Carry out A/B tests on smartphone apps with Optimimo

We at YourStory have been closely associated with the app developer community in India. One issue that plagues almost everyone in the app developer in the community is deciding upon which feature to stick with and which ones to do away with or what colour suites their app better or which button placement is better. One way in which they go about making this decision is asking people and some of their initial beta users. The method is flawed for two main reasons - Firstly, there are only so many people whom you can reach out to and secondly, a feature takes time to get used to and a first impression is almost always wrong. This method doesn't generate hard, actionable data.

Shimply, an Indian app company has launched Optimimo which helps app developers perform A/B tests across various variants of their apps. By adding a .jar file to their apps, App developers can then log into Optimimo's back end and find out how variants of their app performs with different audiences.

Shimply's CEO Rajat Garg, said “We built a mobile app for our product and realized the pain. It was slow to figure out what is working and what just did not work! This product was our own internal creation to solve that problem. When we researched, we found this to be a larger issue and decided to make this product available to fellow mobile app developers at a very affordable rate."

For serious app developers only

Having worked with a fairly large app developer community, we are aware that a large majority of this population are individual developers trying to build apps in their free time or are startups crunched for resources. To use a tool like Optimimo, the experiments A and B need to be defined by the app developer in the first place. This means that an app developer needs to build different variants of the application and a tool like Optimimo will help in testing it on specific target groups based on OS versions, screen sizes etc. Furthermore, the Optimimo can be used for up to 10 different variation. However, to use this, there need to be 10 different variations of the app in the first place, which most developers and even some startups, don't do.

Rajat shared, "Let me put it this way - anyone who's serious about app development can use Optimimo. Most web based products have many variations which they try out on various test groups. Executing such methods for a smartphone app used to be often random and to a relatively small test group. With Optimimo you can get more real time data from your apps' regualr users."

To me, the ability to carry out specific and defined experiments with Optimimo is of great value for anyone trying to build a smartphone app. Decisions made from inferring the data generated by Optimimo will be much more accurate and correct. Rajat shared that they will be adding features like heat maps which will help in making better decisions on which feature to keep and what to do away with.

On a parting note, I asked Rajat what he thought about the Google Play Store's new "become a tester" feature. He says, "It will be used mainly by a niche and savvy users of the app, which is more often than not your regular app user. With Optimimo, we are striving to provide real time data on changes from your app's every day users."

For now, it surely looks like this is a comprehensive and systematic solution to get user feedback on UI and interaction changes for a smartphone app. It will be interesting to see how users adopt Optimimo.

If you're an app developer, check out Optimimo and tell us what you think.