The Tapad Device GraphTM Had Twice the Precision and Three Times the Scale as Next Competitor
New York, September 14, 2016 – Just-released findings of a Hotels.com® study revealed that Tapad’s (part of Experian) cross-screen marketing technology achieved the highest levels of precision and scale among competitors. According to the leading online accommodation booking website, after a rigorous, three-and-a-half month vendor analysis, Tapad achieved twice the precision of the next highest-scoring cross-screen offering and three times greater scale. The two other companies evaluated were not named.
Said Helene Cameron-Heslop, Senior Manager of Analytics of the Hotels.com brand, “Our team implemented an extremely rigorous vetting of open, cross-screen technology vendors. At the outset, we assumed we would have to compromise on either scale or accuracy – particularly given the importance to our brand of operating in a privacy-safe setting. We were surprised to find a complete package, but Tapad’s Device Graph won out on scale, accuracy and privacy; making our choice of partners very clear.”
In another metric critical to the Hotels.com brand, The Tapad Device GraphTM was eight times more “unique” than the next closest offering, meaning Tapad’s graph was found to have a much greater number of connections not seen in any of the other graphs.
In addition to precision, uniqueness and scale, the Tapad Device GraphTM was found to have:
● 100% higher recall
● 47% more incremental matches
● 53% higher North American market coverage
● 101% higher F-Score*
“A valuable cross-device solution should enable partners to get everything they’re looking for from a single vendor,” said Tapad Founder and CEO, Are Traasdahl. “We are deeply impressed with how thorough Hotels.com was in their vetting, and we confidently tackle the complex challenges of the martech industry thanks to our superior technology. Everyone loves a bake-off, and Tapad is no exception – delivering best-in-class results in areas that really count.”
*F-score is a statistical measurement that takes precision and recall together. The calculation is 2*(precision*recall)/precision + recall). It gives you one number instead of two numbers to look at and judge performance.