A/B Testing — 17 Oct 2016
Optimize Your Apple Search Ads Creatives with A/B Testing
Maryna Dorash 
According to Apple, search drives 65% of all App Store app downloads, which is roughly 1.5 billion installs every month. This automatically makes Apple’s newly launched Search Ads one of the biggest ad networks for iOS. And since search has always been a channel that drives some of the most qualified and loyal users (both on the App Store and on the web), Search Ads is definitely worth trying out.
Today we take a look at how you can optimize your App Store Search Ads for maximum ROI.
If you’re familiar with Google Adwords, you should be quite at home with Search Ads. It is a very similar though much simpler beast. You bid on keywords you want to target and when users type in one of your search terms – your ad shows up on top of search results.
There’re several targeting options available including device (iPhone or iPad), age and gender, and whether users have your apps installed. Ads are only limited to the US at the moment, but we can expect a global rollout to be on the way soon.
There are several ad types: with and without screenshots. The tricky part is that the actual ads are automatically generated from your app page.
You can not set up and test different ad copy and creatives without altering all of your meta data. Which means updating the app if you want to try a different set of screenshots and a mandatory 16+ hour wait before any changes are applied to your search ads.
This makes the process of optimizing your search ads screenshots and descriptions extremely complicated and quite risky. Because the ads operate on an auction basis where you pay per tap on your ad – Tap Through Rate (the mobile world’s CTR) is critical. If your ads are underperforming you’re likely to either pay more or lose the auction and not get your ad shown at all. So the cost of error is very high because you can’t instantly roll back if you mess up.
Most of us would call the system a marketer’s nightmare; Apple would probably label the approach as ‘courageous’.
At SplitMetrics we’re proud to unveil our brand new Search Ads Experiments. A new type App Store of A/B testing that will help you effectively and safely optimize your Search Ads screenshots and descriptions. You can quickly test and iterate ideas and concepts without putting your store conversions at risk.
It works similarly to our experiment for organic search. You specify the search term and the competing apps that you want to test against and Splitmetrics creates a landing page emulating app store search results.
You create several versions of your ad listings that you want to compare against each other – something Apple doesn’t allow you to do. With SplitMetrics you can A/B test any element of you Search Ad:
Once you’ve set up your variations you can then run an ad campaign on Facebook targeting users with interests matching or related to your app to drive traffic to the experiment. Splitmetrics will distribute the visitors across your variations and show you how they perform.
You will get detailed stats on performance of each of your search ads variations and your competition and use these insights to take away and apply in your Search Ads campaigns.
And the best thing – by improving your search ads you will actually boost the performance of your app in organic search results as well. The formats of the search ads and regular search results are very similar – and our earlier studies show that users don’t seem to treat ads too differently.
Here at SplitMetrics we’ve run several search ads experiments of our own to see how much can be improved. The short answer – A LOT!
Disclaimer: The stats below are for experiments performed on SplitMetrics emulated pages, not the actual App Store Search Ads results. Prisma was taken just as a sample app.
Search Ads Experiment 1: Screenshots
Search Ads Experiment 2: Descriptions
