Header bidding has been an important inflection point for the programmatic industry. By replacing the traditional waterfall, header bidding has given publishers increased competition, revenue and transparency into their auction dynamics.
In my last post, I discussed how open source header bidding wrapper technology improves efficiency and transparency. Today, I would like to unpack how moving to server-side header bidding, in particular, reduces latency.
Traditionally, latency is caused by a combination of a partner’s network latency with page execution. It can cause ad-call timeouts which in turn will impact demand and auction dynamics. If you want to learn a little more about causes of latency and how to optimize, check out what our architect, Abhinav Sinha, suggests.
Client-Side Header Bidding
Header bidding code, along with ad calls to requisite demand partners, can take place in the header of the page (this method is also known as client-side header bidding). By having multiple header bidders on the page, the publisher can achieve increased revenue. Many publishers manage their multiple header bidding tags within a container called a wrapper tag.
Knowing the right number of partners can be tricky, however. We frequently see publishers get yield improvements as they add up to five or six partners, and then diminishing returns above that number of partners. Publishers need to evaluate the latency versus monetization tradeoffs of adding incremental partners.
Each additional header code within a wrapper increases the page weight as well as adding to the ad calls made to demand partners from the browser. Thus, making the page heavier and causing latency. Connection speeds can impact latency as well. For example, if a user is accessing a page through a mobile device, client-side header bidding can have an even greater impact on page load times.
Latency is problematic because it slows the ad and page load times, which disrupts the user experience, leads to timeouts, reducing fill and ultimately monetization.
Server-Side Header Bidding
To better manage latency, publishers are moving to server-side header bidding and wrappers. Many industry thought leaders consider the move to server-to-server (S2S) the next innovation in header bidding.
Server-side wrappers offer many benefits but one the most impactful is reducing latency. Server-side header bidding can greatly reduce latency because most of the execution is taken off the browser and moved to a server.
Moving the execution to the server makes the page lighter, plus content and ads load faster, resulting in a decreased impact to the user experience. Server-to-server also allows for a faster response time between DSPs and SSPs.
S2S header bidding is still in its early stages but holds a lot of promise to help publishers achieve increased speed and monetization. For a variety of reasons such as the availability of demand or complexity of managing user syncs across demand partners, some publishers are not ready to move completely to S2S. These publishers can leverage a hybrid solution. A hybrid solution which combines client-side and server-side header bidding, thus improving access to demand and improving monetization.
PubMatic’s OpenWrap solution is a complete hybrid solution which publishers can leverage today. It allows you to maintain your current client-side header bidding demand in addition to managing server-side partners within a single solution.
OpenWrap allows you to leverage all of its UI based tools, controls, and reporting for both your client-side and server-side header bidding partners, as well as bringing you the value of a fully supported enterprise solution with technical support and a dedicated account management team. As publishers deepen their header bidding expertise, the popularity of hybrid solutions offering both client- and server-side integrations, such as OpenWrap, rose from 13.6 percent to 20.7 percent adoption rate between September and December 2017 according to ServerBid.