WTF are server-to-server connections?

This article is a WTF explainer, in which we break down media and marketing’s most confusing terms. More from the series →

Header bidding has been a hot topic all year among publishers looking to squeeze more money from their programmatic inventory. But the method, which lets buyers bid on online ads simultaneously, has its drawbacks. The more demand sources that can be added to a wrapper tag, the higher the risk of page-load latencies: the stuff of nightmares for publishers, especially in a year rife with ad blocking. It’s also a prime reason why not all publishers are sold on header bidding.

That’s why many are now eyeing something called “server-to-server connections,” which isn’t new technology; it just hasn’t been used at scale for the evolution of the kind of unified auction header bidding has opened up.

For some, server-to-server is where the market is leading. Others in the industry have called header bidding a short-term solution and a “hack.”  Here’s a look at what it actually means.

So, WTF are server-to-server connections?

Server-to-server is an evolution of the principle of header bidding, just using different technology and processes. It still allows for a unified auction (rather than waterfall), but instead of the bidding happening in the publisher’s browser, it happens on the server of whoever the partner is: either an independent ad tech vendor that provides server-to-server connections or the bigger tech players developing in this space, like Amazon.

Great. What does that mean?
“It means the vendor now has to do the heavy lifting, so the publisher doesn’t have to. We have the bandwidth and can absorb the infrastructure,” said Paul Gubbins, U.K. country manager of programmatic platform, and header bidding provider, Pubmatic.

How exactly is it different from header bidding?

Header bidding was brought about as a way to flatten the waterfall process and help increase the value of a publisher’s programmatic ad yields by letting buyers bid simultaneously. In doing so, that’s driven up competition in the auction, which has led to higher yields for the publishers. But the problem arises once a publisher wants to use header bidding for more of its inventory: Plugging in more demand sources (additional pieces of code that get added to the publisher’s header) causes page-load times to drop. Wrapper tags have helped to a certain extent by allowing publishers to collect tags in one place. But by going for the server-to-server option, publishers can, in theory, offload the entire auction bidding process to the ad tech vendor’s server.

What does that mean for publishers?

The main benefit is that in shifting the auction process out of the publisher’s header, they can avoid risk of page-load latencies. And that also means that publishers needn’t cap how many demand sources they plug in. Meanwhile, the publisher will also have fewer partners to deal with, because they’ll all just integrate into the server.

Amir Malik, programmatic chief at national newspaper Trinity Mirror, said the publisher is already using a server-to-server option, and that the efficiencies that method has driven are “unbelievable.” He believes that using this method gives the publisher the ability to speed up programmatic ad serving in a big way. “We can report within three minutes on an impression that is bought server side from a provider that is faster than what a typical SSP can provide,” he said.

And the vendors don’t have any issues with this?
This is being worked on, according to some ad tech execs. One exec, who preferred to stay anonymous, said it’s still the case that nobody wants “to play” in anyone else’s wrapper. “Part of the problem we have is that no one really trusts the other ad tech partners, and everyone thinks whoever has the wrapper will have competitive advantage. It’s just fear of the unknown, but we’ll get there,” said the exec.

What are the downsides?

The same with anything in ad tech: Publishers must do due diligence on what the scope of the partner is that they’re supporting, and how many exchanges are actually working with them. “You need to be sure that you’re not missing out on whoever you were working with on the buy side, when you were still using a client-side solution,” said Drew Bradstock, svp of product for Index Exchange. After being burned more than a few times, publishers are more closely scrutinizing their ad tech partners and demanding more transparency on fees and bid processes. “Publishers won’t go back to arbitrage and opaqueness. They’ve had a taste of knowing what’s been happening and aren’t going to put up with it,” he added.

So what does it cost to switch from header bidding to server-to-server?
It shouldn’t cost anything for publishers. Or so some ad tech vendors believe. Header bidding is an open-source piece of code that sits within a publisher’s browser. “It should be free, not another opportunity to charge more ad tech tax,” said Bradstock. This shift in server-based auctions is also going to lead to a lot more consolidation among header-bidding vendors, he added.

https://digiday.com/?p=213977

More in Media

AI fatigue sets in among workers and company leaders

About half of business leaders report declining company-wide enthusiasm for AI integration and adoption, according to a recent EY pulse survey.

Media Briefing: The top trends in the media industry in 2024

This week’s Media Briefing takes a look at the top trends from 2024, from AI licensing deals to referral traffic challenges.

WTF is agentic AI?

Generative AI is being shoulder barged out of the way by the latest term du jour: “agentic AI.”