While much of the programmatic world is still adopting header bidding, here in the UK, this transition took place long ago. We are now in a stage of maturity whereby we’re all looking at ways to optimise rather than increasing supply and demand. As noted in a recent eMarketer report, nearly nine in 10 digital display ad dollars will be spent on programmatic inventory in the UK this year. And that adoption isn’t slowing down.
So, with this focus on optimisation in mind – as well as the seemingly infinite amount of data and insights potentially available – on which key points should programmatic players across our ecosystem be focused to ensure they’re operating as efficiently and effectively as possible? That’s the question, and one that we – two analysts looking at opposite sides of the programmatic coin – feel well-equipped to answer.
The bid cycle
For those on the buy-side, understanding the criteria required to turn a ‘Bid’ into a ‘Bid Chosen’ (i.e. the highest bid in an auction, which is then submitted to the ad server) is imperative. As an ad exchange, we typically bucket and classify the various stages of the bid cycle as the following: Bid Error, Bid Time Out, Bid Pass, Bid In Review, Bid Blocked, Bid Eligible, and finally, Bid Chosen.
Understanding the first few stages (i.e. Bid Error through to Bid Blocked) allows marketers and buyers to dissect and analyse what might be preventing a bid from becoming eligible for auction. In the case of ‘Bid Error,’ for instance – which measures the total number of times an exchange receives a response containing an error from the bidder – a bid may be missing data or showing an invalid seat ID, thus preventing it from ever making its way into the auction. Understanding these metrics, and the elements which may be preventing a bid from ever being submitted to the ad server, is imperative to boost your win rates.
Now, let’s say you have a bid that’s successfully made it through the bid cycle and is ready for submission to the ad server. Another important element to consider is bid participation – a figure calculated by looking at the bids recorded within the exchange, commonly known as ‘bids chosen,’ compared to the bids recorded in the ad server. Put differently, it’s the answer to questions like, “How many bids actually made it into the ad server? And what percentage of bids were ‘lost’ along the way?”
There are a number of ways through which bid participation can be improved, particularly around the technical implementation of header bidding solutions. If you’re reading this as a publisher, the mapping of the integration is one of the first things to analyse. For instance, if there are changes to the configuration of the slots on your site, this must be reflected in the integration of inventory within your header bidding solution of choice (alongside ensuring this is being correctly targeted within the respective line items).
Further, timeout has a major influence on bid participation, as there is a direct correlation between timeout and bid participation. There are an array of factors that need to be considered regarding timeout. We recommend utilising the available data to help drive this decision – consider everything from the volume of bids timing out to the average latency each showed in order to correct and capitalise on any missed opportunities.
Prioritisation within the ad server
Last, but certainly not least, is the importance of prioritisation and fairness in the ad server. Having jumped through an endless number of metaphorical hoops and hurdles in order to make it into the ad server, the last thing you want (as a buyer or seller) is to lose out on the highest bid. Understanding that different publishers have different set-ups and strategies, and there’s no ‘one-size-fits-all’ approach to prioritisation, there are a few tips and tricks to keep in mind. Namely, ensure partners are bidding in net, bidding in the same currency as the line item set up and that there are no clashes between waterfall tag set-ups and your header bidding partners.
Collectively, by taking the time to evaluate these various metrics and making the necessary adjustments, those on the buy and sell sides of the programmatic ecosystem can significantly boost their monetisation opportunities. It’s all in the numbers – you just have to know where to look.