Kevel launched a new set of APIs on Thursday, called Relay, for publishers and their ad tech partners to build their own programmatic stacks.
You can think of Kevel as a plumber for plumbers.
The company, which rebranded from Adzerk to Kevel in 2020, offers APIs to make it easier for engineers to create and launch a first-party ad server, including APIs for native ads, sponsored listings and retail media. (An ad-server-as-a-service, if you will.)
Kevel’s overarching vision is to provide an AWS cloud-like infrastructure to support online advertising. The purpose of Relay, specifically, is to help publishers get more innovative with their programmatic monetization strategy, said James Avery, Kevel’s CEO and founder.
“So much of what you see in the programmatic space is about getting as many header bidding partners on a page as possible so publishers can make the most of their 300 by 250s,” Avery said. “We want to give people an API interface they can use to customize what they do without having to build from scratch.”
Relay includes APIs for running server-side auctions without tags or cookies and APIs to support direct DSP integrations. Companies also have the ability to set geo-filters, competitive restrictions and other creative controls.
Kevel charges for its APIs based on a SaaS model to avoid having to act as a middleman for its customers, Avery said.
“It’s like how AWS doesn’t charge you a percentage of revenue to host your website and we don’t have to give Zoom a cut to host a call,” Avery said. “We’re focused on solving engineering problems.”
Kevel launched Relay roughly 16 months ago in a private beta with clients, including Blockthrough, which helps publishers recover ad block revenue by adhering to the Acceptable Ads standard. The Acceptable Ads program allows publishers to monetize ad block users who consent to seeing so-called nonintrusive advertising.
Because of the amount of filtering Blockthrough does to make sure the proper creative is coming through, a publisher’s overall fill rate can naturally decrease.
“We realized pretty early on that we needed to build our own demand stack,” said Graham Michels, head of strategy at Blockthrough.
But doing so on the client side would cause additional latency for anyone with an ad blocker on, which was a definite no-go for Blockthrough. “These are users who are very sensitive about their experience,” Michels said.
Blockthrough ended up using Relay as its Prebid server. “That allowed us to add just one additional call and store as much demand inside it as we needed,” he said.
It would have been possible for Blockthrough to build the infrastructure itself, but in the unforgettable words of Kimberly “Sweet Brown” Wilkins, ain’t nobody got time for that. Or resources.
“If we had to invest in our own infrastructure and people, the amount of incremental demand we’d need to generate would be pretty astronomical,” Michels said. “Instead, we’re essentially offloading most of the work so we can focus on building our demand, which is an increasingly important part of our revenue mix, without diverting resources from our core product.”
The process is like the inverse of supply-side optimization, Avery said.
“Whereas before there might be a ton of extra, unneeded requests because of how many Prebid adapters there would be on a page, customers have the ability to only make bids when it makes sense,” he said. “It’s just a lot more efficient.”
Since implementing Relay, Blockthrough has seen fill rate rises between 7% and 12% for its publisher clients and incremental revenue of up to 10% in some cases. Blockthrough keeps a portion of the revenue it recovers on behalf of publishers.