What Behavior Are You Targeting?

“Data Driven Thinking” is written by members of the media community and containing fresh ideas on the digital revolution in media.

Today’s column is written by Pete Sheinbaum, CEO of LinkSmart, which provides text-linking optimization solutions for web publishers.

It happened to me a year or so ago, but it happens to millions of people every day. I was in the market for some new snow tires and did what most do. I asked ‘The Google.’

Without leaving my living room, I was greeted with several good-looking choices on the best snow tires for my vehicle and the climate I live in here in Boulder, Colorado. I clicked on the (second) search result and landed on a great review. But something was strange… the page I was reading tire reviews had been invaded with women’s luxury shopping ads.

Glancing up and to the right of the web page, there was a big, beautiful ad for…Gilt. Strange, I thought, but went back to reading the review. About half way down the page, I chose to click on a link to a more specific review on Continental tires that the author had written. I clicked, landed on that page, and that strange thing happened again — another big ad for Ideeli this time. Why was I getting ads for Gilt and Ideeli on a snow tire site?

Three reasons: Wife’s computer + AdSense + Cookies = Poor Targeting

Behavioral or psychographic/demographic targeting has been around for a long time now, but they are saddled with one fatal flaw: the one-person one-computer/device construct is no longer the norm. We have multiple computers, phones, tablets, etc. that we access every day and every single one is collecting data.

While those devices can create a decent profile of the primary user, we no longer use them in a singular way. Think about how you’ve interacted with technology devices today, we are moving from one device to the next more frequently than ever before.  Also in the case where there is one ‘common’ machine, you have ‘cookie confusion,’ a tangled web of disparate user sessions that generally perplexes the Skynets that are monitoring us. Moreover, if they do get it right about who is using the computer at any given time, they are serving up ads that could be relevant for what I was doing hours, days, even weeks ago. Stale is an understatement.

While the aforementioned may be only applicable to me, I’ve had it with this kind of targeting.  If a person is reading an article about choosing a golf club, show them golf ads. If you’re reading an article about how to make an apple pie, I think you should be shown food ads. But please don’t show me ads about designer shopping when I’m looking for snow tires.

What’s the fix? Targeting users based on intent. Readers and their clicking habits present web sites with a great deal of information that is not stored in the cookie. What page are they on? How did they get there? What links are they clicking on to navigate from page to page? This information can tell you a great deal about what a user’s intent is, and smart systems will pair relevant ads with this knowledge.

Additionally, advertisers need to acknowledge that users’ context and intent can switch many times during a single session. I start on food, then I switch to sports, then golf club reviews, and finally autos. It doesn’t necessarily matter where I was an hour ago, or a day. It doesn’t necessarily matter how old, young, rich, not rich, I am (all of which have been culled from thousands of user sessions across hundreds of sites, and the cookies that are being planted on me for re-targeting). Today systems should be more sophisticated, focusing on the here and now, not the there and then. Ad-serving technologies must leverage what signals readers are offering them in real time and embrace the idea of Real-Time intent.

While any degree of targeting may still be better than no targeting at all, ad buyers should beware of serving ads to husbands surfing their wives computers.

Follow Pete Sheinbaum (@sheinbaum), LinkSmart (@linksmart) and AdExchanger (@adexchanger) on Twitter.