Server-side header bidding may be a hot topic in certain publishing circles. But like any new technology, it has its weak spots. And not all publishers are convinced the trade-offs the method generates are worth it, yet.
The term is being used to describe the next phase of header bidding. Typical header bidding, known as “client-side or “browser-based header bidding” has become popular among publishers because of how much more revenue it can wring from programmatic ads by letting buyers bid simultaneously. Those publishers seeing such uplifts naturally want to increase the number of ads they can sell using the method. And for that, they need to plug in more demand sources.
The problem is that more demand partners means higher risk of latencies in the bid windows, which creates drag on the publishers’ pages. The result: bad user experience.
Server-side header bidding offers to take that headache away: Instead of the publisher absorbing the technology burden that arises from an online auction, it can move the heavy lifting onto a third party, be it an ad tech vendor or it could be the likes of Amazon, which is also poised to release its cloud-based header-bidding solution.
The publisher can then enjoy the increased yields, without any compromise to their page-load speeds, and any latencies forming in the bidding windows. They can also then fill their boots with as many demand partners as they like, driving up the volume of how many ads they can trade via header bidding.
And now, the caveats
But (and there’s always a but) like with any new tech, it’s not all rosy.
One drawback is related to potential user syncing issues. “Supply-side platforms that are running server side would no longer have access to the user’s browser and, much as demand-side platforms are doing now, will have to sync with the one player that has the tag down and is initiating the call from the browser,” said a head of programmatic for a newspaper, who preferred to remain anonymous.
“The concern is that this may lead to reduced match rates and lower yields on unmatched users,” he added.
Meanwhile, not all agencies are convinced that Amazon’s move into cloud-based header bidding will be good news for publishers. For starters, publishers using client-side header bidding — which means the demand sources are plugged into their own page header — get a good view of where the demand is coming from. They can control which of their partners’ bids they choose. “It may be a bit clunky, but it’s end-to-end,” said VCCP Media chairman Paul Mead.
Once they go for a server-side solution with a single vendor like Amazon, for example, that view may be lost. “They will just get served up whatever Amazon or Google, or whoever is providing the server-side solution. So what you gain on one side, you lose on the other,” he added.
Digiday Daily Newsletter
Not all agree, however. Some ad tech vendors are more concerned that server-side header bidding may be more harmful for SSPs than publishers. Unlike client-side header bidding where each vendor can see what is happening on the page, “server side empowers a single vendor to aggregate demand and then run the auction in what can be perceived as a black-box environment,” said Paul Gubbins, independent ad tech consultant.
The reason that’s bad news for SSPs is that particular part of the digital ad market has become fiercely competitive and commoditized. So there are growing concerns vendors that run the auction server-side on behalf of the publisher could be manipulating the auction in their favor, added Gubbins. “That means they could look at everybody else’s bid and then increase theirs to ensure they win the impression.”
Find out more about header bidding, including the lowdown on server-side bidding, at our Hot Topic: Header Bidding event in New York City on February 7. Executives from Turner, Hearst, The Washington Post, Meredith and more will speak.