Grover Abhimanyu

Measuring which features 'stick'? or 'A/B' testing for features within application

Discussion created by Grover Abhimanyu on Feb 7, 2015

Hi, I've been developing web applications from a long time now, some are huge success and some are bust. But I've never thought of successful applications as having made several right decisions and few wrong ones, and vice-versa. So from last few weeks I'm exploring the new ideas of measuring what makes an application success. Is it possible to increase odds of success if we make continuous right decisions based on data? like:

  • which interface version is better, a or b?
  • is 'feature 1' used and liked by more people when 'interface 1' is used, and not with 'interface 2'? etc.

Of course at the end of day a lot would depend on your sales and marketing, but I'm only pointing out to increase the application's stick-ability to the end users.

I want to explore, if anyone's had a success story with such kind of system? Has anyone thought about trying it? What are the challenges of putting such system in place? Disadvantages? Is it worth? too complex? impossible?

Whatever you have to say about it - please share.

Outcomes