• by Jana Fung , Op-Ed Contributor, February 13, 2017

In the past, it was simple. Publishers sold ad space to media buyers, booked campaigns in the ad server, and ads were delivered to users. There was full transparency over all the advertisers running on the Web site, what they paid, and how much revenue was earned. Yet over the past decade, advertising technology has seen a huge shift: Marketers are becoming more savvy with their budgets, which means that publishers have been forced to adapt their technology to optimize for revenue growth. From network mediation to what we now know as header bidding, we are seeing new and improved ways to optimize monetization on Web sites and apps.

Client-side header bidding or header bidding wrappers, from a technology standpoint, place JavaScript in the header tag, so all auctions actually begin in the user’s browser before the page loads. However, with this configuration, the number of exchanges and demand sources that can participate are limited, as each added partner adds significant latency to the ad-serving process.

The concept of server-side header bidding enables publishers to offer inventory simultaneously and directly to ad exchanges, demand-side platforms and direct sources of demand simultaneously to ensure the highest possible yield with no additional latency. This differs from header bidding wrappers because these direct unified auctions evaluate every bid net of revenue share to ensure that publishers get the highest effective CPM yield on every impression.

Now we are seeing a shift, as publishers are moving from header bidding wrappers to server-side header bidding. Here’s why:

Increased Revenue with Static Page Load Times

This may sound obvious: More demand has a direct correlation with more yield for publishers. However, adding more header bidders effectively means that more JavaScript has loaded, which increases page load time. For mobile in particular, the increased page load causes an exponential increase in latency, so publishers tend to limit their header bidding partners to a maximum of five.

To solve this problem, server-side header bidding taps server-side connections to enable an unlimited number of simultaneous bid requests, with no increase in page-load times.

Simultaneous Auctions from Multiple and Infinite Exchanges

Because of the limitations of client-side header bidding, it is difficult to evaluate demand from multiple exchanges for every ad impression. Server-side header bidding allows publishers to include demand from multiple ad exchanges, increasing competition and yield. This comes with many other benefits such as increased fill, diversified advertisers, and more transparency.

Total Control with Standardized Inventory Rules

For browser auctions, each header bidder added to the page actually makes it more difficult for publishers to effectively control their ad-quality standards. Because header bidding wrappers do not provide solutions to manage ad quality, publishers are forced to manage each individual real-time bidding system to track, source, and block ads, making the process more cumbersome as the number of partners increase.

With a server-side auction, there is a single pipeline that creates unified management of ad quality. This means publishers can pre-establish a framework that controls ad quality through one server, reducing the overall time spent on investigations when issues invariably arise.

In a world where it is inevitable that publishers and developers relinquish control over their inventory and user experience to maximize revenue, the S2S paradigms allow publishers to regain what they have lost and scale further.