Many publishers are selling their inventory through server-side bidding because it helps them improve their page speeds. But unlike browser-based header bidding, where bidding code is more visible to all parties involved, server-side bidding operates in a nontransparent environment because its code is only visible to the companies who control the server-to-server connections between disparate tech vendors. Eric Berry, CEO and co-founder of native ad platform TripleLift, spoke to eMarketer’s Ross Benes about how server-side bidding could change the pricing dynamics of programmatic ad buying.
Why is pricing different in server-side bidding compared to browser-based header bidding?
With client-side header bidding there is a self-verifying mechanism that ensures, with some exceptions, the prices are accurate and there's nothing but a genuine auction going on. In a server-side header bidding world, you just generally don't get that. There's an infinite set of possibilities that you could be doing that make things complicated or reduce transparency between the buyer and the publisher.
How might server-side bidding be used to hide pricing tricks?
Some SSPs pay on gross before fees are taken and some SSPs pay on net after fees are taken and [browser-side header bidding products] adjust the bid prices according to whether the SSP is putting in a bid at gross or net. But it’s not clear in server-side bidding that the bid adjusting is occurring. A vendor could submit impressions as gross and beat out those who submit as net through accounting as opposed to the reality of what the publisher is getting paid. It's such a nuance between gross and net that nobody would know what's going on.
Why wouldn’t people be able to spot this in server-side bidding?
Nobody can audit your auction.
Have you seen any companies use server-side bidding to hide their true prices?
I cannot know for sure that anyone is doing this, and I’m certainly not naming any names.