.png)
A perfect storm is brewing in the digital ecosystem, and marketers and data vendors alike are caught in it. Frequently, marketers are willing to shell out large amounts of money to purchase audience segments to target users with whom they have no pre-existing relationships. At the same time, data brokers frequently use black box machine learning algorithms to compile these audience segments. The result is an ecosystem that compounds data inaccuracy, lacks transparency, and bloats marketing budgets.
This storm has been active for many years now — just ask any CMO. Recently, a May 2020 article in the Harvard Business Review — written by Catherine Tucker of MIT Sloan and Nico Neumann of the Melbourne Business School — joins the chorus and highlights this same problem.
Summarized by Tucker and Neumann, "marketing managers do not know whether they can trust the audience data, despite the fact that these purchases typically make up a large portion of their media budgets." To make matters worse, their research indicates that the quality of demographic-based audience segments — for example, segments that target young women— were "particularly disappointing." The solution for marketers, Tucker and Neumann argue, is to be "cautious about deciding to buy audience profiles for targeted marketing,“ and specifically to consider "waiting [to purchase audience segments] until they have reliable visibility into the quality of what they’re buying.
Well, the wait is over. Marketers have an actionable solution to the problem put forth by Tucker and Neumann, and it's being independently developed by Truthset. Truthset computes Truthscores™, the first solution that quantifies the accuracy of consumer data at the record level. That is, Truthscores™ fall between 0.00 and 1.00 and quantify the probability that a given demographic assertion (e.g., is this individual 18-24?) about a given ID (e.g., a hashed email, a person-level marketing ID, a phone number, etc.) is true.
For data vendors, Truthscores™ make it possible to differentiate themselves in crowded marketplaces and call attention to their higher-quality data segments. In addition, Truthscores™ give marketers detailed, direct insight into the quality of the data they use to fuel costly campaigns.
The beauty of Truthscores™ is that they're so easily scalable. For our beta product release, the Truthset team built an engine that evaluates the quality of over half a billion distinct IDs (hashed emails, to start) across 9 different demographics. Taken together, that is way more than half a billion Truthscores™ all computed in a matter of minutes — and that's only scratching the surface. The same Truthscore™ engine can look at any consumer-level identifier (name, address, phone, mobile ad ID, etc.) and measure its accuracy tied to any consumer attribute.
What's more, from this point onward, Truthscores™ and the Truthscore engine™ enable Truthset to re-measure and re-evaluate the quality of consumer data on a consistent, ongoing basis. That is, Truthscores™ are not a static metric, but rather with each successive release, they dynamically reflect the state of data quality in the current digital ecosystem.
If you are a data vendor, marketer, or platform interested in working alongside the first company to tackle head-on the problem of data quality in the digital ecosystem — and in the process, finally distinguishing the good data from the bad data — please contact Truthset at truthset.io.