More on eMetrics

[I’m going to be down in San Diego this week teaching DMU classes and going to the VS User Forum – if you’re there be sure to come by and say hello! BTW – thanks to the several people who commented favorably on the Internal Search Series! I wasn’t sure if that had any resonance. I have at least one or two more posts on that topic but I wanted to get the timely stuff out of the way first.]

I’ve been asked several times what were my key take aways from eMetrics? It’s a tough question. Because I work the booth and meet with people, I don’t get to see a lot of the speakers. So I saw more speakers on the last day than on the first two days combined. Also, I don’t generally go into any Conference with high-expectations of learning much.

Of the presentations I did get a chance to see (and the conversations surrounding them), there were a couple that really made me think. The first was Bryan Eisenberg’s presentation on Persuasion Architecture. I had several extended talks with Bryan during the show and they, no doubt, helped to crystallize my thinking on the subject.

Bryan is a good presenter. Almost too good. I think people take away more of the style than the substance of what he’s saying. And that’s a mistake because what’s he’s saying is deeply interesting.

Surrounding that presentation, I talked with quite a few people I know about how Persuasion Architecture and Functionalism might fit together. I think there is a close fit – and maybe even a nice symbioses.

Functionalism is a measurement methodology. Pure and simple. In as much as it provides guidance to anyone who isn’t an analyst, that guidance is primarily about how the success of a page is going to be measured. And in thinking about that, it can certainly help "crisp-up" a designer or marketers thinking about what the page is supposed to accomplish.

That being said, Functionalism says nothing about how to structure the creative messaging on the page or what alternatives are worth testing. With the advent of serious A/B and multivariate testing, that's a major unaddressed problem. I thought Bryan’s comments on the 10,000 monkey approach to multivariate testing were especially apropos. You can test lots and lots of uninteresting variations. But that won’t help you find a really good creative approach.

The Persuasion Architecture provides a method that seems like it might. Like any methodology, it’s not going to churn out guaranteed answers. But it should focus a marketing and design team on possible messages and variations that might make a big difference.

In much the same way, Functionalism can serve to focus an analyst on the measurements that are most important for any given type of function. I’ve often said that I think methodologies are most useful for new and less experienced practitioners. That’s true of Functionalism and probably of the Persuasion Architecture as well. You don’t need to provide a system for a marketing genius. He or she can just do it. Same with measurement.

Genius, by definition, is scarce. That’s why methods make sense. They make all the rest of us much better. And however good I may be at web analytics, I know I’m no marketing genius. That’s why I’m anxious, when I get some time, to try some of Bryan’s concepts on our own web site.

One thing that’s probably obvious to anyone really thinking about the Persuasion Architecture is that it’s going to be a lot of work to implement thoroughly. That’s not a negative in my book. It seems to be an approach (again, like Functionalism) that could very effectively be implemented in pieces.

You could take any significant page on your site and apply parts of the Persuasion thinking to help you define creative approaches. The more pages you touch, the greater your control of the navigational structure and messaging is likely to be. But in principle you could get good value even within the context of a single page.

Nor do I expect a good methodology to provide answers. The real-world just isn’t like that – especially when it comes to open-ended issues like the best approach to design and creative for selling.

A lot of the page types that Bryan talks about in the Persuasion Architecture (information way points, re-assurance modules, etc.) have close or exact corollaries in Functionalism. So there isn’t any big divide in the way the two carve up a web site.

I’d love to see the two approaches seriously combined – with Functionalism providing a measurement overlay and the Persuasion Architecture providing a creative overlay on the same site. It seems to me that combining the two would be deeply interesting and truly state of the art in the world of web analytics and marketing!

I have two more presentations I want to mention, but in the interest of "shorter" posts, I’m going to stop here and address the next ones in a day or two.