Header bidding is no longer shiny and new, but are you doing it right? Inarguably one of the most widely discussed publisher technologies these days, header bidding has gone from a niche yield optimization tool to pretty much standard operations for many publishers.
Here at AOL, we implemented our header bidding solution on our own properties and now offer it to our publisher clients—because we firmly believe in the power of an open, transparent ecosystem versus walled gardens that stifle competition and publishers' ability to reap the full value of their inventory.
Given how rapidly header bidding technologies and ideologies evolve, we spoke with premium publisher clients Paul Bannister, EVP, CafeMedia and Paul Bell, CRO, 33Across to shed some light on why it’s crucial for publishers to get smart on header bidding, and let us in on some best practices.
Q: Header bidding has taken the publisher community by storm. What factors were involved in your decision to use header bidding? How has this impacted your bottom line?
CafeMedia: We felt that standard, tag-based programmatic was a very poor system for us to get access to programmatic budgets which we knew were growing rapidly. Header bidding was the obvious solution to this challenge. It has been a very large part of growing our revenue in the last two years.
33Across: We took a firm stance on header bidding adoption. We believed it would address critical issues facing the publishing community such as ad request latency, a lack of a true auction environment, traditional waterfall limitations and low fill rates. Our strong revenue growth results confirmed this theory.
Q: What are the characteristics you looked for in selecting a header bidding partner(s)? Are these characteristics the same/differing for container solutions vs. integrated bidders?
CafeMedia: Unique demand, low latency and true incrementality are the most important elements we look for in our header bidding partners. We don't want to add header partners that aren't bringing something unique to the table and also want to make sure that they are truly lifting our overall yield and not just moving dollars from another partner to them.
33Across: We took a selective approach when we evaluated header bidding partners. Specifically, we reviewed a partner’s successful past performance with 33Across, capability to work within our own bidder framework and available support resources provided during the new technology implementation. Non-bias is extremely important to us. We aim to be truly programmatic and strive for equal competition across all demand partners.
Q: Google is almost universally acknowledged as the reason why header bidding solutions were developed; to get beyond the closed relationships of dynamic allocation. Does the recent announcement of Google opening dynamic allocation to select outside demand sources mean anything for you? Does this diminish the need or effect of header bidding?
CafeMedia: There needs to be more clarity about exactly how the solution works and how it is priced, but we don't believe it's a relevant solution for us. The idea is great, but we're concerned that pricing and lack of transparency will nullify any value it does bring. We do think that server-to-server integrations and unification of demand are critical for the long-term success of header bidding/open dynamic allocation, but are skeptical that the current solution is workable.
33Across: Google’s announcement was welcomed as it supports the need for a true auction environment and validated the need for impartial dynamic allocation. However, it is early to tell if it will effectively serve both demand and supply sources.
Q: What are best practices/tips you can offer to publishers that are thinking of adopting header bidding solutions?
- Flexibility of partners is key. Every publisher's setup, approach and properties are different and require a customized approach. One-size fits all approaches can sound appealing but in reality rarely work.
- Don't treat header bidding like standard programmatic - you'll see little return that way. You should completely rethink the way your ad server waterfall works and ensure you are maximizing your value and your advertisers' value.
33Across: Adopt a solution that will not conflict with other header bidding implementations, and continually test ways to streamline the process and optimize the integration.
Q: What are some of the biggest challenges that still need to be solved for?
CafeMedia: There is a still need for education around the ins and outs of publisher inventory configuration and how header bidding and waterfalls impact advertisers on both the direct and programmatic side.
33Across: We’re currently working on a few challenges when it comes to header bidding including establishment of multi-size auctions which aren’t available in all platforms, continued improvement of latency in cases when header bidding times out and universalization of header bidding implementation.
Q: What’s next? Name your top 2 predictions for the publishing industry in the next 12-18 months.
- Private Marketplaces will become a much larger part of the puzzle - it's a way to modernize the direct relationship/IO-based business of the past and bring publishers and buyers together more.
- Unified auctions (via container tag or server-to-server) will start happening, allowing publishers to maximize their yield across header partners, and buyers to get the most valuable impressions.
- Header bidding technology will detect each bid request’s viewability/fraud levels and reject those with low scores, resulting in higher quality inventory within the marketplace.
- Like most marketplaces, there will be some publishers that prefer a one stop shop and others who prefer to customize their monetization stack.