But what is it all about, why has it been so influential, and where is it going next?
In this blog post, we’ll take a deeper dive into this programmatic advertising trend.
What is header bidding?
Put simply, header bidding is a programmatic way of selling ad space that gives publishers a lot more control over the process, including who they choose to sell to.
Impressions can be sold individually, rather than being lumped together by the thousand, and publishers can also gather valuable information about the buying habits and average spends of different advertisers, which helps them make informed decisions about how much their ad inventory is worth.
What’s all the fuss about?
Before we get into that, let’s quickly remind ourselves what the header bidding model was up against: waterfall auctions. In a waterfall-style auction, publishers chain together ad servers in such a way that each impression is offered to the first provider initially, then, if they pass, it goes to the second, and so on.
From a publisher’s point of view, there are two big problems with waterfall auctions. First, you don’t get to choose who you work with, and second, the figure your impression sells for doesn’t necessarily reflect its true value.
That’s because whenever an impression goes onto the auction pile, the “highest tier” (advertisers that have spent the most in the past) get first dibs. If no one in this tier goes for it, it’s passed down to the second tier and so on until someone bids for it.
What this means is that the pool of people bidding for your ad at each stage is limited. There might be someone in the next tier down who is more than happy to pay a higher price, but they never get the chance, because someone in Tier 1 or Tier 2 already offered a low price and had it accepted.
With header bidding, on the other hand, everyone bids at once — which can quickly push the price up much higher, bringing more money to publishers. At the same time, advertisers get a fairer chance to bid for the ad inventory they want, because the higher-tier advertisers don’t have a stranglehold on the most sought-after spots.
Google’s answer to header bidding
The astonishing ascent of header bidding has been a bit of a wake-up call for Google.
You’re probably familiar with Google’s DoubleClick platform, which publishers use to serve ads and offer inventory to advertisers. Google has long used this platform to give itself an advantage over competitors like OpenX and AppNexus, since it could see the whole of a publisher’s inventory and bid on the best stuff first.
This doesn’t happen with header bidding, because the highest bidder jumps to the front of the queue. In response, Google last year launched an open beta of an alternative it calls Exchange Bidding — making the argument that it’s faster than header bidding. In February, Google and its partner in Exchange Bidding, OpenX, said that more than 200 publishers are using it and they’ve increased their revenue by an average of 48 percent.
What should you expect for 2018?
Given the benefits that header bidding brings to publishers and discerning advertisers, it’s not surprising that this has taken off so fast. But what will 2018 bring? Let’s take a look at some of the key developments to look out for.
Smarter and smarter header bidding wrappers
Header bidding wrappers (also called frameworks) have already taken off in a big way. These are used by publishers that work with a bunch of different header bidding solutions at once to make sure that all bids are triggered simultaneously and that these bids are translated into something coherent that the ad server can understand. The aim is to make sure that bids are processed in such a way as to take advantage of the best offer every time.
We’re now seeing a “new wave of header-bidding wrappers” that give publishers even more control over the process. As David Hayter, the director of programmatic, data & technology at Shortlist Media, explains:
“Publishers will gain more control over their programmatic yield with a new wave of header-bidding wrappers entering the market, which allows publishers to close a second-price auction based on first prices from all demand partners. This, in turn, leads to closer ties between buyer and seller.”
The rise of server-side header bidding
Initially, all header bidding was of the client-side (or browser-side) variety. This uses a snippet of code embedded on the publisher’s site to connect to all the different sources of buyers that might want to bid for advertising inventory.
The big problem with this is that it can drag down site speed.
If a lot of advertisers are connecting to the header bidding wrapper directly, that’s going to run a lot of JavaScript processes on the page, increasing the latency of page and ad load times. What’s more, everyone bidding will have a different response time, so the whole auction can only move as fast as the slowest bidder.
Some publishers try to tackle this performance issue by limiting the number of partners that can bid in the first place, but that, of course, goes against the big draw of header bidding. The whole point is to have more people bidding for your inventory, pushing up the price!
That is where server-side bidding comes in.
Server-side header bidding tackles the latency problem by removing the actual auction process to an external server. Publishers still need to embed the snippet of code on their sites, but rather than sending requests from the browser, they’re sent from the ad server to all the relevant supply-side platforms (SSPs).
This takes less implementation and makes little difference from the ad-server side — you can opt to set up just one set of line items across all partners, your site runs faster and the whole process becomes far more scalable.
If you’re a publisher and haven’t already made the switch, this is the big technology development you’ll want to adopt in 2018.
As published on MarTech Today