An old colleague of mine, Mike Peralta, recently joined a newly launched start-up called Magnetic, which you can read more about here. In short, the company provides DSPs and their kind with search behavior data to use in display retargeting campaigns. I told Mike, I find it odd that this is the first time I’m hearing of a provider who does this. Especially since Yahoo! based their entire behavioral targeting business on this principle years ago. But here we are, with another way to retarget valuable customers, and I think its a really good one for a few reasons (not just because Yahoo already did it, which isnt always a good reason.)
– It will get more display marketers thinking about the interplay between display and search.
– Unlike offline data, search data is easy to update in real time or near real time.
– There is a lot of it, so a lot of people can play and experiment to find what works.
So while I have publicly cautioned against the “more is more” pile-on effect we are seeing in the data space right now, I do think search data has a justified place at the table.

Of course, as I mentioned in a recent post, there are risks with this kind of targeting. Among them is one I may not have mentioned recently, that marketers must be careful not to be “too good” at targeting messages to key audience segments for fear that they look like they are stalking customers. I actually had a conversation on stage at the x+1 Next Targeting summit where Eric Roza of DataLogix said just that; the data we can now access as interactive marketers is so good that we now have to make sure to tone down the customization of the creatives or risk privacy backlash. I hope this will not always be the case. I hope instead we create a new quid pro quo with consumers that we have to show them ads, in which case it benefits them if we provide custom messaging when and where they want us to. But in the mean time, I think a marketer’s best option is to test data targeting first to find factors that generate lift, and test targeted creative second, cautiously.