The future of mobile app bidding is here: best practices for publishers to test the impact of bidding solutions

Tuesday, October 15, 2019
By: Sara Penchina

At GameDaily Connect USA, MoPub’s Head of Strategic Publisher Partnerships, Americas Kavitha Gadde took the stage to discuss the state of mobile app bidding — a topic that has become the focal point of considerable hype in the adtech community — and the results publishers and buyers alike are experiencing. She was joined by Conrad McGee-Stocks, Director Growth at Uken Games, Brian Truman, Executive Director, Digital Ad Revenue and Operations at GSN Games, and Amit Bhojwani, Head of Partner Management, Facebook Audience Network, who contributed insights and tips for publishers to accurately test the impact of in-app bidding solutions. Highlights from their talk as well as a video of the full session can be found below. 

 

"Header bidding" for mobile apps has become a hot topic — but does it actually work?  At GameDaily Connect, Kavitha, Conrad, Brian, and Amit discussed what results publishers and buyers are seeing from in-app bidding (at MoPub, our solution is known as Advanced Bidding), including impact on ARPDAU and filled supply. They also unveiled best practices for publishers to accurately test the impact that bidding solutions are having (hint: a real, effective A/B test involves much more than just turning on bidding and measuring ARPDAU). Key insights included:

1. Why bidding? 

In-app bidding is seen as beneficial across the ad tech ecosystem as a whole: for publishers, buyers, and mediation platforms alike. In-app bidding allows for greater competition and fairness, including increasing the ability to access inventory. It also comes down to control and automation: in-app bidding allows for less waterfall management, so that ad monetization teams can spend more time thinking about how to generate more inventory from a product perspective. According to Uken’s Conrad McGee-Stocks: “A benefit of in-app bidding from the UA [user acquisition] perspective is being in a situation where all partners can access inventory equally, and from a buyer perspective, not having to drag people through the waterfall.” In addition, app publishers can focus more energy and resources on building their apps and creating better user experiences. Ultimately, in-app bidding is beneficial because it creates long-term, sustainable value for the publisher ecosystem at large.  

2. Testing: how do you determine if it really works? 

To properly test the results of in-app bidding, a true A/B test is recommended as opposed to a “pre/post” test, in which a publisher measures results before in-app bidding is activated versus after in-app bidding is activated. Panelists who utilized A/B tests shared their thoughts during the session, discussing how they ensured proper test set up. According to McGee-Stocks, “MoPub was a great partner helping us run two separate waterfalls, test everything… exactly the same stack. I think that was really important and anyone who’s looking to evaluate the success of something like this should do that, if possible. Depending on the size of the publisher who is running the test and the more inventory you have, the more potential inventory you have that you’re missing out on from a revenue perspective.” 

GSN Games ran their A/B test in-house. According to Truman, “We were able to segment our audience (at least the enabled group that we turned MoPub’s Advanced Bidding on), through our own mechanism.” GSN Games split up an audience and decided which percentage of the audience would get a particular waterfall ad unit from MoPub, versus a group that did not use Advanced Bidding. They then set up the same waterfall on both sides. Truman noted: “We ran it for quite a long time. It takes a little bit of time for that bidder to adjust — just like any network does when you first test them — so we let that run for two weeks before we really started making any comparisons, then we ran it for another two weeks to see what the differences were. We ran it on both interstitials and rewarded video. While results may vary, it will vary on a number of things: 

  • Complexity of the waterfall today - if you’ve already invested a lot of time and management into a waterfall, you may not see as much lift as someone who hasn’t. 

  • Volume of inventory

  • Ad units, and whether you’re caching or not

  • Latency: decreasing your latency shows improvement [in test results].”  

For more details on testing methodology used during MoPub’s recent Advanced Bidding tests, check out pages 7-10 of our whitepaper here

3. Results: 

Buzz around the potential of mobile in-app bidding is at a fever pitch, as was certainly the case at GameDaily Connect. Per discussions of the importance of A/B testing above, MoPub’s A/B tests show that the hype behind the talk is real. According to Truman and McGee-Stocks, in conjunction with proof points evidenced in MoPub’s recent Advanced Bidding whitepaper, results showed: 

  • An increase of 5% to 15% in publishers’ ARPDAU.

    • Uken Games experienced close to 5% ARPDAU in Q2 2019, and GSN Games saw a 13% increase in ARPDAU on Solitaire TriPeaks over the course of a three-week test in Q2 this year

  • An increase in filled supply.

  • An increase in supply access for all programmatic buyers (Advanced Bidding networks as well as MoPub Marketplace DSP buyers).

  • An increased share of wallet for all programmatic partners. 

Based on the panel discussion, it appears that networks who do participate in in-app bidding experience a higher share of voice, and ultimately see an increase in revenue. From a network perspective, Facebook Audience Network’s Amit Bhojwani noted, “We’re seeing very exciting signals from some of our publishers not only in ARPDAU lift but even the way they’re quantifying the efficiency of the resources that they’re putting into these waterfalls and their management.” 

For the full results publishers saw using MoPub’s Advanced Bidding solution during recent testing, check out pages 11-12 of our whitepaper here

4. Why focus on ARPDAU?

When looking at metrics, the panelists agreed, ARPDAU (average revenue per daily active user) is key — more so than CPM. Often, as Truman mentioned, “when you have the conversations today with [ad] networks, it’s always a conversation around CPM because they’re trying to jockey for a position in the waterfall. When you no longer have to do that, the CPM metric really goes away.” From the network perspective, according to Bhojwani: “in the waterfall challenge today, it’s a multiplayer game, but single-player action — as you move to a metric like ARPDAU, you’re normalizing some of the variables that you either know about, or in some cases, you don’t know about. That’s what helps you save time and it improves the efficiency of decision-making in general.”

5. As a publisher, how can you prepare for in-app bidding?

From the publisher perspective, the biggest takeaway is to first and foremost ask your network partners to participate in bidding — getting network partners involved and on board will help level the playing field across the entire ecosystem. Next, make sure that all parties are committed to testing (including the development work). If you’re going to commit to testing, commit to doing it well and doing it right. Last, make sure you understand your supply side partners’ tech stack. Bhojwani closed the panel with this: “Overall, it comes down to resources that you have and optimizing the resources to what you’re actually looking to get out of [in-app] bidding short-term and long-term.”

For more insights and advice from the panel, check out the video recap from this session. 

Sign up for our newsletter to stay up-to-date with our articles, news, events and more.