Atlas Solutions

Atlas Insights Series: Uncover more conversions with people-based metrics

By Rebecca Sharpe, Atlas Insights |

I researched my holiday vacation plans over the weekend at home using my personal laptop and tablet. I made my reservations on Monday at the office using a work computer. Advertisers using cookies for measurement believe my reservations were not influenced by online ads, because there’s no way to connect cookies alone across browsers or devices. My purchase was “unattributed” and the credit should go to TV, print or perhaps out-of-home advertising. But advertisers using Atlas people-based metrics know better; with Atlas, my purchase can be tied back to the ads I saw across devices and that it’s an online-attributable conversion. People-based metrics reveal a larger number of conversions, and therefore lower cost-per-action (CPA), than cookie-based metrics alone. Previously we’ve reported that on average advertisers can uncover 21% more conversions using people vs cookie-based metrics1. In this study, we consider the impact of people-based metrics on conversion counts for advertisers in our travel vertical, specifically hotel and air travel.

Remember how the online marketing world measures conversions: when a conversion pixel fires, we look upstream into the recent online history to see if an ad was served prior to purchase. If we find an ad in the history, the purchase is called a conversion; otherwise it’s just an organic purchase. Using cookie alone, the ad history will be limited to the device/browser of the conversion and only go back a finite period of time. Using people, the history can span browsers and devices and can last indefinitely. We investigated how the number of identifiable conversions increases with the time window considered prior to purchase. The longer the period of influence (lookback window) the bigger the difference2. See the graph below. Note that even for the most conservative lookback period – 1 day – people-based metrics reveal a 7% increase in attributed conversions.

Why does this matter? It means that your digital media is performing better than you’d previously thought. When you consider the CPA of your TV, print and digital buys side-by-side, you overstate the CPA of digital, artificially making it look expensive versus your other channels. You have all of the costs, but when using cookies alone, you’re only capturing some of the conversions. Using people rather than cookies, the CPA goes down by 7 to 22%, depending on the period of influence considered. Good job advertisers; you’re doing better than you think.

Atlas internal data, April 2014
2 Considered conversions in November 2014. We condensed multiple purchases during the conversion window to a single conversion.