Since the late 1990′s we’ve been promised an amazing jump in advertising effectiveness from targeting only possible online. Initially, it was the claim to “know who the individuals are”. Today, it’s become “we track them and know everything about them”.

Truth is, online targeting hasn’t delivered, but the idea remains prominent. Part of the reason was summarized today in an excellent post by Byron Sharp discussing a recent HBR “advertorial” for McKinsey. (Link here.) Sharp criticized them, in part, because “there is an in-built assumption that hitting someone at the moment when they are thinking about the brand/category is the only advertising that works.”

Seems like such a smart idea. Except, let’s add reality. Any salesman knows that if we wait for the “best point” to ask for the order, we won’t sell much. At the same time if we always ask for the order at the worst time, we will never make money. In a way, traditional media planning walks this middle line focusing a reasonable amount, but casting a wide enough net to get the surprise purchasers.

The Digerati Code. But the Digerati ignore this reality and wholly embrace the assumption, leading to the following hypothesis:

Given a specific product or brand…

1. We CAN use digital tracking to determine points in time when consumers are most receptive to receiving ads about that product or brand.
2. Once we detect when they are most receptive, we CAN present ads that hit them at that most receptive time.
3. When ads are presented in this way their effectiveness is dramatically higher than using traditional methods.

If any of these three are wrong, we’ll never locate the Holy Graal of digital marketing.

What Does Experience Show? There has been no general increase in marketing dollar effectiveness as a result of digital efforts. And digital response rates (CTR’s) are pathetic while online CPM’s remain a fraction of what traditional media is able to demand.

There seem to be two realities:

There are times when online targeting pays out exceptionally – usually for highly targeted direct response campaigns. My guess is that constant low cost testing helps them figure out, of the target criteria available, which ones are connected to response. But brand campaigns don’t have this luxury.

So broader brand efforts haven’t seen a dramatic increase in ad dollar effectiveness. In fact, many campaigns curtail their online spending because there are simply too few places to spend money that achieve acceptable results. (For example, Pepsi and Best Buy have both recently redirected media funds away from digital and back to traditional. Apple has always succeeded with heavy traditional spending and minimal digital spending.)

Is Targeting Fundamentally Flawed? My sense today is that there’s a big problem matching market to online variables in a way that generates impact. Because online variable must be limited.

Assume Criteria List A tells us what makes a good target consumer. Unfortunately, if you have defined your audience well, nobody’s tracking what’s important in your list.

Because, in truth, Criteria List B is what is actually tracked.

But hoping to cross this gap, some digerati take “B” and expand it to Criteria List C using database correlation.

SO, this leads to three questions:
1. Can we effectively reach our audience “A” when we only have “B”? Not really.
2. How accurate are the projections in “C”? Probably not very accurate.
3. So can we reach “A” when we add the projections “C”? Probably not.

So, it turns out that online buying is pretty much just like buying TV with Nielsen rating data – just with some additional criteria. But “more” criteria doesn’t mean “the important” criteria. What we’d really like to know is predisposition to absorb some of what we say and have that cause action in the future.

There’s nothing wrong with this situation – TV has succeeded for years with Nielsen’s. But what happened to the consumer targeting nirvana we were promised?

Watch Out for VC Hype. VC’s know this theory is a clever way to try to increase value for their ventures. Also, this theory is one of the few really unique things the digerati can claim.

Even worse, ad agencies are suckers for this theory – probably due to a common agency disease I’ll call “Nielsen frustration”. (This is the tendency to blame weaknesses in Nielsen audience measurements for the fact your ads didn’t work. “If only the RIGHT consumers had seen our rats dumpster diving for sandwiches people would have bought the sandwiches in droves…”.)

That’s all nice. But online behavioral targeting hasn’t shown huge results in practice. Even worse, the theory has hung around for a decade without dramatically changing results. How much longer should we wait?

Until proof arrives, it seems best to conclude that online targeting is nice, but no magic pill.

Copyright 2011 – Doug Garnett – All Rights Reserved