…is that it's not quite good enough. We've all had it. Go to a retailer's site, have a bit of a browse not really intending to buy anything, and then you're followed around the web for the next month by said retailer with (usually) not very enticing banner ads that sometimes contain the products you looked at all those weeks ago on the retailer's site. More recently, my own personal corporate stalkers have been a trainer company and an online retailer.
Such retargeting networks are dependent on tracking technology to identify users that have been defined by their browsing behaviour from large 'cookie pools'. In my case though, what the technology doesn't know is that the only reason I went to the retailers site was in the course of doing some research for a project I was working on. What the technology also doesn't know is that after visiting the trainer company's website, I subsequently bought a pair of trainers elsewhere and am now a satisfied customer who doesn't need another pair of trainers right now.
That's the thing about shiny new technology – we get all excited about the possibilities and then the early execution has a lot of holes in it. That's kind of OK, as long as the technology developes rapidly. The problem in this case with being not quite good enough is that it's been not quite good enough for quite some time and can actually prove quite annoying. I was once followed round the web by another online retailer for six months after I visited their site. I realise that it's dependent on the limitations of the data but I can't help but feel that if more care were taken around the execution (more use of frequency capping, more sophisticated application of the data to close the loop better with subsequent user behavoiur to redefine the context, for example) it would be better for everybody. It's about time we made it a bit better I think.