4 Ways Location Data Can Change Campaigns

By August 25, 2010Expert Opinions

Could the places you go reveal your partisan preferences?  We’ll likely soon know, thanks to the explosion of location based services (LBS) like Foursquare and Gowalla, and especially now that Facebook has gotten into the location game.

Digital strategists are excited about location services’ potential as another connective tool in the social media ecosystem – a way to humanize candidates, create an intimate connection with supporters, and distribute their message in unique ways to voters.

But with recent news that Foursquare is in talks with search engines to license their data, it’s time to consider the ways in which rich location data could add strategic value to candidates and campaigns.

  • Opportunity 1 – Customized mobile ad-targeting:  From a consumer products perspective this application is fairly straightforward: if someone is online and they’re walking past your store, offer them an incentive to stop in and spend their money.  But in a campaign context, customized mobile ad-targeting isn’t as obvious.  If an individual’s location data can be linked back to other online profiles it would be possible to identify locations that are more heavily frequented by people with distinct partisan or issue preferences.  For example, most any young person living in Washington DC can tell you that Smith Point is a Republican bar and that Palace of Wonders is a Democratic bar.  But what watering holes might swing voters in Columbus, Ohio frequent?  Do you think sushi eaters in St. Paul have pro-environment preferences?  I don’t know, but imagine the power of being able to deliver a customized political appeal to online mobile users targeted to the known political inclinations of a destination’s patrons.
  • Opportunity 2 – More accurate microtargeting models:  We already know that consumer data (where a person shops) and geographic data (where a person lives) have high predictive value when it comes to establishing a person’s political preferences –  it only follows that where a person travels in their day to day lives would also be predictive.  Consider the data exhaust generated each time someone checks in with an LBS: location, time of location, frequency of that location, and unstructured but very rich text data (“shouts” in the Foursquare idiom); not to mention all the linked “social-graph” data associated with users that have checked in to that location.  Surely this rich raw data has some predictive power in establishing a person’s partisanship or issue preferences.
  • Opportunity 3 – Smarter yard sign strategies:  As much as I hate to say it, location based data might actually finally bring some value to the oldest, cheapest and (until LBS) most worthless form of political advertising: the lowly yard sign.  Yes, that’s right, the bane of campaign staff everywhere could get smart with the infusion of location data.  Consider it a low-tech analog version of customized mobile ad-targeting.  Theoretically, if location preference were attached to political behavior, the placement of yard signs and large road signs could be optimized for maximum exposure to critical electoral groups.
  • Opportunity 4 – Real-time event outreach to influential supporters:  Your campaign is holding a rally, meet and greet, or other large event where it’s non-obvious who the candidate and his or her surrogates should be spending their time with.  Encourage people to check-in to the event (a feature recently enabled in Gowalla), and perform some quick social network analysis on the group to identify attendees with a high degree centrality, betweeness centrality or network reach as high-value time and attention targets for the day’s event.

To be clear, these are not things that can be done today.  For now, there are three primary obstacles to doing these types of things effectively:

  • Obstacle 1 – Data availability:  Most significantly, all of this depends upon making raw location data available to third-parties.  As of now, there are no official plans to provide access to the data, though reportedly Foursquare is in talks to license their data to search engines, and the rollout of Facebook Places saw this revealing line in a recent Wall Street Journal article: “Facebook says it isn’t monetizing the service, at least not at first, but may consider ways for companies to make use of the data ‘down the line.’”
  • Obstacle 2 – Linked data:  Each of the above is ultimately dependent upon linking LBS accounts with other data that is verifiable back to known individuals, even if anonymized.  Location data alone will tell us little.  Instead, its real power comes from having at least some sense of the person behind those check-ins: are they a Republican or Democrat? What issues do they tweet or blog about? How many friends, followers or fans do they have?
  • Obstacle 3 – Adoption and usage rates:  Finally, all of this may very well be a moot point until takeup rates increase for location based services.  Currently only about 4% of Americans use LBS, though with the release of Facebook Places that number is set to grow exponentially.  But the truth is, until there is widespread adoption and an appropriately sizeable data footprint, location data likely won’t be heavy enough to have a significant impact for prediction or segmentation.

As you can see, right now location based services remain at the bleeding edge of political tech.  But ultimately, these services clearly have enormous potential as a data-rich resource to guide a campaign’s strategic decision-making.

– Alex Lundry

Cross-posted at Techrepublican.com

Leave a Reply

Contact Us

66 Canal Center Plaza No. 555
Alexandria, Virginia 22314

P – 703.535.8505





Sign up for email updates