I’m just returning from Philadelphia where, in addition to the usual frenetic round of meetings and lunches, I had the pleasure of attending the WAA’s Philadelphia Symposium on Customer Analytics. I got see almost all of the half-day event, having only to duck out for one call during the proceedings. I thought David McBride did a super job putting together content. Every presentation was substantive and enjoyable. Not a single presenter abused the audience with a sales-pitch or thinly-disguised vehicle for self-promotion. What a pleasure. And you have got to love the Ritz – a beautiful room indeed!
In addition to the usual suspects (of which, I suppose, I am one), there were three presentations that I found particularly interesting. Elea Feit reprised the (really, really good) presentation she did at X Change, and while I’d seen it before it has so much content that I still took something new away. I realized that one of the Wharton studies on Social Media was extremely relevant to my recent blogs on Social Media monitoring. I also very much enjoyed Ryan Caplan, ColdLight’s CEO. He did an unconventional, rather conversational presentation that, from an unexpected direction proved to be a meditation on the changing role of the analyst with the advent of machine-learning systems. It was entertaining and excellent, though I disagreed with one aspect of it and I intend to post on that as well since I think it’s interesting and important. My favorite presentation though (probably because I’d already seen Elea’s), was probably from Comcast Interactive’s CFO, Daniel Spinosa.
Yes, it’s certainly nice to get a speaker at that level at a Web Analytics event and it’s always more interesting to get an outside perspective since I see so much of the inside perspective. Yet this presentation considerably exceeded my expectations. What struck me about it was how exciting and big the data opportunity is at Comcast. And by big data opportunity, I don’t mean big-data at least not in the sense of what’s interesting.
Yes, Comcast has loads of data – and not just Web data. The numbers presented were probably conservative in comparing the amount of set-top data collected in Web-like terms (and it was still enormous). But lots of data isn’t inherently interesting and, in a meaningful sense, isn’t really unusual either. Organizations like Facebook and Comcast may each be alone in their own particular big-data universe, but at some point the problems around big-data merge. Or so I suspect. If you can solve CNN’s or eBay’s data problems, you can probably solve Comcast’s or Facebook’s.
What was far more interesting, I think, is how intriguing and potentially important the data available to Comcast really is and the fact that these analytic opportunities take several distinct forms. That multiplicity of forms is important because it suggests that if Comcast Interactive has multiple data analysis opportunities important enough to be interesting to the CFO, other enterprises might have some too. A single problem might well be quite unique (and, in fact, one of their problems is pretty unique).
His presentation was entitled “Is There Anything Good On? Customer Value Through Analytics” and while only a part of the presentation really pertained (at least directly) to the first part of the title, I found it the most thought-provoking portion. With the explosive growth in broadcast content, Comcast’s XFINITY delivers an extremely rich experience. Alas, if you are a Comcast subscriber (like me), you know that the set-top interface is pretty horrible. I like almost everything about Comcast better than DirecTV except the set-top interface (okay…and the football package)! With lots and lots of content and a poor interface, finding out what’s on and good is, indeed, challenging.
Improving that interface embodies a classic UI design problem as well as an extraordinarily rich opportunity for personalization to support navigation and consumption. The ability to guide people seamlessly to the content they like is a huge and not fully explored advantage in the industry. Yet the analytics techniques for solving these problems and driving this personalization are well-understood and can, I suspect, be ported effectively from the Web.
I’d bet my house, for example, that our Use-Case Analysis techniques would yield rich fruit in this set-top soil (though I’m not so sure about Functionalism). And it almost goes without saying that traditional Market Basket analysis would drive outstanding personalization recommendations. There are obviously significant monetization opportunities in this type of optimization, but the bigger advantage is around customer acquisition and, particularly, retention. Giving up my DirecTV interface was a wrench, but that interface, while cleaner, still lacked ANY substantive personalization. Provide a really rich and personalized content consumption experience, and you’ll almost certainly boost retention, consumption and monetization. As George Costanza on Seinfeld would say, “Trifecta!”
Which leads into the second large data opportunity – one that really resonates with the message I’ve been discussing for most of this year. At Comcast, the big customer data opportunity is around retention and attrition. Moving to a customer-level view and analytics based on customer lifetime value ran through every aspect of Daniel’s plan and presentation. That customer lifetime value view of the world is a huge change for most organizations; it’s analytically demanding, data intensive, and intellectually challenging. It’s so much harder, for example, to balance the long-term impact on retention with a short-term monetization opportunity around display than to simply “optimize” your revenue, that the two tasks can hardly be compared.
A huge and potentially game changing UI and personalization opportunity that could leapfrog the customer experience in the industry; a desire to measure every aspect of marketing and operations in terms of its long-term, true impact on the customer and company. Those are big ideas. Important enough to be interesting at ANY level in the enterprise and deep enough to excite even the most jaded analyst. Good stuff.
I can’t close without a word on my “noveau” panel. I think it went pretty well, though being on the wrong side of the podium it’s really not for me to judge.
I will say that it was hard work.
I put more work into the preparation than I ever do when I sit on a panel, so I’m not sure how keen I am to repeat the experience. On the other hand, I did enjoy the actual doing of it. It’s nice to be asking questions instead of answering them. I really like listening to folks and, here’s the big bonus, being able to ask whatever questions I want! Thanks to all the panel who endured my relatively unscripted format, the unplanned follow-ups, and the “omigod – what question is he asking me now” experience with grace, verve, good humor, and, in my opinion, considerable intelligence!
I’m still not sure I’m a fan of the format. In reflecting on it, I think it might be better to try a one or two person panel/interview. That’s me, though, always tinkering…
