Verint Analyst Days – Four Takeaways

My latest travel run had back-to-back events the same week, starting with Zoom Perspectives in NYC, then on to Dallas for Verint’s Analyst Days event. I’m doing separate reviews for each based on the same format, so this article covers my Verint takeaways, and my takeaways from Zoom are here.

Fellow BCStrategies colleague Blair Pleasant attended Verint as well, and her coverage is here. I urge you to read Blair’s review as well, and to provide a more complete picture, I’ve tried to minimize any duplication of her coverage.

1. Why CX Automation is the way forward for Verint

Every technology vendor – especially those who are not household names – needs a thing – a clearly-defined offering and value proposition that distinguishes them from the competition. At last year’s analyst event, that thing for Verint was Open CCaaS, a concept that most of us couldn’t grasp. We got the Open part, but not the CCaaS part, which seemed like a stretch given their pedigree as a WFM vendor, and their more recent focus on chatbots for customer service automation.

Clearly, Open CCaaS didn’t take, as we didn’t hear a single mention of it at this year’s Analyst Days event. Nobody asked about it, and I don’t think anybody missed it, but we were all ears to hear CEO Dan Bodner and others talk about CX Automation, their flagship offering. In terms of takeaways from the event, this was the first one for me – CX Automation (aka CXA) is now their thing. They talked about CXA last year as well, but it wasn’t the core theme, and I do like the focus.

Part of allure seems to be that the CX Automation isn’t really a mainstream concept, so it’s easier to position yourself as a leader to stand out. Verint isn’t the only vendor, but they view themselves as having the most complete, advanced offering. This is a fairly defensible position when compared to major CCaaS vendors, for whom CXA isn’t a core focus. There are also vendors in the helpdesk space – like Zendesk and Capacity – who are very strong in CXA, but not so much for the entire contact center environment.

I’ll stop there, as these comparisons get messy, and the analyst community hasn’t done its own due diligence to validate Verint’s positioning. That said, Dan Bodner and his team clearly understand the challenges contact centers are facing, along with the need for more CX automation. More importantly, Verint’s focus for enabling automation is not for sake of efficiency, staff reduction and cost savings, but for driving business outcomes. That’s where AI really brings value and can transform contact centers from transactional service operations to strategic hubs that grow the business.

In short, a CX Automation solution requires an open platform that integrates with everything else the contact center is using, AI-powered bots to perform various automation tasks at scale, and the ability capture data across many sources and convert that into intelligent applications to improve CX as well as agent performance. Contact centers are struggling to transition from being telephony-centric to becoming data-centric, along with deploying AI to automate effectively, and that’s where Verint is having solid success.

2. Behavioral data is the key to better CX

Having a degree in Social Psychology and being a 30+ year market research practitioner, this takeaway really resonated for me. I rarely hear anyone talk about behavioral data in this industry, and was very happy to see Verint connect the dots here for making CX better. I have never understood the appeal of vanity metrics like CSAT and NPS, which provide zero indication of how customers will behave. Intention or sentiment is one thing, but it’s behaviors that drive business outcomes.

This is why contact centers have always been potential gold mines, since they are data-centric by nature. Agents are managed by a plethora of KPIs, and with the rise of digital channels, contact centers now have incredibly rich veins of customer data coming their way 24/7. Some of it is behavioral, and some of it not, but this is where the magic of AI comes into play. Not only can AI capture unlimited amounts of data across all these channels, but it can be processed, analyzed and fed back to agents at wire speed.

Never before have contact centers had so much data to work with, but the vast majority lack the tools to harness it, and even fewer fully understand the opportunity. Until contact center leaders recognize the value of all these data streams across multiple channels, they will remain siloed and unconnected. They will have some level of informational value, but when filtered through AI-based analytics, the data can be transformed into powerful indicators of customer behavior. However, contact centers cannot do this on their own, and this is where Verint’s Data Hub provides the solution (see Blair’s coverage for more detail on this).

3. Getting to the Eureka Moment with CX Automation

This was another message that resonated for me, and CPO Jaime Merritt explained it very well. Basically, when he talks with customers, they think “rip and replace” when it comes to modernizing and addressing their CX challenges. As analysts, we only see the leading edge with vendors, but it’s a very different reality in the trenches, where adopting new technology is not a simple matter of flipping a switch.

