Updated: Apr 24
Prebid deployments start with a simple mandate: bring quality demand to a publisher’s inventory in a transparent and fair auction. Over time, Prebid implementations become more complex and difficult for publishers of any size to manage. To give publishers back control of their Prebid implementation, PubWise pioneered the Prebid as a Service space with our industry first offering and are continuing that leadership with machine learning driven Smart Path Optimization Technology.
In a typical Prebid implementation, a publisher sees a notable revenue lift and is very happy with the results. As a publisher begins to increase their use of header bidding, they may deploy a config for each section of their site. However, after spending more time using Prebid, the publisher’s AdOps team might realize a single script isn’t meeting their needs. They might realize, for example, that they have an ad unit that is only on certain pages, or there is a different config for one section versus another. They find that while the config works, it’s actually suboptimal in terms of presenting the best view of their inventory to demand partners.While Prebid doesn’t complain about these suboptimal configs and many demand partners continue bidding this way, this leads to low calculated win rates and other metrics that may be used to adversely judge a publisher’s inventory.
After dialing in site layout and ad layout, the AdOps team will begin to see differences between mobile and desktop performance. They will also begin to notice that the U.S. has a much higher bid rate when compared to other countries. There are even locations where some demand partners simply won’t bid.
For a typical publisher, these combinations and options would look like this:
5 Ad Units
5 Page Layouts
3 Connection Speed Categories
Now that the publisher is paying closer attention to this level of optimization, they quickly realize their prebid deployment has more than 2 million potential configurations. Trying to manage that type of number with any manual system is a not feasible.
We have a solution.
PubWise Smart Path Optimization Technology
Traditionally, all ad servers have relied upon a great deal of human intervention for testing and yield management. At PubWise, we think this heavily manual process is an artifact of a bygone era. Having the capability to directly insert machine learning into the adserver and the header bidding config makes such a system obsolete.
This is why we built Smart Path Optimization Technology (SPOT) into our Prebid as a Service Platform. With SPOT, you arrange a baseline configuration. SPOT then analyzes variations of that config to determine the optimal setup across all of the possible combinations. This reduces the test/update cycle from months to days — and even hours. Whether you have a small set of combinations to analyze or hundreds of thousands, SPOT can dramatically reduce your workload. And it does so without development work on your part.
SPOT gives publishers the ability to manage a large number of demand partners. It can reduce bid volume by as much as 40% while reducing page latency, increasing ad viewability and improving the user experience. Since SPOT is able to target demand to various audiences, the gross number of demand partners active on any site can be expanded. This increases competition and pushes up revenue in header bidding and AdX.
At the same time, demand partners benefit from SPOT by reducing their data infrastructure costs by the same volume. This gives them the assurance that their bids are not being wasted on misconfigurations. Furthermore, running SPOT also improves win rate.
All of this comes in an easy-to-use, point-and-click, hands-free package. It’s a self-driving car for advanced Prebid optimization, and it’s available today.
Whether you’re a publisher with a millions or billions of impressions, or if you are looking to relieve the stress of managing your Prebid implementation, contact us. If you are a demand partner looking to reduce the costs associated with header bidding get in touch for a demonstration of Pubwise Prebid as a Service with SPOT.