• by Mike Proulx , Op-Ed Contributor, October 12, 2018

The right place at the right time with the right message. In the ad industry, how many times have we heard this rallying cry as one of the greatest benefits of digital marketing? Certainly, there’s a lot of promise when it comes to personalization, one-to-one, and people-based marketing. But to actually deliver on that promise requires not just data (the currency all of us in marketing covet more than ever). It necessitates clean, accurate, and up-to-date data—and therein lies the problem. 

Facebook’s Cambridge Analytica scandal helped shine a mainstream light on ad-targeting practices and put increased pressure on social media platforms, ad tech companies, and data aggregators to be more transparent about consumer data. As a result, many are now providing windows into the data they use to deliver personalized ads. As a digital marketer who relies on the accuracy of this data, I’ve been particularly interested to see how I’m being targeted, so I sought out to piece together my data profile. 

What I found is a person I barely recognize. Sure, some of the basic demographic info about me was generally in the right ballpark, but much of it was over a decade old. And when it comes to interest-based targeting, much of the data about me is totally off the rails—not only inaccurate, but just plain false. One data provider thinks I’m interested in cooking, antiques, and smoking tobacco, while a major social media platform targets me on things like Olympic weightlifting, the Republican Party, and beer. 

Those who know me are #ROTFL because in real life, I’m an organic prepared-foods buying, modern-décor loving, off-the-scale anti-cigarette, “my body is my temple,” gay Democrat (who’s never thought about Olympic weightlifting…until now). At best, about 40% of the data I found that’s used to target me was accurate. That would also explain why a year ago I came home to a package (addressed directly to me) with baby formula samples that said, “As a mother, you’ll discover strength you never thought possible.” 

With so much waste in our data stacks, what’s a marketer to do? Here are three things to help get your data strategy off to a good start: 

Ask the tough questions: When choosing a third-party data provider, find out what they’re doing to ensure their data is accurate. Ask them about their data cleansing, updating, and maintenance practices. Understand if and how they’re using third parties to validate their data integrity. And find out what percent of their data is deterministic (known people) versus probabilistic (look-alike models), with a preference towards the former.

Test your targeting: Often you can’t control or choose your ad platform’s data source(s). But you can prioritize first and second-party data to increase the likelihood that your data is accurate and maintained. When third-party data is being used, create a “test factory” with data and systems that you own. For instance, you can tag and track exposure to targeted campaigns against your customers via your DMP. While limited in reach, you’ll get a good sense of how accurately you’re delivering to your intended audience before activating your full dataset.

Start broad, then narrow: Accepting the fact that data is flawed, start your campaign off with only basic targeting parameters and then use re-marketing tactics to narrow, personalize, and optimize based on how your initial target engages with the various paid, owned, and earned assets that comprise your campaign. You might think that you’re wasting impressions by doing this. But you might also find an unanticipated audience response to your brand. It’s both okay and smart to give your targeting configuration a little breathing room. 

In the end, the promise of effective personalization is a great thing. At its best, it provides relevant, welcomed content and experiences along the customer journey, efficiently. We’ve seen time and time again increased conversion rates for clients when we’re smart about the mashup of data, ad targeting, and dynamic creative. But let’s not fool ourselves—It’s neither perfect nor precise. Personalization is just one of many tools in a marketer’s tool box—It’s not the Holy Grail.