As well, he finds that many customers have been trying for years to modernize, but just aren’t seeing the benefits. There are many reasons for this, and we’d need a few articles to cover that ground, but not today. Bottom line – contact center leaders want to deploy new technology to help their people be more efficient, and they know the answer isn’t to keep hiring more people.

Invariably, these conversations all lead to the same conclusion – the need for more intelligent forms of automation both to make CX better and make agents more effective. That’s what AI brings, and when framed as a path to better business outcomes – rather than cost reduction – the benefits become clear. Plus, with Verint, they can do this today, not years from now. As Jaime says, when they hear that message, “you see the lightbulb going on, and that there is an alternative”. Eureka.

4. Differentiating on transcription accuracy

Verint’s CX Automation isn’t about creating a super-bot that does everything with infinite capabilities. Rather it’s about having an army of bots, each built and trained for a specific task. Last year, they talked about having 40 such purpose-built bots, and now it’s closer to 100. I think this is the right approach, and during the sessions, they walked us through several specific examples, such as wrap-up bots, coaching bots and quality bots.

Blair’s article covers this further – both in her narrative and video interviews with Verint executives – but I wanted to detail one in particular, their Exact Transcription Bot. I’ve been closely following AI-driven speech technology for years, so this one caught my eye. AI is very effective at analyzing text, as there’s less ambiguity than with the spoken word. However, not only is speech much more difficult for AI to work with, but most interactions with customers are speech-based. The use of text channels is clearly growing, but for CX Automation to deliver full value, the AI tools need to handle speech really well.

It turns out that every vendor offering transcription claims to have the best accuracy – nobody wants to admit to being anything less than number one in this space – so you have to take these claims with a grain of salt. While it’s fair to say that all transcription engines are pretty good, small edges in accuracy can make all the difference given the incredible richness of speech and the human voice.

I have long maintained speech is the most important frontier for AI to conquer, as it defines us as being human more so than any other faculty, and it has impossible complexity and idiosyncrasies that makes even the best AI application look foolish. As such, hitting 100% accuracy is pretty unrealistic, but AI doesn’t need to be that good – it’s already reached 95% at a broad level for speech, and that’s good enough for plenty of use cases.

The standard metric for accuracy here is Word Error Rate (WER), where the lower the rate, the more accurate the transcription. As per Verint’s data below, they lead the pack at 12.87, with the next closest being MSFT Azure at 15.05. In the spirit of creating a matrix to show competitive advantage, the other axis is cost, defined by dollars per kilo minutes. The numbers are drawn from data among their customers’ contact centers, and based on this price/performance model, Verint shows best, both for transcription accuracy and cost efficiency.

Having the best performance at the lowest cost is a tough value proposition to beat, so if you take these numbers at face value, speech transcription is a real differentiator for Verint. I can’t vouch for how true this is, apples-to-apples across vendors, but the premise holds for me, and if I was buyer, this would be a pretty strong selling point.

Conclusion

There’s a lot to like for what Verint is doing, and compared to last year, I think there’s a clearer focus and sense of purpose here. Yes, they’re going all-in with AI – but really, who isn’t in this space? – and based on solid financial and business metrics, they seem to be executing quite well. Their heavy focus on bots may be ahead of the curve, and with so many challenges to manage, contact center leaders may have trouble seeing CX Automation as their silver bullet.

Also, there’s no guarantee going this deep with automation with make CX a more human, authentic experience, so there’s always a risk that contact centers will keep it back of the house, with limited customer-facing engagement. CX Automation really needs the latter to deliver full value, and that might require earning a higher level of trust than currently exists with AI.

That said, Dan Bodner provided a commonsense notion that ties all this together nicely. His take was that if you give good customer service, they’ll call you more. Conversely, they’ll call you less if CX isn’t good, and from there customers will buy less or worse, they’ll leave altogether. However, if CX Automation makes them happier and they call you more, guess what – you’ll need more capacity, and AI is the best way to provide that.

As such, there’s a virtuous circle with AI, where better business outcomes drive more customer interaction, which in turn drives the need to provide more customer support. This would be an impossible equation to solve with legacy technology, but it all fits with AI, making this a bigger – and better - story than just deploying some bots.