Foursquare’s Attribution Solution a Step in the Right Direction, but Still Leaves Gaps for Marketers

Foursquare—the location-based social network that now calls itself a “location intelligence” company—recently stepped into the analytics business. The company’s entree is a product called Attribution Powered by Foursquare that is intended to help brands measure how media impacts foot traffic in brick-and-mortar locations.

To fuel this intelligence, the company pulls data from a panel of 1.3 million opted-in users of its Foursquare and Swarm apps in the U.S., who have agreed to leave location sharing always-on, regardless of whether the apps are open or not. The solution works by comparing a group of media-exposed users from the panel to a similar non-exposed control group, and comparing location visits for the groups to measure incremental uplift.

For the moment, Foursquare’s solution only reports the basic demographic data of the visits: age, gender and location information. While this solution is an interesting new way for marketers to bridge the gap between online marketing and offline actions, it still doesn’t account for actual sales. There is still plenty of headroom for innovation when it comes to creating a holistic, multi-channel attribution strategy.

Smarter Attribution Means Measuring More Touchpoints

Foursquare’s solution will certainly be a boon to multichannel marketers, but as a standalone it still suffers from the same issue that plagues the online industry: single-point attribution.

The most accurate attribution models are those that give credit to not only the final media touchpoint preceding an action (for online, the famed “last click”), but to all of the assisting touchpoints that might influence a person to act—whether it’s walking into a store, searching for a brand name, or engaging with a mobile banner. In our connected world, a competing brand is never more than a mobile search away, so it’s critical for brands to attribute credit to the marketing campaigns, channels and vendors that influence a person throughout the customer path to purchase. These are thus called “customer journey” attribution models.

In theory and in practice, customer journey attribution is no walk in the park. It requires defining how important each touchpoint is in achieving the desired outcome. For example, a business with a long sales cycle may attribute the highest percentage of credit for a given sale to the first touchpoint. By defining this “attribution curve,” marketers can then split the effective media costs across the various touchpoints. Multichannel advertisers have it the hardest, because they must take into account media touchpoints and actions that happen not only across different devices, but also offline.

Cross-Device Attribution

With today’s U.S. internet-savvy population using an average of six devices, being able to not only track user interactions on each device, but to also connect those actions across devices will be paramount for complete customer journey attribution.

There are two schools of how to match audience profiles across devices: probabilistic and deterministic. Probabilistic user matching is statistical; it analyzes large amount of interactive data (like connected networks, visited locations, and visited websites), finds patterns in that data and pairs likely matches. On the other hand, deterministic user matching depends on a unique identifier, such as an email address or customer ID, to link up interactions that happen across devices.

Probabilistic matching has its limitations. While having the benefit of scale, probabilistic is most often criticized for a lack of accuracy. Deterministic, on the other hand, enables marketers to not only implement consistent segmentation and personalization strategies on the execution side of marketing, but to afford more accurate attribution on the measurement side by enabling marketers to measure media exposures for each logged-in device. The issue is that it requires the user to sign in (for example, on mobile web or on apps) in order to be effective. This is why apps, with their persistent log-ins, will be so important for the next generation of digital marketers looking to fill the gaps in the customer journey.

Multi-Channel Attribution

Multi-channel attribution, connecting offline and online media exposures and actions, can be even more challenging than cross-device, but it’s not hopeless for those looking to include offline marketing in their customer journey attribution.

When it comes to measuring the impact of offline marketing like radio and TV on online sales, marketers often use correlation analysis to make an educated guess: if direct website traffic increases dramatically around the time that a major TV campaign runs, there is reason to believe that some or much of that traffic came as a result of the TV ads. This is the method that Foursquare used to demo its new attribution solution just following the Super Bowl.

On the other hand, in order to accurately measure the impact of any type of media (offline or online) on offline sales, advertisers must make use of a unique identifier that can be matched both in-store and online. Perhaps the most reliable example is the customer loyalty card: by tying all customer activity to a single loyalty ID or email, the advertiser can track both in-store and online purchases.

The Road To Customer Journey Attribution

For digital attribution, it’s clear that single-point models must go. But without an accurate attribution strategy in place that takes into account the individual across the entire customer journey—across devices, online and on foot—advertisers might find themselves prioritizing incorrectly and inefficiently using their budgets.

Whether focusing investments more heavily towards new, potential consumers or towards re-engaging valuable customers that may have been dormant for several months, a cross-device, multi-channel attribution strategy is vital for advertisers who are looking to maximize their advertising spend. Time will tell whether Foursquare can scale to play a valuable role in that model.

JB Brokaw is the president for Sociomantic Labs, North America, and is responsible for sales, revenue operations and developing the company’s advisory services teams. He brings more than 15 years of experience in digital marketing and search technology, which includes previous positions as both chief revenue officer and chief client officer for digital agency, iProspect.