Google vs Header - Bidders from the perspective of publishers
28 November 2016
Google’s domination of the technology behind serving adverts onto websites continues to grow. But the relatively new and fast growing header bidding technology may just disrupt that. Should Google be worried? And what is best for publishers?
First, what is header bidding? At its simplest, header bidding is a way for a publisher to receive multiple bids for ad inventory from multiple sources all at the same time. Importantly, header bidding gives publishers much more visibility of bids actually made than through the traditional ad-serving techniques. It is different to the system of real time bidding through the waterfall because a publisher can see in advance what advertisers and media buyers will bid compared to the waterfall process where bids are accorded to a pre-set order of priority, sometimes meaning that a bid from higher up the waterfall will be accepted by the ad server when there might have been a higher bid from lower down the waterfall.
A further advantage of header bidding is that Google's structural advantage is removed. Google supplies the dominant ad-server and is also a key SSP (Google's Doubleclick for publishers known as DFP). Because of the dominance of its ad server Google can see the value of bids from other SSPs and then is able to trump them with bids from its own SSP.
Further, to publishers the Google ad server is effectively a black box solution because, while publishers know that ads are being served by the Google ad server and are receiving revenue for those ads, publishers don't really know how Google has chosen the ads. There is a suspicion that Google might choose one bid over another to suit Google's own commercial interests. The reason Google's competitors are promoting header bidding is that it allows more transparency in the bids received by the publishers and this removes Google's advantage of being both a major SSP and the dominant ad server.
Once adopted and working on the publisher's side, header bidding should mean more bids from other SSPs which should mean inventory is sold to the ad buyers willing to pay the most, which means yield for the publisher should go up. OpenX (an ad tech company which is a proponent of header bidding) claims its clients have experienced a sustained increase of 50% on their ad revenue.
For a much more detailed explanation of how the waterfall and header bidding works, visit www.adopsinsider.com.
Google's First Look
Google is not sitting by and watching a competitive process take hold without a fight. Its solution is a new product called Google First Look. It works in a similar way to header bidding because it displays all bids for ad inventory made through the Google ad eco-system to the publisher, who can then select the highest.
A draw back to header bidding is that there is a technology investment that needs to be made up front by the publisher. For Google First Look the publishers simply need to agree to subscribe to it if they are already using the Google ad server - there is no upfront cost. Further, load times do increase using header bidding, so potentially pitting a publisher's ops team against its own ad sales team. For these reasons many commentators are predicting that header bidding will not take off.
Using Google's First Look is cheaper to implement but involves signing up to more of the Google infrastructure. But Google's First Look may result in lower improvements to yields for publishers than header bidding because only advertisers who bid through Google's DSP will have their bids shown in First Look.
Facebook's entry into this battle could change everything. Facebook has unparalleled data on its users and is able to use that data to generate significant revenues by selling ads on its own platform.
But now Facebook wants to challenge Google where Google is currently dominant; in the world of ad-serving. Facebook Audience Network ("FAN"), which has been around for a while now, allows advertisers who advertise on Facebook to target the same individuals or lookalikes on other publisher's platforms. FAN is heavily focused on mobile and uses Facebook's own data to target impressions at consumers or lookalikes. And now Facebook is launching its own header bidding technology with FAN.
Having FAN as one of the bids for publishers using header bidding is an exciting development for publishers and could tilt the balance towards header bidding, since currently Google's DFP does not currently allow FAN as part of its service.
Advertisers who already successfully market on Facebook will be attracted by the increased reach other publishers can offer and publishers will be attracted by the higher yields that the quality of data Facebook can offer will hopefully bring.
So which will win?
Ultimately, publishers need something to improve their online ad yields. As income from print ads continues to fall, the revenue from online advertising has not yet got anywhere near replacing it. Publishers need to monetise their online content somehow and will be willing to try alternatives.
If header bidding really does increase yields by the double digits that its proponents claim then Google's current dominant and extremely profitable position must be under threat, even if Google's First Look does become the preferred solution over header bidding. If nothing else, the rise of header bidding means that publishers have options and can demand more from their ad tech partners. The response by Google in launching a header bidder like solution in First Look means that publishers no longer have to settle for the waterfall system through Google's ad server.
If publisher's online ad revenue yields do not increase then ultimately no one will win in the long term. So the developments in ad tech should definitely be welcomed.
As a final footnote - watch out for our next article on the privacy challenges inherent in ad tech from the publisher's perspective.