The product thing: Justin Anovick and David Bowen
In his address to Ascend 2018, Episerver CEO Mark Duffel highlighted the success of what he called “an experience-driven approach to the customer,” citing a 150 percent growth in North American cloud subscriptions in 2017, outstripping 90 percent YOY growth worldwide.
Of course, broadbrush commentary about the importance of competing on CX, while it’s still needed (skeptics abound), doesn’t get into granular detail about how a cloud-platform like Episerver is promoting that aim. Okay, it drives online commerce solutions (websites), and empowers content management for marketers. That’s nothing new. I sat down with Justin Anovick and David Bowen to dig deeper.
Anovick, who had ditched last year’s cloud pants in favor of a shocking scarlet tuxedo, is VP worldwide product. I asked him to sketch recent developments. “At the lowest level, we’re tracking data: That’s essentially what we call the profile store. It’s been in development for a couple of years. On top of that is Episerver Insight, which is the view into that data, so that you can segment, show the customer journey, and the content details. We’re surfacing that to make it understandable to the marketer.
“The analytics suite, which includes the dashboards, is really the umbrella over those pieces. That whole thing launched in December and January, so it’s pretty new.” Previously, Episerver customers had tended to lean heavily on Google Analytics. “But there’s a lot of things we can do in tracking the individual which Google Analytics can’t do, so it’s augmenting Google Analytics.”
Importantly, added Bowen (head of product management), data from Google Analytics is anonymous, and can’t form the basis for personalizing the customer journey at an individual level. “We can track the customer; there’s an API so clients can add more data from touch points in the system. It gives them a richer analytics capability and a truer view of customer lifetime value.” The additional data can include transactional data, call center logs, CRM; anything that’s available.
“We’ve had a product recommendation piece for a while,” said Bowen, based on shopping behavior. “Insight allows you to see those product recommendation algorithms, but add other data.” Anovick explains why this creates much broader opportunities. “Product recommendations are way easier than content recommendations. There’s an explicit behavior: Somebody buys it. You can relate that to what else they buy, and the journey they took to get to that point. The content side is way more implicit. Is it based on someone being on a page for 29.6 seconds? Maybe not. If your KPI is how much time a user spends on a site, that’s not really indicative of them being happy, or finding what they need. It’s taken us a lot longer to launch the content part [Episerver Advance], because it’s way more complex.”
Content analytics, to be actionable, must happen fast? “It’s in real-time,” said Bowen. “You get intent signals from where they land. Where they go in the early click path directs the machine to present the next content or product sequence.” The machine models are automatically re-built and re-tuned in response to user behavior (yes, AI). “We do have a rules-based element,” said Anovick, “which allows a marketer to input ‘if-then’ statements, but we also have this black box of machine learning. What we’re working on is converging the two. With Microsoft, we’re working on displaying the decisions the machine made, so that the individual can change those decision-points.”
But is that a simple thing to do? “We’re keeping it simple enough,” Anovick affirmed, “that the user doesn’t have to know anything except their business.”