When we first looked at the numbers behind one of our redesigned features at Any.do, the first thing we said was (raise your hand if it’s sounds familiar):
“There’s probably a bug in the analytics”
It’s gotta be a bug right? In some cases, it is, but in this case, it wasn’t.
There’s a saying in the startup world that 90% of experiments are bound to fail. Luckily at Any.do, we are way below that threshold. But I would like to take the next few minutes of your time and explain more about that failure of mine. It’s one failure I took to the heart because of the time it required to design, iterate, develop and test. It’s one failure that occasionally haunts me because UX wise, “it’s such a well-crafted product” (a quote from one of the leaders in our design community).
A brief history:
The feature we redesigned was Any.do Moment. Released in 2013, Any.do Moment was aimed to increase productivity & engagement among our customers by allowing them to review their tasks for today and then decide if they would like to tackle it now, later or whenever:
Feedback from V1 of Any.do Moment was fairly positive. It was a fresh attempt to address a known problem, baked within the Any.do minimalistic design language.
But there were problems we found in our research leading to V2, these were the main ones:
- Data showed customers interacted with Moment a few times a month. The main goal of this product was to increase engagement to a level of a few times per week.
- Customers had no idea what they have scheduled in their day when they postpone tasks for later.
- Customers had to make a decision regarding all of their tasks, they couldn’t just skip a task they were not interested in tackling .
- Customers who used the app to jot up just about anything saw they needed to make decisions regarding tomatoes and onions in their shopping list.
- The interface didn’t support more than 30 tasks & was designed for 4-inch screens. It didn’t utilize the big canvas larger phones offer today.
- With all the fireworks & animations, the interface was pretty slow, and you needed a good 3–4 seconds until you can actually interact with the product.
Designing Moment 2.0
Following the success of Any.do’s calendar tab which allowed customers to view their tasks inside their calendar, we decided to integrate it into our redesign and implement the rest of the feedback we got from our customers.
And here comes the bride:
When we did user testings with an advanced prototype, potential users loved Moment 2, while existing users loved the motivational messages during the experience that stayed loyal to the original product.
We addressed the main issues
Customers had the ability to scan their backlog of tasks and magically drag and drop their tasks within their daily agenda for the perfect A-HA moment. They also had the ability to adjust the timeframe required for a specific task — another A-HA moment.
After a successful beta, we released Moment 2.0, and I was like:
Data showed Moment 2.0 was a failure. It wasn’t a catastrophic failure: The numbers were pretty on par with Moment 1. Even better: customers who had calendar events were better engaged with Moment 2 compared to Moment 1. But from an analytical point of view — it didn’t perform significantly better.
So what went wrong?
The first rule of building a valuable product is that it should be valuable for someone. Moment 2 did that, but the main problem was the majority of our user base didn’t have many events in their daily agenda or weren’t calendar oriented at all.
Thus, 70% of the canvas was pretty irrelevant to them. Dragging and dropping tasks in your agenda is awesome, but dragging them between your busy day is far better and rewarding. Many users missed that golden feeling.
Moment 2 also featured bugs that I felt were harming the core experience, but when deciding what’s next on our plate, we decided not to invest further into fixes that can increase numbers by a little but won’t change the big picture.
I have been designing digital products for 7 years. Luckily for me, most of them succeeded tremendously. But some of them did fail. For whatever reason, I was emotionally attached to this feature, and it wasn’t even the main product. It’s one of those products that get you really excited about, and that’s one of the essences of our profession.
Learning from failures is a critical aspect of product design. Moment 2 on mobile was buried, but it will return in a different form factor, this time on the large canvas of the web. Stay tuned.