Journey Analytics and Artificial Intelligence
DCX Podcast #4
An interview with Will Thiel, the founder and lead product architect of Pointillist the customer journey analytics and orchestration SAAS tool. Pointillist was recently acquired by Genesys, the global leader in cloud customer experience and contact center solutions. Will has been at the forefront of bringing AI and machine learning into the customer experience domain and he is now the head of the core platform for Genesys Cloud. I'm excited to speak with him today about journey analytics, his path to journey analytics, where he sees things going in the space, and AI and machine learning in particular.
7 Key takeaways from the conversation:
There are three important components to Journey Analytics: a. the ability to describe milestones; b. the ability to look at the patterns and aggregate, and c. the ability to look at what happens when individuals diverge from those.
Journey analytics is a practice, not a software solution or technology
The industry/practice is relatively new and there aren’t any best practice standards set
Three main use cases that span across industries include a. Digital containment; b. Escalation management and c. Product management
The contact center is becoming much more of a strategic opportunity to improve the customer experience.
Journey Analytics needs to be an integral part of the operations of a business and the contact center, not siloed with the analysts
AI is not a panacea for companies wanting answers. It’s becoming more of an amplification of what decision makers have to do, vs making those decisions in areas like agent allocations and staffing, routing of calls, and root cause analysis.
Welcome to the DCX podcast where I interview leaders in the customer experience space about how digital is changing the landscape and how you can leverage these changes for success in your business.
Today I'm excited to be talking with Will Thiel. He's the founder and lead product architect of Pointillist the customer journey analytics and orchestration SAAS tool. Pointillist was recently acquired by Genesys, the global leader in cloud customer experience and contact center solutions. Will has been at the forefront of bringing AI and machine learning into the customer experience domain and he is now the head of core platform for Genesys Cloud. I'm excited to speak with him today about journey analytics, his path to journey analytics, where he sees things going in the space as well as AI and machine learning in particular. So, Will, thank you for joining me.
Thanks, Mark. It's great to talk to you again.
So maybe we could start with just give us a little bit of background on yourself. You know, where you grew up, studied, kind of how you got into journey analytics.
Yeah, sure. So I grew up in sunny Buffalo, New York. And I did everything that I could to build downhill vehicles and injure myself with all sorts of crazy contraptions and then decided to pursue that professionally with a mechanical engineering degree. And then despite all of that activity with machines and real stuff, I somehow ended up getting sucked into the vortex, of big data analytics, post-school, and so completely ended up in the abstract non-material world of data and math and systems. And actually, I mean, at the time, it was hard to avoid that world. As you may remember, the data scientist was at one point in some professional magazine called the sexiest career, which at the time, and still, seems completely incongruent with who a data scientist actually is. But I did end up following that path.
I spent some time in management consulting, running an analytics practice inside of a consultancy that gave me a lot of exposure to the problems that or challenges that were being faced, specifically in telecom. And, as you know, in telcos there are many different channels through which customers interact with the business, including outbound channels and inbound channels, passive and active. And the challenge that our telecom clients were facing was figuring out how the interactions across all of those channels impact customer health and business performance. And at the time, I couldn't, we couldn't, find technologies that were sufficient to give us that view into the data and what was going on. So that problem spawned in the end, what ended up becoming our startup, Pointillist, which was a technology built from the ground up for journey analytics to be able to answer those sorts of questions, and over time, we developed a partner ecosystem which included Genesys and then, of course, recently, we had enough overlap in our objectives that made sense for Genesys to acquire us and now I'm part of Genesys and in addition to working on facilitating bringing the Pointillist in Genesys platforms together, now playing a deeper role within the core platform of Genesys cloud itself.
So before journey analytics tools were available, right before they existed, sounds like companies had all sorts of different ways of assessing journeys and major pain points. So what specifically were you trying to solve with creating Pointillist?
Yeah, I think that if we just sort of reflect on what's meant here by Journey, versus other types of business intelligence. A Journey specifically is when you, as someone who has a goal or a hypothesis about how you want customers to be interacting with your business, when you describe that as a set of milestones, and then you take a look at how are your customers actually progressing along those milestones. And when they're not doing what you want them to do or what you think they're doing, what are they doing? So that you can go in and make strategic decisions about how to change that customer experience and ultimately accomplish the objectives that you're looking for.
