"AdExchanger Politics" is a weekly column tracking developments in the 2016 political campaign cycle.
Today’s column is written by George Tarnopolsky, director of account management at Adelphic.
The 2016 election promises to be exciting not only in politics, but also in the world of advertising. For the first time, mobile and multidevice political advertising will play a significant part in the advertiser media mix.
Historically, political campaigns were challenging because of lack of ability to precisely target registered, prospective and on-the-fence voters. But in the 2016 election, candidates will be able to reach voters with more relevant and engaging campaigns, due to advancements in advertising technology and user identification.
While significant developments have been made in deterministic and probabilistic user identification, however, the industry still needs to conquer measurement, privacy and other challenges before it can realize the full potential of cross-device for political advertising.
Political views are deeply ingrained. In recent years, scientific thinking has emerged that a person’s political views are in direct correlation to a person’s personality or genetics. Therefore, targeting users on-message in a political campaign is more important than ever. Failing to deliver political ads with pinpoint precision can create wasted ad spend and campaigns that are at best irrelevant, and at worst laughable.
In the 2004 and 2008 elections, advertising was entirely contextual, with media buyers placing deals on sites or ad networks as a proxy for audiences. Although innovative at the time, this type of contextual-only buying was highly wasteful, especially for political campaigns. The 2012 election year signaled the emergence of demand-side platforms (DSPs) and the promise of audience buying at scale, but early DSPs were slow to adopt the technical and policy changes necessary to incorporate political audience segments into their platforms. In addition, early DSPs used cookies for targeting and measurement, making them unreliable in mobile and across multiple devices.
Another historical challenge for political campaigns: location targeting. In a presidential campaign, the focus is on high-value swing states, such as Ohio, Florida, Colorado, Nevada, Virginia, North Carolina and Iowa. Ad targeting must go a level deeper, down to the congressional districts that act as the fulcrum point for the entire state. While early DSPs included capabilities to target users based on state, city, ZIP code and DMA, they lacked the precision to target congressional districts, which are irregularly shaped and can include an overlap of multiple ZIP codes.
Mobile Location: Getting Better
Going into the 2016 election year, the ecosystem has matured to improve the precision and relevance of political advertising. New technical capabilities allow marketers to go beyond the cookie and IP address. Advertisers can now target prospective voters based on their mobile location, as well as probabilistic and deterministic user identification, all of which can fuel high-performing mobile and cross-device campaigns.
Mobile location signals, consisting of latitude and longitude data points, increase the precision of targeting prospective voters. By using latitude and longitude signals, custom polygons may be created for targeting users at the time the ad is served, as well as based on past location history. Latitude and longitude targeting allows media buyers to reach users with precision inside congressional districts.
Yet, these signals are fraught with problems, leading the marketing industry to demand greater transparency around latitude and longitude data integrity and validation. The majority of latitude and longitude signals are dubious, forcing some mobile data partners, such as Factual, to go to great lengths to clean and validate latitude and longitude data. Marketers thirst not only for better measurement and verification of mobile signals, but for supply-side certification of inventory as well.
New advances in ad tech can also solve the challenge of precisely identifying voter audiences. Third-party data providers today feature robust political data segments. For example, Neustar and BlueKai offer targeting of political audiences, including segments of registered Republicans, Democrats, independents, voters in specific elections and on-the-fence prospective voters. Voter data files are matched to household IP address info and translated into registered, voted and undecided audience segments, which can be targeted on the open exchange and on specific sites and apps.
CRM data targeting can also prove instrumental in reaching prospective voters, with several large publishers now supporting the uploading of CRM data for targeting within their ecosystems. For example, voter and donor records containing email addresses can be hashed and uploaded into Facebook Custom Audiences for targeting prospective voters within Facebook. While precise, this type of deterministic targeting raises questions around user privacy; while matching is done on hashed “depersonalized” data, initial inputs like user email addresses are decidedly personally identifiable information. In addition, CRM targeting in Facebook, Twitter, Amazon and other publisher ecosystems are confined to the publisher’s own walled gardens.
A similar flavor of addressable targeting may also be done at scale via a third-party data provider or data management platform. For example, voter CRM data may be uploaded into Neustar or LiveRamp for activation everywhere a DSP has reach. While a compelling tactic, this type of targeting creates for open questions around data fidelity as data gets translated from CRM to digital identity to person.
Lastly, advancements in attribution usher in the most exciting developments for political campaigns and the advertising landscape as a whole. By leveraging their expansive login data and device graphs, Google’s DoubleClick Campaign Manager and Facebook’s Atlas ad servers allow for seamless measurement of user path and conversion, even across multiple devices.
The key is accessing a turnkey integration with these leading ad server partners that allows for measurement of the efficacy of cross-device and mobile advertising efforts. A political advertiser will be able to measure voter engagement and conversion across multiple devices for the first time.
Still, some measurement gaps remain within these publishers’ walled gardens. For example, only Facebook’s Atlas ad server is allowed to fully serve and track ads inside the Facebook ecosystem. DoubleClick view tags are not allowed for Facebook Custom Audiences or in the Facebook mobile app. Twitter allows DoubleClick click tracking for promoted tweet products, but does not pass device IDs from its app into the ad server tags that enable cross-environment tracking.
So while publisher walled gardens promise a unified view of the customer journey, that view can also be difficult to assemble. Hopefully, this is just a temporary setback as the market naturally moves toward efficiency and consolidation.
The next election year will be exciting to marketers and voters alike. Marketers will have unprecedented ability to target users consistently across devices with unparalleled audience and location precision. In turn, prospective voters will see ads that are not only relevant, but also informational, engaging and influential for their future voting decisions.