Commentary

The Last Two Feet

Last week the New York Times ran an article in which the headline posed the question, "How Many Hits? It Depends Who's Counting." The author began the article by wondering how many people had visited Style.com last month, citing several disparate estimates.


But hits, as readers of this column probably know all too well, are not people.

My colleagues in the Web Analytics part of online metrics will tell you that Web site hits are things that can be empirically known -- counted, trended, analyzed, correlated. That's because hits are things that machines do to other machines, and as we all know, the Internet is the most measurable medium, thanks in no small part to all those handy machines and their digital trails.

Technology has truly revolutionized the science of marketing information over the past 15 years. We have data warehousing, database integration, set-top-box data, click-stream data, transactional data, behavioral targeting. Yet sometimes it seems as if the more empirical behavioral data we amass, the less we actually seem to know about people.

People, unlike machines and much to our collective chagrin, remain as difficult to measure as ever. People -- whether as readers, viewers, site visitors, or consumers --remain the flaw in the system, the ghost, if you will, in the machine. People watch, listen, chat, text, IM, read, browse, chat, transact. Not insignificantly, they delete cookies . They become aware, engage, develop purchase intent, buy stuff, talk about products, generate their own media. They multitask. They block pop-ups, fast-forward through commercials, and forget your ad the minute they see it. They've even been known to talk on the phone, discipline their kids, and just get up and walk away from the computer for no good reason -- all without the courtesy of closing the browser window through which they have accessed your site.

But let's get back to online metrics.

Yes, the Internet is the Most Measurable Medium . Digital technology enables us to empirically know pretty much everything that happens online. We know, for example, how many times an ad is served; we know how many clicks this serving generated. We even know how many subsequent online sales accrued as a direct result of those clicks.

If we work back from the point of that sale, there is really only one thing we still have trouble understanding: the last two feet of the funnel. These are the two feet that separate the user from the screen. It is as if we online metrics practitioners can track everything that goes on in here, but we find ourselves huddled together against the screen from the inside, rapping on the glass, wondering "Hey, what's that guy out there doing?"

As measurable as everything is that happens on this side of the glass, that guy out there is harder to measure than ever. He is less likely to sit through your survey. His media behavior is more cluttered and fragmented than ever before, his time more precious. And now, as he distributes less time across a continuously greater array of media options, you want to know how "engaged" he is with each.

I would argue that the fundamental difference between Web Analytics and Audience Measurement (the two overlapping halves of Online Metrics) is that Web Analytics is about making optimization decisions based on machine-generated data, whereas Audience Measurement is very much about the last two feet. So when the New York Times asks "How Many Hits?" it doesn't so much depend on who is counting, as it does on what it is you want counted. If you want to know how many "unique cookied browsers" hit the site, well then, that's machines talking to machines, and Web Analytics will provide you with a pretty good answer. But if you want to know how many different people visited the site -- and maybe their demographics, and their distribution by country (or, let's go the other way, local market), and what other sites they visit -- then you need to worry about that last two feet. That's what Audience Measurement is for.

I think that digital media are by nature so darned measurable that we forget the distinction between the medium itself, and the person accruing the exposure. We're going to see this same phenomenon play out in TV as set-top-box data elbows its way into the ratings puzzle; and we're going to see it in Mobile, the next great (and all digital) media frontier. We're going to know more than ever before about the medium itself. But the soft, squishy people using the machines -- to truly know them, we're going to have to traverse that last two feet.
Next story loading loading..