A Journey specifically is when you, as someone who has a goal or a hypothesis about how you want customers to be interacting with your business, when you describe that as a set of milestones, and then you take a look at how are your customers actually progressing along those milestones.
So some important components to what I just said, are the ability to describe these milestones, the ability to look at the patterns and aggregate, the ability to look at what happens when individuals diverge from those. And, previously, you could do these things, but you could kind of only do one of them at a time. So, if I wanted to look at who is achieving certain milestones, I could do that with a BI platform, but it would be building a chart and building a bar chart, breaking that bar chart down by customer segments. But I didn't have the ability to look at this over time. As a time series. I couldn't see who is hitting this milestone and then going to this milestone and then going to this milestone and see where the dropout was occurring. That was a really complex hard thing to do. And I had to do that personally. And I know that we had a whole in multi-month engagements, just around doing this type of analysis, is very manual, very labor intensive.
Now I could get a time series view, but I couldn't get that in aggregate. So I would call this today a customer timeline where I could see for this one customer wants their story, what did they do, but how do I look at a bunch of customer timelines in aggregate? I mean that's actually kind of a complex problem when you think about how you visualize something like that, when you have 10 million customers and every single one of them timelines is different? And so journey Analytics wasn't exactly solving a new problem. It wasn't, it's not exactly doing things that are completely unique. But it uses some sophisticated visualization, sophisticated underlying analytics to bring together these different techniques that were independent of one another and put them in the same place so that you can get that aggregate view of what customers are doing and where they're not doing, what you want them to do and what they are doing that you can fix.
Yeah, and I think one of your key points there is the aggregation of data into a single system that can analyze that data. And with speed. That was a huge benefit I know when we started using it at Comcast, it was the ability to very quickly understand, is something going on here? How many people are getting into this situation? Is this a problem? And to very quickly analyze the data from different perspectives. So, sounds like the challenges early on were there are ways of doing this but it would take weeks or months. And now it's moments, which is a huge benefit.
Yeah, that's right. And I think, you know, Journey analytics that the name didn't really exist 10 years ago, it kind of was made up around 2014-2015 is when, when why that sort of started to solidify just as the name of a practice. And I would say that journey analytics is more of a practice than a description of a certain product or technology. And a lot of companies are doing journey analytics. But we should remember it's not something that's like taught in business school. It's not like there's the journey analytics class. This, it's all sort of new. And so even though companies are doing journey analytics and analyzing customer behaviors, there aren't really standards for what to expect like how fast I should be able to do this, what's the best practice for doing journey analytics? So all of that as we're discussing this. We're still discovering and agreeing as an industry on what the best practices is and what the standards are for how quickly you should be able to solve these problems. And Pointillist is an example of a technology that can facilitate that journey analytics practice.
We should remember it's not something that's taught in business school. It's not like there's the journey analytics class. It's all sort of new.
Yeah, absolutely. So Pointillist, Genesys work with a lot of different companies. Are there different industries that are using journey analytics solutions today and what benefits are they seeing?
Yeah, I think the benefits go across industries. I mean, I'll just describe some of the use cases. So one use case, which came up early, in fact, is the use case where, that Pointillist emerged from, was digital containment. So looking at the reasons why customers are leaving the expected self-serve journey, and then coming into the contact center. And of course, that's largely a cost reduction journey as the goal. Keeping someone inside of digital containment, but more and more It's a customer satisfaction journey, as the expectations of customers become more around self-serve and self-install and self-guided experiences, especially in the post COVID world where we're all kind of expecting to be able to do things for ourselves in our own spaces. So digital containment is one and obviously that applies across most industries, where customers now expect there's an app or a digital experience for just about anything that they would want to be able to do.
Another really important use case that again spans industries is escalation management. And this would be when a customer calls in or emails which generally has a complaint about their experience. Now sometimes these can be simple, but sometimes they require some deep investigation into what exactly was happening with this customer. Where did things go wrong? And in our discussions with analysts, we found that they were having a lot of trouble and were spending a lot of time stitching together data sources from lots of different systems and different channels with disparate customer identifiers in order to put together these timelines, and then recognize patterns across those timelines and figure out if we have something going wrong for this one customer. Is it going wrong for five customers? Or is it going wrong for 100,000 customers? And with a tool like Pointillist, a journey analytics tool, you can just get both of those views the individual customer view, but then also those aggregate patterns very quickly. And again, that's across industries.
