Header bidding, sometimes referred to as advance bidding or pre-bidding, is a technique used by publishers to provide content to multiple ad exchanges at once—before ever making a call to their ad servers. If you’re not familiar with the process, this may sound hopelessly complex, or even useless to you, but the reality is it’s a potentially powerful mode of serving ads, and it’s undergoing a rapid evolution.
The Automation Problem
In the digital advertising world, publishers and advertisers are crazy about automation—the process of matching and delivering ads without any manual interference. This can help maximize efficiency by improving ad relevance and ensuring that every ad served yields the maximum value. But behind those forms of automation are multiple parties, multiple platforms, and a ton of code trying to manage extremely complicated exchanges. Accordingly, the programs we develop to further automate and increase efficiency are becoming increasingly complex—and while they sound great in theory, they also demand refinement and improvement if they’re going to become long-term institutions.
Header bidding is a perfect example of this paradigm. With it, publishers (who feature header ads, or any other type of ads onsite) can instantaneously and simultaneously send out a request to (potentially) every ad server they have. Rather than adopting a “waterfall” style approach, which starts with one ad server and works its way to others, one at a time, this allows multiple ad servers to bid at once, theoretically increasing yield for the publisher.
There are a handful of advantages to the strategy. First, everything happens at once, avoiding the one-at-a-time dilemma from the old model. Also, because multiple bids can be received and evaluated at once, publishers have the opportunity to choose higher bids and maximize their potential returns. Also, because inventory doesn’t have to move back and forth, it allows publishers to sell ads on a calculable per-impression basis, giving potential customers great insight into the actual value of the ads.
For publishers, there aren’t many downsides. The first is the automation complication factor we addressed above; trying to create and install a header bidding script that actually works is a developer’s nightmare. But perhaps more importantly, the simultaneous callouts to multiple ad servers can severely slow down the load times of the publisher’s site, dragging down the usability factor.
Why Does the Future of Header Bidding Matter?
This article is looking to the future of header bidding. Why? Partially because it exists in a kind of “unfinished” state. In theory, it’s a good solution—and certainly better than the models publishers relied on previously. It helps secure a higher yield and does more to get the best ads in front of customers. Unfortunately, it’s sloppy and imperfect, designed as a kind of temporary solution to the ad serving problem.
The future matters because this somewhat sloppy solution has a lot of potential to be refined—it’s like a word on the tip of your tongue, and all it needs is a good push to become fully developed. What that push is and where it comes from may hold tremendous value for the future of header bidding.
Google’s Latest Rollout
The most significant development in recent years has been Google’s latest rollout to DoubleClick, its ad server. Previously, Google’s ad server would estimate what a potential outside exchange might bring in, and disallowed bids for every impression. Header bidding circumvented this issue—serving as a duct tape fix—allowing publishers to get real prices, and not just estimates.
Now, Google has rolled out a new feature called “dynamic ad serving,” and it’s changing the game. This new feature will allow for server-to-server communication, enabling multiple ad servers to submit real bids at once, essentially creating a kind of shortcut in the old header bidding process. Google has evidently been working on the feature for months, and has developed it purely as a way to improve the average user’s experience, while still maximizing the potential ad yield for publishers.
There are still a few hiccups to address with this configuration, however. For example, buyers’ technology costs could rise if they’re met with the same query from different exchanges. Plus, if buyers submit different bids to different exchanges (a relatively common occurrence), they could end up outbidding themselves.
Three Main Avenues for Development
In any case, Google’s dynamic ad serving solves many of the problems that header bidding faced previously, including slower load times and complicated information exchanges. It’s not perfect, but it’s a good next step.
Header bidding could theoretically go three different places from here:
It’s too early to tell exactly which one of these paths header bidding will follow, as Google’s dynamic ad serving is still new. If the feature works well, it’s likely that it, and other corporations, will pursue this line of development, working out even bigger and better features to aid publishers in multiple-bidding scenarios. If the product is just as hectic as header bidding, or if a powerful new alternative mode comes up, it’s likely that the concept of traditional header bidding will fizzle out entirely. If Google’s option isn’t satisfactory, and third-party developers can outrace the corporate giants, a remastering of header bidding could also take place.
If you want to stay up on the latest in ad serving and video advertising technology, or if you’re interested in improving your video advertising campaign, get in contact with Verta Media today!
May 16, 2017
May 03, 2017
April 27, 2017
April 26, 2017
March 14, 2017
January 18, 2017
January 17, 2017