The phrase “Google Surge” (aka a “network blast” for those less keen on using a militaristic vocabulary for all things political) has been coined to describe the practice of blanketing websites viewed by inhabitants of a particular area with geo-targeted online adverts in the last few days of an election campaign.
The increasing sophistication of online advertising options , with the ability to serve up different adverts to different visitors, means the advertising blanket covers not just local sites, such as those of local newspapers, but also covers people from the constituency visiting broader websites too.
First used in Republican Bobby Jindal’s 2007 Louisiana campaign, the tactic has popped up in several other American elections since and was even honoured with a feature article in the October 2009 edition of Campaigns & Elections.
The ability to blanket online voters in a specific contest has a particular appeal in the American media market, where the alternative blanket options – principally TV advertising – can be hugely expensive, particularly if the coverage areas for TV stations do not match up with the boundaries of the election contest.
With a general election in the UK due by June 2010 – and most likely taking place in May – will this tactic be the next American import to British politics?
Tight geo-targeting of election advertisements has been tried before in the UK. In both the 2004 London elections and 2005 general election, along with Parliamentary by-elections since then, the Liberal Democrats ran geo-targeted adverts, for example. However it is telling that the company with the most advanced technology that I used during those advertising campaigns subsequently greatly scaled back its ambitions for selling such advertisements to the political sector.
Whilst geo-targeted adverts had their role, their impact was relatively muted. In part this was because of the relatively small size of constituencies in Britain compared with the US. The smaller the constituency, the harder it is to do run geo-targeted adverts which actually hit the right audience. In practice blanket coverage of the sort required by a surge ends up very unblanket like.
In the last few years, Google has significantly improved the granularity of the geo-targeting options available in the UK, which suggests the past may not be such a sure guide to the future.
However, UK constituencies have not gotten larger. Typically with 70-80,000 electors, it is very common for electors to work in a different constituency from the one in which they live. Even if the computers serving up geo-targeted adverts work out with perfect accuracy from which constituency someone is accessing the internet, this does not reveal whether they live in that constituency – or just work in it.
Geo-targeting is more likely to have a possible future role on a larger scale, such as blanketing a region that contains several marginal seats.
What may make geo-targeting a big success, though, is the media’s reaction to it. As in the US, the media in the UK love running process stories about elections and in particular about the Internet and political campaigning. However badly done or incomplete the coverage, a story about a “Google surge” may earn sufficient media coverage that it becomes a success regardless of how the adverts perform.
Relying on that sort of media coverage would be a risky approach, because the alternative media story could be to knock a Google surge as being a return to the sort of blanket newspaper and billboard advertising that has been seen in past British elections and subsequently widely discounted as an expensive way of achieving little.
Exciting modern American import or thrown back to old fashioned and inefficient campaigning? The coverage – and technique’s success – could go either way.