Following the Prebid Leadership Summit in October, Dikshant Joshi (pictured below), director of product management at AdPushup, penned this exclusive article for ExchangeWire on the importance of Prebid for publishers, and how the technology is set to evolve.
Why prebid is popular among publisher communities
Header bidding is a response to the inefficient waterfall method of ad serving and the walled gardens of the duopoly. When header bidding first emerged, a lot of players started building proprietary tech—poor standards and lack of cooperation between ad tech vendors further led to black box solutions. However, AppNexus decided to make Prebid open source, meaning, publishers did not have to be an AppNexus client to use the wrapper. As Prebid grew in popularity, other companies started collaborating to bring together the vision, guidance, and engineering capabilities of the ad tech community to improve header bidding. Simply put, Prebid is popular among publishers because of the power of the open source ecosystem behind it. It’s free-to-use, transparent, delivers market-leading performance, and integrated with more than 150 demand partners, over 15 analytics partners, and many managed solution providers.
The essentials for an effective wrapper
A good wrapper should follow the guiding principles of agility, openness, transparency, neutrality, and fairness. The key features that a good wrapper should offer include the ability to integrate multiple demand partners (while being demand agnostic), unified reporting for transparency and openness, and frequent security and feature updates. Another major feature a good wrapper should have is the ability to select the best bidders with respect to different users and their demographic and targeting data, based on the geos and device type, for starters.
Selecting number of demand partners
For a single website, this would depend on the geos, device type, ad sizes, etc., as there would be partners excelling for particular audience segments. Given the different network bandwidth of users across geos and devices, it makes sense to have different number of partners on mobile vs. desktop, for instance on the mobile, you could have the top 5-7 performing partners whereas on desktop, you could have 10-12 partners—this will help keep latency in check while delivering good performance. I wish there was one magic number for every website in the world, unfortunately there isn’t, hence continuous A/B testing is the best bet.
Key learnings from the Prebid Leadership Summit?
The community is actively working on solving major bottlenecks in the adoption of Prebid Server, Prebid Video, and Prebid Mobile SDK, while grappling with the industry challenges like the user identity crisis due to the war on cookies, which could result in a tectonic shift in how ad targeting works. The key takeaways from the Prebid 2020 Roadmap are:
– Making Mobile SDK integration easier
– Dynamic Floors for all bidders
– Unified Reporting API
– Standardisation of User Identity Modules
– Programmatic Guaranteed is coming to Prebid Server
– Monetisation of longform video via Prebid
Are server-side wrappers/tech a threat to Prebid?
The industry adoption of Server Side technologies has been really fast, with the introduction of Google’s Open Bidding (previously called EBDA) and Amazon’s UAM/TAM—these solutions have become a part of the monetisation strategy of many publishers. However, this doesn’t mean that Prebid has been or will be displaced, rather, a lot of publishers are utilising a hybrid implementation to get the best of both client-side and server-side technologies. It’s still early days though, and some view server-side technologies as the new walled gardens, so on the contrary, they may pave the way for adoption of Prebid server. When the adoption increases, client-side header bidding will evolve further to work via the Prebid Server.
The future for Prebid
Publishers and ad tech partners working on the web inventory have adopted header bidding as an important technology component in their monetisation stack. 2020 will be about how different wrappers start utilising massive bid landscape data, and apply big data and machine algorithms to further optimise for higher revenues. This could mean automated prioritisation and personalisation of demand partners based on their bid rate, win rate, etc., in different geos, devices, and ad sizes. This could result in significant lifts, the algorithms could be handy for configuring smarter timeouts in Prebid, based on the response times in different geos/devices.