From Header Bidding on Web to Unified Auction on Mobile (Part 1)
What is the most popular ad tech buzzword in the past two years? Depending on who you ask, many ad pros would start their list with “header bidding”. Many of the same people, however, would struggle to give a good elevator pitch on the concept.
With that in mind, we are kicking off a mini blog series on the topic of Header Bidding. In this first blog post, we will cover the basics of header bidding, followed by a blog post about the pros and cons of header bidding, a summary of our recent publisher survey on header bidding, and closing it out with the Chartboost manifesto on header bidding. Stick around for this exciting blog series!
So what is header bidding? Simply put, it’s a unified auction conducted by publishers. Header bidding replaces the traditional waterfall model of monetizing ad inventory. Header bidding adoption has been a revolutionary change for web publishers in the past three years. About 75% of top US sites are now using header bidding, with Index Exchange and Prebid as top header bidding technologies, and AppNexus, Index Exchange, Amazon, OpenX, and Rubicon as the top five ad platforms for display advertising.
Now mobile app developers are looking into similar solutions for in-app monetization. In this article, we’ll explain the basics of the waterfall approach and header bidding, track the growth of header bidding over time, and explain why header bidding is inevitable for in-app monetization.
The Basics: Waterfalls and Header Bidding on the Web
The waterfall approach on the web is headed for replacement because it has built-in flaws. Here’s how it works: when an ad placement on a web pages starts to load, the publisher calls its ad server, which is often Google’s DoubleClick for Publishers. Then a sequential process takes place: direct orders get served first until exhausted, followed by programmatic line items. To build up more direct orders, publishers spend resources negotiating with advertisers — but favoring direct orders don’t necessarily result in the best yield. That’s because programmatic buyers in lower waterfall positions might be willing to pay higher CPMs than direct orders. In short, relying on a waterfall loses revenue from high-value programmatic bids.
Header bidding creates a step ahead of direct orders, by conducting a new auction outside the primary ad server. This new step is still fast: ads are requested as soon as the page loads, simultaneously from all demand partners. The difference is that the partner with the highest bid wins unless the publisher has manually created an exclusion. Web publishers have seen a significant increase in eCPM and revenue after switching to header bidding.
The Trend of Header Bidding
On the web, interest in header bidding grew fast. Google Search trends for “header bidding” shows the term originating in mid-2015, reaching its peak at the end of Q1 2017, and maintaining popularity ever since*.
But what are users searching for header bidding via Google trying to find exactly? Here are the top 10 related search queries**. The results suggest people are most interested in knowing what header bidding is, and some top header bidding solution providers.
What countries or regions is the interest stemming from? Here are the top ten countries. This ranking probably won’t come as a surprise.
(date range for all graphs above: 1/1/2015-11/30/2018)
These trends indicate how fast header bidding has moved: in just three years on the web, it went from a nascent trend to a wildly popular technology.
Why Header Bidding Is Coming to the Mobile App
Seeing the success of header bidding on the web, the mobile community has begun asking whether there can be similar solutions for the in-app environment. Some companies in the mobile ad tech industry have already developed header bidding solutions under product names such as “unified auction” or “parallel bidding”. However, many of them are flawed because they don’t fully embrace the equal opportunity for bidding and instead implemented a hybrid model where there might still be a direct deal involved getting the first bid opportunity before it equally calls on all the other ad networks bids. Various hybrid models exist but yet they are communicated as “unified auction” or “parallel bidding”.
As it did on the web, unified auction is likely to continue growing fast in mobile. That’s because mobile advertising itself is exploding: mobile ad spend in the US is estimated to be $76.17 billion this year, the highest among all advertising channels, and grow to $141.36 billion in 2022, more than double the current TV ad spend.
How It Is Different from Mediation
Beyond the massive ad revenue many publishers are raking in, unified auction looks like a welcome relief from the race to improve mediation, the process of managing multiple ad networks with minimal SDK integration.
On one hand, the unified auction infrastructure relies on standard RTB protocols and forces the traditional ad networks to operate as programmatic bidders, reducing the number of SDKs needed (publishers today can have north of 20 advertising SDKs in a single app).
On the other hand, the dynamics of a true auction per impression give all demand partners the ability to bid on every ad request, which makes the publisher’s inventory more competitive, compared to mediation where each demand partner is called in the appropriate position on the waterfall.
Mobile app developers have experienced multiple rounds of monetization solution changes in the past decade, from the increase in SDK ad networks to the surge of third-party mediation, to programmatic sales. To keep up, publishers have needed to constantly evaluate new mediation tech and integrate new demand partners within ever-more complex waterfall implementations. And despite all the work, the waterfall can never overcome its basic flaw: inflexible sequential ad serving.
The Challenges to Overcome in Mobile
However, there are still hurdles to overcome in the adoption of mobile in-app header bidding or unified auction. Some publishers are very comfortable with pre-negotiated direct deals, and as a result, feel reluctant to switch over to pure unified auctions. In addition, some ad networks don’t have the capability to bid programmatically yet. As a result, some mobile in-app header bidding solutions are allowing a mix of hard bids and soft bids, which presents a compromised solution and inserts risks for publishers’ ads revenue.
What is clear today is that the monetization tech stack needs to be disrupted. As an industry, we need to address some of the biggest concerns of publishers: high yield with transparency, control, fairness, and low latency. 2018 is the birth year of mobile in-app unified auction (or in-app header bidding depending on who you talk to), but there is definitely more to come in 2019. The company who is able to solve the aforementioned concerns will own the future of mobile in-app. And every day it’s delayed, publishers lose.
Stay tuned for our next post in this series, which will cover the pros and cons of header bidding in more depth!
* Interest over time: numbers represent search interest relative to the highest point on the chart for the given region and time. A value of 100 is the peak popularity for the term. A value of 50 means that the term is half as popular. A score of 0 means there was not enough data for this term.
** Related queries: users searching for “header bidding” also searched for these queries. Top related queries are the most popular search queries. Scoring is on a relative scale where a value of 100 is the most commonly searched query, 50 is a query searched half as often as the most popular query, and so on.
*** Interest by region: see in which location your term was most popular during the specified time frame. Values are calculated on a scale from 0 to 100, where 100 is the location with the most popular as a fraction of total searches in that location, a value of 50 indicates a location which is half as popular. A value of 0 indicates a location where there was not enough data for this term.