When my company's co-founder, Michael Provenzano, helped start the first DSP in 2007, the ability to bid on an individual user in real time was a novel idea.  Today, however, anyone who has ever shopped online has been stalked by the items in his or her abandoned shopping cart.  The pair of shoes from Nordstrom that was almost purchased.  The book from Amazon that was marked “buy later.”  Retargeting, as this tactic is called, became the bread and butter of programmatic buying. 

Whatever the cause, consumers sometimes hover over the buy button without actually finishing the transaction.  Facilitated by the programmatic ecosystem, an ad for those shoes will appear on Facebook, and information about the book will show up as a banner while the user is reading The New York Times. 

Abandoned-cart retargeting is one of the most basic types of retargeting, and it’s very simple to execute as long as marketers only care about desktop users.  But as mobile shopping rises in popularity, abandoned-cart retargeting is no longer as easy, nor as practical, as it was when consumers were limiting their online shopping to their desktop computers. 

On a practical level, this is a cookie issue: Most phones do not permit cookies -- so it’s not as easy to follow a consumer across mobile web or apps -- but it also speaks to the nature of the platform.  While consumers who are shopping across websites are often sitting at their desks, consumers who are shopping on their mobile phones are usually on the go or in actual physical stores.

It’s not that we want to forget about desktop retargeting; it’s that we want to apply it in a better way for mobile purposes.  Acknowledging the way that consumers use mobile means acknowledging the fact that mobile gives physical context to consumers.  Because geographic information is part of so many ad requests now, location-based retargeting can be done more intelligently.  The consumer who didn’t finish buying those shoes might be less likely to complete her purchase if she’s in a business district (presumably for work) than if she’s at home.

This sort of location-based retargeting, however, requires a lot more effort (and data) than traditional methods, which is where programmatic again comes into play.  Whether the data being used is proprietary first-party data, or whether it’s third-party data that can be enriched through in-house analytics, to achieve this form of nuanced retargeting requires far too much effort to collect, analyze, and deploy through manual ad buys.  Programmatic systems glean insights based on consumer movement and behavior, initiating mobile inventory purchases that would never have been imagined previously.

If done well, consumers will be reminded of items  in a way that they’ll appreciate, and at a time and place that feels relevant.  To have maximum impact, that cart doesn’t need to follow shoppers all around the internet. It just needs to appear at the right place and the right time, in the real world.