a business is in charge of its work content, they can take one piece from their overall content stack for their website, another piece for their chatbot and another for their IVR while all of it is available to their agents. 

So, when the platform is in place, do all updates go throughout the organization?

Yes.  When the data is centralized in one location, as soon as you update it, it is instantly available in every channel. Whether the data is internal or external, it comes from the same source. We want to avoid customers getting different information or answers in different channels, for example on the phone versus on the website.

On the self-service side, the customer sees the same information as the agent does, although the agent might see additional internal information that can be suggested as next-best-action.

Have you found that when customers can access a more robust database, they call less frequently because they are able to better self-serve?

There are two things that we see quite often. Usually, what we see internally is that the overall phone experience improves. Some customers may still call first, but since the agents feel more confident they can provide accurate information quicker , that helps the business from the customer experience side. 

The second part of this is if the organization can provide the right information on the website, it not only reduces call volume, but makes for happier customers. In the 21st century, fewer and fewer people want to pick up the phone anymore, especially for simple FAQ information which involves a response to a single question. Sometimes, it also depends on the industry and what is the product that they are inquiring about. Utility and mobile phone bills are something that people can just look up without a call, but a credit card bill is more personal and variable, so people still want to call to have a more deep-dive conversation with an agent.

In what ways does having information more easily accessible benefit the agents? Are they happier having this type of system?

The number one challenge is the search. When you’re talking about a knowledge base, what agents expect is a more Google-like experience. We all know it from our personal lives; usually when we search something, we type in a keyword and are able to easily find the information we need. We try to duplicate that for agents so they’re confident the information is accurate and up-to-date—which is what they need to do their job— and they always know exactly where to find it and just need to do a simple search. 

The most significant impact that we see-- besides reduced average call handling time and first contact resolution—is the turnover rate. The one constant challenge in the contact center space is having agents leave because they are frustrated. What we have found is that if you provide them one single source of reliable, updated information, the confidence and satisfaction of agents goes up, which translates into lower attrition. 

No matter what technology a business provides, the job is still challenging because agents are dealing with customers on the phone and it’s an understatement to say that some are not in the best of moods when they call so agents always need to try and maintain a positive attitude. If you combine this with having trouble finding the information need and having to deal with data stored in multiple places, it becomes even more frustrating. 

Does having the knowledge base shorten the amount of training time necessary?

One of the advantages we offer from an ROI perspective is that we target the onboarding time for new hires. For one customer, we were able to reduce the onboarding time from six weeks to one week. Since the knowledge base eliminated the need to search many different places and always remember what information is where and because we have incorporated e-learning in the platform, training is far easier and definitely faster. 

Agents just need to refer to content that is readily available and combine it with what we call internal knowledge text and they are ready to go. Another advantage is that from the first day agents come into the organization, they are introduced into the knowledge platform which advances their satisfaction level. We offer our own training, sometimes integrated with the company’s Learning Management System (LMS), but even if the company doesn’t have one in place, the e-learning we provide is usually enough.

If a company deploys chatbots, are they using the same system that an agent or customer would be using?

Once again, we want there to be one single repository of information. In our world, a bot is simply an additional channel to access it. Our bot is fueled by the information from the knowledge base and we combine it with natural language processing so the bot can communicate from the customer side and retrieve information from same base of content that the agent uses. 

If a company wants to get started and their information is in multiple places, what is the first step in pulling everything together? How long should it take to centralize it?

Usually, our recommendation that the first thing that needs to be done is to ask the agents what content is being used in their daily activities. Because there might be a lot of information that is not being asked for at all and we don’t want to integrate it. The agents are the main stakeholders in the process and can tell you what they need to be successful. Businesses can see where the right information is stored which can then be mapped out and transitioned to the one single source of truth.

The timeframe for implementation depends on the number of documents a company has in multiple locations and number of agents they need to have access to them. Typically, the implementation process takes three to six months. That is of course for the additional rollout and we recommend that they make it a continuous process, so the issue doesn’t come up again a year or two down the road. It’s crucial to ensure that the information is updated regularly.

What sort of results are your clients experiencing?

In terms of ROI, it’s the three KPIs we’ve discussed: onboarding time is faster, average handling time (AHT) is reduced, agents spend less on the phone because they can quickly find what they need and third the first contact resolution rate (FCR) goes up dramatically. One company said they have a decision tree in place to guide agents through a call step-by-step. This is particularly important for businesses that have a long service process. 

