Nancy Hua, CTO and co-founder of Apptimize, looks at the lessons to be learned from Instagram’s short-lived tap-to-advance navigation feature launch.

Shortly before the new year, Instagram unintentionally rolled out tap-to-advance navigation to app users. Shortly thereafter, Instagram rolled back the change with an Instagram spokesperson saying that the feature update was intended to be a small test that then went much broader than expected.

While consumer reaction to the change was swift, I dont think Instagram did anything bad in this instance. Sure, maybe they had a mistake in how they configured the test versus their intentions, but its not the big deal it’s been made out to be thus far. Every change causes some users to be unhappy, but resisting change is futile. Instagram is a great product that drives change and innovation instead of being reactive, which raises my opinion of them.

There can be hiccups is the testing process, but the fact that theyre trying new features and testing out their changes demonstrates the importance of testing and iteration for maintaining quality products. Product development is change and thats why its hard. Instagram is an important app in consumers lives and feature flagging, staged rollouts, and A/B testing are state-of-the-art strategies that products like Instagram rely on to innovate, maintain the platform and hopefully grow their user base.

In this tap-to-advance feature test, Instagram seemingly made a user error in their rollout configuration versus what was planned. The company then learned that a proportion (who knows how large) of users did not like the change. Without an eye on the data from the test, for all we know, the vast majority of users in the test may have liked the feature a lot and are just less vocal about it on Twitter. To truly understand the impact of the change, Instagram would need to run the test for a longer period and measure engagement in the long term.

Tests generally need to have holdout groups that show how the changes youre testing affect your users over a longer period of time. A specific change and the reaction to it is not something companies should generally worry about. Though faced with blowback in the moment, other brands should recognize that these tests mean Instagram is executing on current best practices in how product development needs to work now.

This situation would have been worse if Instagram had rolled out the change to everyone immediately without understanding the impact of what they were doing. However, an app as sophisticated as Instagram would never do that. The practice of rolling out any change to your entire user base at once is an insane and unnecessary risk in todays world, when A/B testing is an option.

Any change will affect some users in an established user base negatively, so its important for companies to understand what these people are reacting to so they can learn more about their users. Maybe theres something the company can do to improve the feature for those users, or maybe the users can teach the company something about a different use case.

Marketers need to understand how to break down their audience and measure the long-term revenue and engagement implications of any idea. Companies need to A/B test every new feature and roll it out slowly and strategically to receptive audiences (for example, new users) to avoid consumer blowback, while letting power users learn how to interact with the new features. As brands learn more about their consumers through testing, they can come up with more sophisticated segmentations down the road.