Google Analytics Provides an Easy Solution to the Problem of Web Lead Attribution
It’s close to the end of the year, and that means that the budget cycle for 2021 is finally coming to a close. At both Chase and Capital One this process involves ensuring each tactic passes minimum return hurdles in a stressed ROE situation or tightening credit standards. While some of the credit risk pieces are specific to the financial services industry, it strikes at the heart of the problem: marketing is a cost center that has to prove its worth. And proving worth for marketing has long been a challenge — how do you quantify the impact of a TV campaign, or a display ad, or a billboard? Is the return on these individual tactics enough to justify the cost of creative and execution?
Common Attribution Models Don’t Solve This Problem…
If we looked at this problem from a traditional last-touch (or even first-touch) attribution model, the answer is a simple no. How many people are going to see a TV ad or a billboard and remember the correct URL that attributes traffic to that particular channel? Hint: it’s an astoundingly low figure. Really, none of the models above are going to properly tell us what the impact of a specific tactic is.
“But that’s just an issue for traditional media, and we can use last-touch models for digital media” is a common rejoinder that frankly isn’t quite true either. Display ads via DoubleClick and video ads (especially pre-roll or mid-roll) generate immensely low conversions because they interrupt the consumer while they’re doing something else. Once again, the models above can’t tell us the true impact and return of display ads or other paid media tactics.
… But Existing Tools Can Help Solve For This Problem
As I’ve written before, there’s no reason to build your own product when an existing service does it really well. And lucky for us, Google Analytics actually does this for us, and does it very well. The above model looks at site traffic to tell us what sources end up converting, and which combination of searches give us the most conversions. However, if you also use Google DoubleClick, it can also tell you how many of these conversions ended up seeing display ads (even if they didn’t click on it) along the way.
Match Market Testing for Traditional Media
So that helps explain how to attribute all the web traffic in a clean out-of-the-box solution, but that still leaves traditional media as a big question mark. How do we apply this same methodology to pieces of direct mail read, number of billboards driven by, or number of TV ads seen? Thankfully, with enough sample size, anything is possible through A/B testing.
For targeted campaigns such as direct mail, it’s relatively straightforward to hold out a control and measure overall conversion for the targeted population vs overall conversion for the holdout group. That delta will tell us what the overall impact is and how direct mail slots into the multi-touch attribution models. For broader push campaigns such as TV or radio, partnerships with firms like Neilsen, or just running match market tests can give us our answer. Running a TV ad in Philadelphia and not running a TV test in Chicago, for example, can let us know the relative lift we’re seeing across all conversion channels.
While e-commerce happens online, there’s no reason that it has to only rely on digital advertisements. With the right tools for attribution measurement, any company can ensure that its marketing investment is providing an out sized return for the company.