If we are talking about the customer-facing side, it’s the overall reduction in call volume that fuels ROI. While we can’t measure it empirically, it is logical to extrapolate that this also has a positive impact on NPS. It also obviously contributes to lowering customer effort. We carefully examine customer feedback to see if there are any problems and usually see very positive comments. 

Businesses need to think about where they are going over the next few years: everyone talks about chatbots, but not all companies are ready to take that journey. If they map out their complete customer journey, it involves different channels and touch points. They need to examine each channel and ask themselves if they are prepared to promptly provide accurate information on it: this includes having the right process, people and technology in place. Having a comprehensive knowledge base is the first step toward delivering a better customer experience.

VoiceOps is a platform that enables sales call centers to bring all their call coaching functions together in one place: whether it’s sending agents a quick note, setting group sessions, or formal 1-1 training.  It allows supervisors to leave feedback for reps directly on the transcript and have asynchronous feedback discussions without pulling the team off the phones. Maxx Reiner, Enterprise Sales Manager for VoiceOps sat down with us fill in the blanks.  

Can you give our subscribers an overview of your solution?

VoiceOps was created for high-volume contact centers that engage in more transactional conversations. We operate in several core verticals: insurance, financial services, travel, collections, online education…any type of business where a sale can be closed in one to three calls and reps are making 50 to 100 dials a day. What we do is embed our tool within the organization to transform the way they coach and train on a daily basis. Traditionally, contact centers have a difficult time with this because the manual processes they use are very outdated. Our goal from day one is to remove those inefficient processes and allow for more effective and efficient training.

What are some of the less than optimal processes that need to be updated?

It’s an across-the-board problem that permeates the entire industry. Before we start working with them, just about every customer goes about training the same way. They take a call…often at random…and listen to it in its entirety and perhaps jot down some notes on a sheet. In order to coach the rep, they need to sit down in a 1-to-1 session where they might listen to the call again. In the case of a 20-minute call, it might take an hour to listen to it and another hour to do the actual coaching. This could involve someone trying to bundle an auto insurance package, or for someone booking a family travel reservation or in financial services, getting information on a loan or even collections. So, if I’m a manager overseeing 10 or 15 reps, I only have the capacity to do this once every week or two. Reps are thus left in the dark on what skills they need to improve: the process is extremely inconsistent.

How do you go about making changes in these organizations?

At the highest level, we take calls and transcribe them to extremely accurate text. We operate at about 95% accuracy. We then take those transcriptions and push them through a machine learning model that uses the text to identify the key behaviors that the rep is capitalizing on consistently or missing out on. The benefit is that the manager now has this dashboard that tells him or her where the rep is either missing the mark or doing well. The manager can then go through the individual calls without having to listen to them by reading the transcripts and tell the rep where they did a good job or what needs to be improved.

What improvements are being made by companies who use your platform?

We initially measure at two levels: our first goal is to enable companies to coach more frequently, perhaps five to ten times per month for each rep. That’s the first step. The next thing we do is work on behavior change. Most contact centers have a call flow or call script that reps are expected to follow. Our platform customizes the behavior for which it analyzes based off each customer’s call script. We can then inform the company how often the agent is either closing the sale or at least attempting to do it correctly. Let’s say that in a specific environment, that figure is 15% of the time. So, with more frequent and more effective coaching, we hope to see closing rates and proper procedures go up to 30% or even 50%. That ultimately leads to less time wasted and more profitable operation.

How does the agent receive the added coaching?

Ours is a manager-driven tool. Most of the set-up involves a manager going into transcripts and taking advantage of the functionality that allow them to provide feedback to individual reps. The agent will get an email stating that the manager has coached them on a specific call and will then be able to open the dashboard to see the coaching on it.  That’s not always how it’s done: managers might still bring in a rep for a 1-to-1 and review individual or multiple calls together.

Is the coaching mostly for outbound calls as opposed to customer service?

Our platform is used for sales: in some environments, that is a mix of inbound and outbound, but the majority of training is for those placing calls.

How does call calibration work in your company?

Before we start working with a client, our in-house customer operations team gets on the phone for a calibration call with whoever is responsible for training reps and whoever is determining what the call flow should be. The goal is to gain a better understanding of the behaviors they want from the reps.  We take that information and create definitions out of each behavior and use them to identify situations where the behavior is occurring. We then calibrate our model according to their desired training goals. Calibration is done with a manager who does the training and the person who is setting up call flow, which could be a VP of Sales or even a QA specialist.