Attribution measurement can be a key tool for determining which marketing channels work and which can help cut customer acquisition costs.
Marketers in verticals like financial services have long understood that they need to play the long game when it comes to acquiring new customers. Considering the potential payoff, it can be common to see numerous special promotions and premiums offered (free toaster anyone?), and acquisition costs in the hundreds of dollars. In the past, that was perfectly acceptable.
Today, a new breed of financial service startups are doing more with less. New business models for online banking, mortgage lending and even wealth planning services are mirroring the experiences learned in direct mail and online to capture mindshare. But as these disrupters barrage consumers with multichannel offers, they are also promoting churn and potentially eroding the stickiness of customer relationships.
Both traditional institutions and this new breed of online-first financial brands need to look to attribution measurement to see which channels are providing the best marketing ROI.
The flaws of the legacy attribution model
Historically, attribution has been determined by sampling and probabilistic forecasting. Unfortunately, samples based on small match rates can easily skew model results, reducing accuracy. But these legacy models aren’t just inaccurate, they are late. Typically, marketers commission attribution studies that arrive months after the campaign, and the findings often reflect what the marketer wants to hear. While the legacy model is a challenge for all marketers, it’s especially burdensome for financial marketers for two reasons.
First, high acquisition costs and churn will exacerbate the pain of a slow and wasteful attribution model. As a result, financial marketers dig themselves into a deeper hole—one that leaves fewer dollars for innovation in various media channels. Moreover, the longer you are beholden to a legacy model, the deeper that hole becomes.
You May Also Enjoy:
- Attribution is Still A Huge Challenge for B2B Marketers: Report
- Attribution Models: Tips for Deciding Which is Best For You
- Driving Results With Granular Measurement
Second, the legacy attribution model creates inherent legal risks. Digital targeting, combined with slow and inaccurate attribution, could be exposing the issuer to redlining violations without knowing it. For example, a credit card company might use credit scores to build an audience for its campaign, but if the advertiser doesn’t have transparency into who it’s targeting, how do they know if their campaign is discriminatory? Put simply, you have to be able to validate the audience in order to correct wrongdoing before it becomes a legal problem. Providing transparency should be top of mind for financial advertisers in light of the government’s decision to bring suit against Facebook for discriminatory advertising practices.
A new model for attribution requires correlating an increasingly broad range of channel signals, from customer postal addresses to device IDs, IP addresses, emails and mobile geo-location. It is key that this technology is able to match offline and online signals to a common reference point at scale. Through location-based tactics like postal, GPS and geo-fencing, marketers can improve the scope and accuracy of omnichannel attribution.
The other key change needed is attribution that is integrated directly into the campaign workflow. Financial marketers have an opportunity to shift from a backward probabilistic attribution model to one that is deterministic across both traditional and new marketing channels.
By incorporating attribution into the daily workflow, marketers will be in a position to optimize media spend while the campaign is in progress. They’ll be able to pivot from email to display to connected TV, reducing waste and better targeting the most profitable customers.
There’s a lot of business up for grabs, and winning that business now represents decades of revenue. But a mobile-first, digital native, cord-cutting millennial simply can’t be accounted for by the legacy attribution model. It’s time for a new attribution model and a new workflow that can keep up with the consumer.