Base product decisions on user actions…not feedback

We have all seen what happens any time a new version of software gets rolled out. The most obvious examples are when Facebook or Twitter do anything to their platforms, their users squawk and threaten to quit etc.. And then do it all over again the next time they push any significant change or redesign. This is human nature.

Users figure out how to use your app and then the automatic part of your brain takes over. By changing the app (hopefully for the better!) you force your existing learners to relearn how to use your app. We’ve all experienced this over the years, hello iOS music app…

Some users are happy for something new but many just want to keep doing what they are used to doing. After all, they are not here to learn how to use your app, they are here to use your app to perform a function. Changing the way the app works just gets in the way.

This video is not entirely relevant but still has a lot of truth to it and always makes me laugh anyway!

When we pushed our new LingQ Reader 4.0 three weeks ago, we expected the same reaction. We have seen it every time we have made significant changes to the app, and we saw it here again. However, this has been without question the smoothest upgrade we have ever done.

There are likely a few reasons for this. First of all, we feel like we really did push a far superior product. Of course, we always felt like our previous versions were good too so that may just be hindsight saying this. But, we have also learned a lot over the years and the product has evolved and gotten significantly better over this time. Users will put up with learning new behaviour as long as they see an obvious benefit.

Secondly, we took our time with our release. We had some of our members beta testing the new version for months before release. We were able to respond to many concerns and issues long before the update hit the bulk of our users. These included both bug fixes and adjustments to functionality.

Finally, we left our “Classic Reader” in place so that existing users could continue to use it if they preferred.

The majority of feedback we have received has been very positive. Of course, we had a few users who were not happy and continue to be unhappy. But, that is inevitable when doing something like this.

The biggest complaints centred around our move from a scrolling reader to a paged reader interface, more like an e-book reader. Users who have always scrolled became frustrated when they could no longer do so. Here is what I posted on our Forum in response to this complaint:

I understand why you feel this attachment to classic mode. You’ve enjoyed using it and feel like it’s an old friend. That is always the way when tools that you are using change in some way and you have to learn new behaviour to do what you want to do. You find yourself wanting to scroll because that is what you are used to and when you can’t you get frustrated. I, too, found myself wanting to scroll initially and experienced that feeling of frustration when I couldn’t.
The numbers so far, bear us out. In the last 3 days 91% of all LingQs have been created in the new reader. Those are mostly existing members who prefer the new format as there are only 3 days worth of new members. There has also been a 123% increase in LingQs created over the preceding 3 days. That is a massive increase. Yes, it may be excitement over our new release and it is undoubtedly a small sample size but those are the numbers that will drive our decisions.

It’s now been 3 weeks since we launched our new reader and we continue to see significant improvement in site usage since launch. That was the major driver behind most of the changes we made. We know that if we can drive more usage, we will drive more paid conversions and we will retain our paying members longer. So far, this is proving to be the case.

We continue to see that 82% of LingQs are created in the new reader. And users are twice as active as they were before our update. Feedback we have received for the new reader has been overwhelmingly positive and we are seeing significant improvements in paid upgrades and reduced churn.

One thing we have learned over the years is that while you have to talk to your users and find out their pain points, their response to change is not a realistic barometer for measuring the impact of changes. The only way to accurately judge the effects of any changes are by measuring key metrics and comparing the results. In the past, we have spent too much time trying to please too many people which has lead to feature creep and lack of focus.

By being more focused, we should be able to make sure that the functionality we do provide is built and maintained to a higher standard. We continue to hear from those users both for and against our update and we value all their feedback even if we don’t always act on all of it. We have a lot more big improvements coming soon that should make things even better. With our new approach to pushing updates, I am looking forward to our further changes and confident that we will be making informed decisions to drive our project forward.

— — — — — — — — —

At LingQ, you can learn languages all the way to fluency by following your interests and interacting with our global community. Choose from 14 different languages on web and mobile.