And then as I mentioned, any other sort of self-serve or self-guided journey is becoming increasingly important to customers and to the perceived quality of any business and their customer experience. So places where journey analytics can especially help there are new product management. So when a new product or new experience is being deployed, presumably the product manager has done some research and has good reason to believe that this is going to be the right experience. But it's only when it's out in the wild that you get to observe, are customers actually following the self-serve experience? Are they able to self-guide themselves? Are they able to install or activate or set up the account? Are they adopting this new feature? That's something that you only get when you put it out there in front of customers. And you don't want to wait one or two months to figure out if you did it right. You want to have it be right in as little time as possible. So getting that data into a journey analytics platform, being able to see those patterns as they come together and recognize opportunities, really decreasing the iteration cycle. That's a place that journey analytics can really add value in product management, and also in management likewise of bots and automating this distance where there's similar sort of story, the bot or automated assistant can't tell you hey, I'm having trouble answering customers questions in this area like a human agent, can. You have to get the intelligence to see that and then quickly iterate to improve that experience.
I think another area that was really exciting for us was an Omni channel view, and especially around a sales journey. Someone comes into the website, where do they end up purchasing? You know, and how long does it take and all the different moments across that journey? Do they call us are they emailing, you know, all of that coming together and being able to see that so quickly. So you know, huge, huge boon to the ability for product owner, experience owner to understand what's going on with their journey.
So you guys recently agreed to be acquired by Genesys. How did that come about and what's the plan for integrating Pointillist which is a cross-journey analytics tool into their contact center tools and the opportunity for joint value creation?
Well, Genesys and Pointillist were partners for a while before the acquisition, and that really made sense because, in just about every Pointillist deployment, there's contact center data. Contact Center is one of the key touch points or the sets of key touch points that customers encounter especially during the highest stakes moments in forming that customer experience. Genesys for instance is one common data integration for Pointillist, as well as the competitors of Genesys, Avaya, and others. Many of those use cases I think, probably at least three out of four of the use cases that I mentioned, touch contact center, at least in part. So for a long time, they've been highly compatible use cases. It's been data and integrations that Pointillist consistently was taking in. It's made sense to have a partnership and to be part of the Genesys app foundry. And last year, Genesys approached Pointillist and we discussed that at a more strategic level and decided that it made sense for Pointillist to become part of Genesys.
Genesys and Pointillist were partners for a while before the acquisition, and that really made sense because, in just about every Pointillist deployment, there's contact center data. Contact Center is one of the key touch points
And so what's the future look like for that integration? Is it to go deeper, I'm assuming, into support for the contact center, but what else is there?
I would say that we are focusing on integrating both at a content level so the journeys, the dashboards, etc, that are pre-built in Pointillist, having those be optimized or offering optimized content for the contact center. Also doing closer technical integrations to make it a truly plug-and-play experience. But the focus of Pointillist still remains on the broader customer experience and bringing that all together. And the reason why that is highly compatible with contact centers and with Genesys is that certainly a decade ago, five years ago, even contact center was viewed largely as a mission to reduce costs. And now, of course, has changed and COVID accelerated that contact center is becoming much more of a strategic opportunity to improve the customer experience. And in order to do that, managers within the contact center and customer experience managers need to be able to see across multiple channels and need to be able to quickly reduce the cycle of observing an opportunity in a single interaction or conversation with a customer to actually enacting a strategic change that changes the customer experience for a large number of customers and moves the needle on KPIs. So naturally for this to be able to happen the Journey Analytics has to not just live in its own kind of bubble for analysts. It has to come in closer and be part of operations be closer to contact center operations. And that's what the longer-term goal is. It's to bring those journey analytics insights from Pointillist closer and closer and eventually make a seamless relationship between the insights of Pointillist that are coming from across channels and the day-to-day management. And in the moment, decisions that are being made in the contact center to optimize the customer experience.
Yeah, I find a lot of disconnect between the call centers and the digital experiences for many companies. Agents have no idea where the customer has been, what they've been through already, they keep starting the conversation over again they transferred to somebody else, they start the conversation again. Having an understanding, really, like you said impacts that overall customer experience, the sense of trust that you have in the company, that they're valuing your time. So look forward to more integration of that.
