At the dawn of the digital age, there seemed to be an emerging promise that every action of every consumer would now be finally be measurable. We could finally retire that century old quip from John Wanamaker about how he knew that 50% of the money he spent on advertising was wasted; he just didn’t know what half. With the help of web analytics, John or at least his great- great grandson could now identify and dispense with the excess marketing dollars. True Marketing ROI had finally arrived.
Well, we are now a couple decades into the digital era and while everything that happens online is technically measurable, it turns out that this data doesn't explain everything as thoroughly as we hoped. It might tell us what people do on a given site, but it's just about weak as every other research method as explaining the complex purchase pathways people go down these days: using multiple sites, apps, media vehicles, kiosks, phones and word of mouth on their way to making or not making a purchase.
I’ve started exploring the limitations of existing research methods to deliver insights in this new multi-platform world. It appears that we have a bunch of way of finding insights around discreet experiences but not much that will help us understand the connections between those experiences.
Most of the research tools we have fall into a few broad categories
1) Concept tests or ways of measuring how people respond to different ideas and statements in a qual or quant way
2) Communication tests: ways measuring the appeal and impact of marketing like objects, from claimed behavior to fancy Clockwork-Orange like biometric tools.
3) Usability tests: ways of measuring how well something works or easy it it is use
4) Behavioral data: which isn’t claimed behavior but retail and site analytics which will measure an action taken, a click, time spent, or a purchase made.
A lot of us involved in helping make marketing stuff for this multi-platform/multi-channel world want to know not just what happens in these discreet moments but what happens in between them so we can strengthen the connections, driving a person from one to another. It’s one of the reasons we spend a lot of time drawing what we now call “consumer pathways” or “Journeys” or “experience maps” that conceptually visualize the path from one stimulus to another.
I think too this is one of the reasons we’ve seen a resurgence in the use ethnography as a market research tool. While market research ethnographies have of course never really gone away, they fell into some degree of disrepute back during the rise of web and social analytics. Why follow people around making notes when we can measure what we they doing every second, crunch the numbers and turn them into big charts and graphs? The flaws in this logic are pretty obvious (sample bias for one), but it's pretty hard to over-estimate the power of statistical analysis to American business, no matter how dubious the methodology. Any number is better than no numbers.
Ethnographies are far from perfect, especially the down-and-dirty variety we generally do in market research environments, but I've generally found them pretty useful for reasons I'll explain in the next post.