Rather than one particular university, Lewis concentrates on hundreds via BSD. But much like his Keypath colleagues, he recently found himself questioning established methods.
“The biggest challenge is driving volume at a CPL level that's acceptable for us,” he explains. When Lewis collaborated with the Bing Ads team on a solution that would test mobile bid adjustments, it was important that they could do so in a controlled manner while measuring the impact not only on CPL, but also KPIs like spend and lead volume. Driven by the discovery that the platform they use for mobile bidding on other search engines didn’t handle Bing, Lewis designed something undergraduates typically dread: a test.
“We have to generate CPLs that are well under $20 to be profitable,” Lewis says of the test, which compared a two-month pre-period to a two-month post-period. “Since it was pretty much a fresh slate at that point, we wanted to make a test out of it and say, hey, before we weren't really modifying these bids, now we want to modify them in a way that's controllable and we want to monitor the results.”
Using a tiered approach based on mobile average position, Lewis boosted the ad group level mobile bid adjustments in the BSD’s top-spending campaign.
“We wanted to do it in a way that was somewhat uniform across different things, so we weren't just setting manual adjustments based on random factors,” he explains. “We wanted to do something that was pretty scalable across the entire account.” They created percentage tiers based on the average position. “Anything above a certain position we did at one percentage. Anything between this position and that position we would do at a different percentage,” Lewis says. “That allowed us to roll it out across the entire campaign, in a uniform manner that we could easily monitor and adjust.”