About 4 years ago XING did a huge relaunch that included a revamped navigation. The concept split the the navigation into a horizontal product area on top of the screen and a vertical product notification area. If a user clicked one of the notification icons, a layer with a list of notifications showed up. So people could look into their product specific notifications layer without leaving their actual task / page. Quick access was the key idea.
Of course, we did usability testing with prototypes. A lot of them. And we’ve got good feedback from most of our users. Not all were happy -of course- but we managed to address the most critical usability issues. And we needed a little courage, since you cannot plan really everything beforehand. But being courageous is one of our guiding principles at XING. As an agile company we wanted to iterate, fix and improve things after initial launch anyways.
After launch, most of the product teams tracked if people click notifications in these layers. In the end everything worked, waves calmed down, we got no overly critical feedback anymore. And people just got used to how XING works.
But you know, just because numbers show that people click things, it doesn’t mean that it makes sense to them. Just because people click on an item, it doesn’t mean your product KPIs will suffer if you eventually remove it, if things get less complex, easier to understand. Just because you have to move forward, it doesn’t help if you not improve on existing things.
I got used to how things behave as well. But a while ago I started digging a little deeper. I had the feeling that we just went too far with our German engineering, ahem. The whole architecture was not only too complex, it suggested an even more complex product than it really was. We had dropdowns, layers, tabs, horizontal and vertical navigation items. Just too many things to understand. And guess what: with actual user feedback, further usability tests with real user content and looking into data from a more holistic angle, it was clear that we needed to change things. There was a huge potential for improving the user experience, as well as for having positive impact on the business side of things.
In June 2015 we launched a new navigation concept that is less complex and therefore easier to understand, clean and better looking (I think). It scales better and last but not least, it serves business needs better. We even got mostly positive feedback from our users (or no feedback which is often kind of positive, ahem).
You might ask why we did not recognize it in our usability tests. Or that we could have fixed it earlier. We could, we have, maybe you are right. And therefore I would give these suggestions, some well known, we proved again as true:
- You have to iterate. Do not fire and forget. (You know that. Are you embracing it, really?)
- Clear ownership of cross-functional UI elements helps to improve and move things forward. (If you do not have clear ownership defined, nobody owns it really and nothing will change naturally.)
- The longer you wait, the harder it gets to get buy-in from key decision makers. (They just get used to it and there is always something more important to do.)
- Look into data from a holistic point of view and not data siloed in one product. (Mind the local max.)
- Numbers are just numbers. You can interpret them in so many ways, that they can easily mislead you. (Or you can use them wisely to mislead others and stop creative, valuable discussions.)
- Evergreen: Keep it simple!