How Parties can Leverage Data in this Year’s Federal Election Campaign
How predictable is voter behavior? Could big data really swing an election? The past US presidential campaign and the debate about the effectiveness of Cambridge Analytica have fueled the discourse about the role of data in election campaigning.
In the face of weakening party ties and rising numbers of swing voters, the question of the predictability of the electoral behavior is also becoming more and more focus of German parties. But what influence do data-based analyses actually have on the upcoming federal election campaign?
Forget about big data
While US campaigners have recourse to a sheer infinite amount of data when it comes to targeting voters, German data protection law sets a strict framework for political parties to use data for their campaigns. Particularly with regard to acquiring and merging data sets, the legal scope is very narrow. Consequently, powerful big data analyses with personalized scoring schemes and predictive modeling will not spill over the Atlantic soon. Instead, when discussing the impact of data on the upcoming federal election campaign in Germany, it will not be about big data or buying huge data sets from dubious sources but about smart data meaning the smart assembly and evaluation of easily accessible data.
Why data can still make a difference in elections
In the end, each and every campaign has a simple goal – mobilize as many people as possible to support a candidate, a party or an idea. It is a common mistake to perceive data itself as the Holy Grail of state-of-the-art campaigning. Data – no matter if its big data or smart data – is just a vehicle to achieve the overarching goal. Thus, instead of indulging in big data phantasies of almighty algorithms, it is about demystifying the data buzzword and focusing on the indisputable actual impact data analyses can have. Parties need to focus on how to use the “data vehicle” to leverage their campaigns within the sharp German legal framework and with the limited resources they can draw on. This challenge can be broken down in three simple dimensions: leveraging publicly available data, leveraging owned data and leveraging social network data.
Leveraging publicly available data
Door-to-door campaigning or canvassing experienced a renaissance in Germany over the last years since several studies indicated face-to-face conversation to provide the highest conversion rates compared to other, more impersonal forms of campaign communication. With the data available from the federal election commission, campaigns can analyze previous voting results down to constituency level and even polling station level. Herefrom, heartlands and weak districts can be easily derived. Combined with the voter turnout on polling station level, these data provide a campaign with in-depth knowledge of areas with strong previous results but low overall voter turnout and, thus, help to detect mobilization areas to focus canvassing activities on. Consequently, these analyses allow a far more efficient allocation of canvassers. Via app, these mobilization areas can be easily transmitted to the volunteers and campaigners on site. With Connect17 of the CDU and TzT, the door-to-door app of the SPD, there are already good examples of how this smart use of freely available data has found its way into the upcoming election campaign.
Leverage owned data
Political parties already have numerous data at hand – e.g. member data, data from supporter networks, newsletter signups and further associated data to mention only a few of the available internal data sources. This data is an asset that should not be underestimated. However, most of the data still awaits its use in the drawer. Thus, leveraging owned data requires the spread of a data-driven mindset within a party’s organization. Again, it is about demystifying data, about enabling party members and volunteers to access and use owned data. Therefore, the setup of a modern digital infrastructure that allows for data gathering and evaluation is inevitable. With all the available internal data at hand, the mobilization of the party’s resources and, subsequently, the decentralized organization of the campaign, e.g. steering volunteers via messenger apps, is no longer a big thing.
Leveraging social network data
Social networks like Facebook or Twitter provide campaigners with a variety of opportunities to address their target groups with pinpoint accuracy. For example, ads can be played according to age, sex, degree of education, presumed income or postal codes. For political parties, this means that they can reach their target groups precisely with individually tailored messages and hereby minimize scatter loss. In addition, A / B testing allows different messages to be tested against each other in order to find out which of the messages generates more interaction with the corresponding target group. In the consequence, campaign messaging no longer relies on the gut feeling of the campaign manager, but can be derived from reliable data insights. In this way, valuable deductions can be drawn not only for the social media campaign, but for the entire campaign.
Get the basics right
With the sharp legal framework in mind and limited resources at hand, data-driven campaigning will certainly play a role in this year’s federal election campaign. Due to described limitations, however, it will be mainly about getting the basics right meaning to leverage what is available. So instead of hoping for (or fearing) big data magic, parties should focus on their campaign groundwork – well thought-out strategy and consistent narration – and enrich it with data-based solutions wherever applicable.