British Gas sees progress weaning itself off Google’s ad server
British Gas scrapped Google’s ad server for an independent alternative 10 months ago, and hasn’t looked back.
Migrating ad servers is a costly and complex task — a common deterrent for both advertisers and publishers. But for British Gas, gaining a clearer view of how it could use its own data to drive conversions made the switch worth it, according to Patrick Smith, the brand’s digital marketing manager.
For instance, in January British Gas ran a display campaign through four demand-side platforms, and quickly realized that half of them underperformed. The result: Money was pulled from the two underperforming DSPs and pushed into the two top-performing ones where the cost-per-acquisition was lower. The campaign’s CPA was 42% more efficient after the display budget was consolidated into the two top DSPs, said a spokesperson for the advertiser. The attributed conversions from the two DSPs were 139% higher than what they would’ve been had British Gas stuck with the original mix of four vendors, said the spokesperson.
British Gas has been using Flashtalking since last September.
“With the likes of Google, due to their sheer size, they are able to simply price other ad servers out of the minds of advertisers,” said Matt Thorne, media account director, at media agency Agenda21. “Facebook, alongside Google, are also able to provide their attribution offering for free. So there is an obvious commercial element, especially if there are contractual agreements in place.”
For advertisers, the commercial savings associated with working with the walled gardens don’t always stack up against the potential costs of not being able to sense-check campaign data — which is harder to achieve using parts of Google’s ad tech. Last July, Google clamped down on how much of that data advertisers could take from its walled garden, in order to comply with the General Data Protection Regulation. But that put some advertisers in Europe in a tricky spot. For instance, prior to the data crackdown British Gas had planned to use Google’s wealth of user data, which spans a huge pool of synched cookies across publishers and advertisers, as the backbone of its own programmatic buying, from cross-site frequency management to attribution.
“Google then made their changes to remove log-level data which made us re-evaluate the way we were going to do attribution,” said Smith. “The ad server is core to where you start when it comes to making sure the data collected is correct and at a level of granularity that’s enough for us to measure the effectiveness of our media as well as know how many impressions we’re serving.”
The challenge for advertisers is how to secure transparency when regulation both in Europe and the U.S. could potentially limit the amount of data that gets verified. While Google isn’t as strict with its data in the U.S. as it is in Europe, advertisers anticipate it will be once the California Consumer Privacy Act comes into effect in January 2020.
Not every U.S. advertiser however, will look to explore alternatives to Google’s ad server in the same way British Gas did. Rather than switch ad servers, the head of programmatic at a CPG advertiser who spoke on condition of anonymity, said they would rethink how they work with supply-side platforms to create a unified ID outside of the walled gardens.
“We use Google’s ad server and because we’re a U.S.-based advertiser we still see the benefits of using the Doubleclick ID from Google,” said the same executive. “I doubt we will switch to an independent ad server. As challenging as Google can be they have the best-of-breed technology that allows our marketers to be efficient and effective in terms of how they buy our media.”
More in Media
Why podcast companies are investing in AI-generated podcast translations despite questionable quality
While podcast networks like iHeartMedia, Spotify and PodcastOne have publicly announced plans to debut AI-generated audio translations, few have gone live yet.
Legal pressure on AI companies illustrates the myriad challenges for companies that want to use or build generative AI tools.