App Store Keyword Optimisation for Acquisition
Part 2 of a series on: Experimenting for sustainable App Growth
By Caterina Toniolo
This series is based on a presentation I gave at the Epic Growth Conference in Moscow in March 2018 and revolves around app growth at Skyscanner.
In this mini-series, I’d like to talk about app growth at Skyscanner. In the first post I spoke about the challenges of App Growth we face at Skyscanner and shared the first case-study of this series about the HALO EFFECT in India. Follow us to keep up with the next post in this series and others from the team at Skyscanner.
This is post #2 of this 5 post series.Today, I’m going to continue by sharing a second case study on app store keyword optimisation.
Case study #2 Acquisition: App Store Keyword Optimisation
Tribe: EMEA Growth / Squad: Russia
The Challenge: Despite all the valuable work done on ASO (App Store Optimisation), we believed that there was still room for improvement. There were a few opportunities we thought we didn’t take, like:
- We were using only 95 characters out of 100.
- Using long keywords — example: бронирование takes 12 characters, which is 12% of overall keywords. We could have included two keywords instead.
- We could have used more relevant and high-volume keyword. For example, the word автомобиль (car) is not relevant for us. Although it receives a lot of a traffic, the competition is fierce and brutal.
- More keywords variations and combinations could have been achieved by using adjectives
- No iOS auto-suggest keywords have been considered.
Based on these considerations we created a new optimised list of keywords.
The Hypothesis: If we add relevant and high-volume keywords in iTunes Connect, we’ll increase rankings of these keywords in App Store.
Metrics: The primary metric will be ranking of keywords, the second metric will be % increase of organic installs.
How will we know if we’re successful? We’ll consider the experiment successful if we manage to improve rankings of at least 50% keywords otherwise this will be deemed a #failforward
Method:
We’ll use a two-sample T-test and compare two sets of data with the confidence level is 95%.
- Control group: is the historical average rankings for 3 months before testing the new keywords
- Test group: is the new average rankings for 1 month after the introduction of the new keywords.
Results: We improved the ranking of more than 50% of the keywords, and in general rankings were 77% better (lower).
Tip! Use a two-sample T-test to see if there is a statically significant increase but you cannot do a proper split test or causal impact analysis.
Grow with us
Fancy working with a great team in one of our global offices across 10 locations? If you see a role you like and interview successfully we’ll offer to relocate you. Take a look at our perks then our Growth Marketing, Design and Engineering job roles within our Growth Tribes.
About the Author
Hi, I am Caterina, I am part of Win Market by Market Tribe and I’m currently leading a cross Tribe initiative towards becoming really “App First” in our growth efforts.
Previously I was Growth Lead for the Growth Strategy squad, working on enabling Growth to prioritise efforts and to have the best growth metrics to measure success.
