Best Not to Creep ’em Out

1 Dec

Big-brother-is-watching(Update: As this was being posted the FTC released a proposed Online Privacy Plan that would require consumers to opt-in to all online tracking. You can download the proposal here.)

After looking at pair of shoes on Zappos, you get an ad for those same shoes served across four sites. Is Zappos stalking you? A financial services company ad seems to know that you’re nearing retirement; does it set off alarm bells?

Are you creeping out your customers? Before you scoff at the idea, think about it. Are you retargeting with your ads? Using localization? Modeling? If so, it’s quite possible you are setting off privacy concerns.

Sure, your intentions are good. We all understand that serving more relevant ads is in everyone’s best interest. Users get offers that matter to them. Marketers get better results. But in our zeal to make it better, we really may be screwing things up.

That was a topic of great debate at a recent IAB Innovators Roundtable dinner here in Atlanta. It was a great discussion that involved a number of us agency types and key players from companies like ING, Coke and The Home Depot.

No one doubts that making ads more relevant is critical to creating better online advertising. But if consumers don’t understand how we’re doing that it’s quite possible we unleash a firestorm of privacy concerns. One marketer even worried that it could touch off a wave of regulation from politicians who think they’re protecting consumers.

Maybe we need to be more transparent. Ads could be more like Amazon or Netflix. “Because you liked xxx we thought you’d be interested in yyy.” Or perhaps the industry needs a better PR effort explaining to journalists just what’s going on and how it’s no different than decades of accepted direct marketing practices.

Or, maybe we just need to ask ourselves “why” a bit more often. Seriously, I think that one little word can make a big difference.

Yes, you could segment your ads based on site demographics. You could also cross that with a predictive model based on years of purchase data and reinforce that with  retargeted offers.

With all the data available to marketers these days, pretty much any type of segmentation you can imagine can be done. But why?  Is it actually going to provide enough lift that it is worth the effort?


Comments are closed.