So I want to talk a little bit about AI. Using artificial intelligence seems to be part of everybody's strategy at this point, and it's kind of thrown around quite a bit. But you guys saw an early opportunity I think for AI in Pointillist to help with the analysis of data. So what does it mean to you in terms of Journey analytics, and orchestration? And where do you see the biggest opportunities for AI in the coming years?
Yeah, early in Pointillist’s history, we were caught up in the wave of this belief that AI could solve and reveal larger problems that were more strategic in nature. We had customers telling us that they wanted the AI to just tell us what was wrong with the customer experience or just make the customer experience better on its own practically. And we didn't know how exactly to accomplish that, but we did make attempts at that. We had some successes in that direction. But ultimately, what we and what I think that the whole industry has learned is that machines often don't have the necessary context to be able to make decisions of that magnitude. And people also don't believe them or trust them. People usually once they see what an AI produces, probably have better ideas than the AI because they actually know what's going on with the business. And a concrete example, of a lot of hype and optimism and investment and technology and impressive technological accomplishments, kind of failing to meet expectations and imploding a little bit is Watson, IBM Watson where we're actually hearing stories of it being pulled back out of places where it was deployed with a lot of energy and optimism, because it was just trying to do too much. It was trying to meet the call, or something that could just tell you what the opportunity was. To some extent, it probably achieved that. But in the end, it doesn't seem like that's actually what people want. So, now our focus has shifted to using AI in a much more targeted way, which is to amplify what decision-makers have to do on a day-to-day basis rather than trying to make those decisions. The difference between you know, like a mech suit in Alien and an Android. AI can be a great mech suit. But AI is not really the Android executive that people hoped it would be. So in the contact center, there are some really concrete examples of that. There are decisions that contact center managers have to make in the moment about their agent allocations and staffing and other things, and routing of calls. And those are all places where AI can be deployed so that A. the amount of time that a manager needs to spend thinking about it can be reduced and B. the end results can be predicted ahead of time and then optimized. And then escalation management, for instance, is another place where AI can help figuring out what the root cause of a customer complaint is, isn't just a matter of pulling together data. There's a lot of thinking and observing and hypotheses that need to be tested. And that's a great place for AI to find those patterns and surface them. Not necessarily answer the question, but say, hey, in the context of these people that are having a similar problem, these are the kinds of things that seem to be anomalies for them. Maybe you as the human who actually knows what's going on in the business might want to look at these things instead of boiling the ocean. There's a lot of other similar sorts of examples that we're exploring at Genesys now. But in the end, they all come down to asking our customers, what are you doing today? And what is working for you? And finding ways to make it better. Instead of this sort of SciFi vision of creating brand new opportunities for AI, which sound really great. If you're trying to come up with some competitive differentiation, or if you're a data scientist looking for something cool to do, but in the end, if they don't map to what people are doing now and the challenges and pain that they're facing personally, in the contact center and customer experience and product management, then they're just solutions looking for problems.
Yeah, I think the areas that I've been most interested in around that is sifting through large sets of data to bring out ideas that can be assessed that not necessarily giving the AI the ability to make the decision and to move forward but to bring options, bring information based on this particular customer or sets of customers that can help a human make better decisions. Like you said, maybe sift through certain options beforehand and bring those up.
I think an important learning for us and AI is ultimately that people buy solutions to their problems and not technologies. And something funny that we just realized is that a lot of these industry analysts are still analyzing you know, the hype cycle, for instance, is still based on technologies as opposed to solution. So even industry analysts can be a little behind the times on this still focusing on 'deep learning' and these these buzzwordy techniques, but where we are now is finding that those are just math. No one goes out and sells linear algebra. They sell a system to up to optimize operations. It doesn't matter what the math is underneath that. And the same is going for AI where the term AI is being thrown around. I think that people cared because they wanted to have an AI strategy of their own. But now we're all just sort of getting used to it. And the focus is really pivoting toward solutions as opposed to the technologies
An important learning for us in AI is ultimately that people buy solutions to their problems and not technologies.
Will, I really appreciate you taking the time out of your busy day to come and talk with me and I wish you all the best in your new role and looking forward to seeing how Genesys and Pointillist and just all of Journey analytics grows in the coming years. So thank you.
My pleasure. Thank you, Mark.