What if you could easily perform tests to find out what increases your advertising revenue? Opti Digital launches an A / B testing solution that allows you to make agile monetisation decisions. The company will present this solution at Ratecard Publisher Meetings on 29 and 30 January in Barcelona.
A / B testing applied to monetisation: an Opti Digital innovation
We were familiar with A / B tests for improving the conversion tunnel for merchant sites, A / B tests and multivariate tests for improving the lead conversion rate a landing page by advertiser-oriented solutions such as Google Website Optimizer. Opti Digital innovates and applies the principle of A / B testing to advertising monetisation. The tool makes it possible to display two different configurations in parallel, for the same period, to two groups of Internet users, and to analyse the KPIs.
Example of an A / B test: Adjustment of the floor price of a mobile space
To illustrate Opti Digital’s A / B testing solution, we analysed the case of a travel site from 16 to 22 January 2020, in France, with the mobile audience. (See screenshot above.) The publisher’s inventory is divided into 2 groups: a control group and a test group. You can easily define the percentage of traffic sent to the test group. The higher the percentage, the more reliable data we have quickly to make a decision.
- In the Control group, a minimum floor price of € 0.20 CPM is configured.
- In the Test group, we apply a more aggressive floor price, at € 1 CPM, for one of the mobile advertising spaces.
Everything else remains the same: the viewability rates are optimised thanks to the Opti Smart Tags developed by Opti Digital. This technological component improves the quality of the inventory: viewability, protection against malicious advertisers, responsive behaviour, loading speed, etc. The programmatic selling partners are the same, competing in header bidding (Prebid.js and Amazon).
Analysis of the results of the A / B test
In this example, eCPM impressions are better in the Test group. Which could lead us to think that this version is the most profitable. But by applying a higher floor price, we increased the share of unsold inventory and the number of impressions per page decreased. The consequence is a slight drop in the net Page RPM, the publisher net income per 1000 page views and the most significant final KPI. In conclusion, the Test version is less profitable than the Control version. The CPM floor price of € 1 is too high.
We use the A / B Testing solution when we want to connect a new seller or confirm an optimisation idea. Some publishers also send us test ideas and ask us to set them up. This tool allows us to deploy the test and analyse reliable data quickly.
Sébastien MOUTTE, President of Opti Digital.
A / B testing ideas to improve ad monetisation
Here are some A / B testing ideas to implement on your website:
- Native Google style / Native style adapted to CSS site design
- Add / Remove a selling partner
- Extend / Shorten header bidding timeouts
- Add / Remove an advertising space
- Competing the mega banner and the skin / Keeping a leader board when a skin is displayed
- Add / Remove certain sizes within a space
- With / Without the automatic insertion solution for ads within Opti Digital content
- One CMP provider / Another
- Adjust the design and UX of the CMP
Find out more
Be creative! Send us your A / B testing ideas and we will work together on to implement them in your site. Contact us
