I’m Sam, co-founder of a start-up based in San Francisco. I’m VP product in charge of developing our new kick-ass B2C app.
It’s no surprise that we use the ‘Build-Measure-Learn’ iterative process, better known as ‘Lean Methodology’.
The fundamental activity of a startup is to turn ideas into products, measure how customers respond, and then learn whether to pivot or persevere. All successful startup processes should be geared to accelerate that feedback loop.
This way of thinking seems obvious. However, sadly, it’s not what we did in the past two years. Our clients (the guys with the checkbook) came in the way, twisting our product with their requirements and made us shifting priorities.
It was easy money and exciting stuff to develop, but at the end of the day, it brought no value to the company and made the app less appealing to the end user.
This time, we’ll do it right. The end user is our one and only focus. Period.
Start with very few testers
What if your own team doesn’t use your product? It means that your product sucks. This is why I think the first version (usually called a Proof Of Concept — or POC) shouldn’t go out of your office. Leverage your whole team. Everybody should have fun using it. Don’t take “I’m not the right target” for a valid answer. Be realistic: no fun = bad product.
Iterate super fast
POCs are prototypes. Don’t bother using the top-notch technologies. You will have 100 testers at most. Any crappy database will do the job. Choose a framework that you are confortable with. And just … code!
Track, speak, watch
Keep in mind that prototypes are only living to gather feedback. If you can see the WOW effect on your user face, that’s a big win, but generally, people use apps in their intimity. You want to make sure to have access to some key analytics to really understand their usages.
Speak with your users, but don’t take what they say for granted. There can be a huge difference in between their real usages and what they tell you.
I love to watch users using my product. It’s even better if they don’t know that I am watching them. To me, it is one of the best feedback you can gather.
Keep the momentum, improve and push releases quick
Usually, users are super excited by testing new stuff. Don’t break this momentum. Fix things quick and release often (every days for minor fixes, every 2 weeks for major versions). Be transparent on your daily roadmap.
Build your community in parallel
Testing your app with few users doesn’t prevent you to build your community of beta testers in parallel. Once you feel confortable with the product, you’ll be ready to broaden your circle. Mailbox, back in 2011, is a good example, with thousand of beta testers stacked up in a huge waiting list. It was super exciting to see the number of guys in front on you decreasing every day until you were able to, at last, test this awesome app and be one of the “privileged person”.
I hear too often that the User Interface (UI) of a prototype can be crappy because that’s not the point. That’s wrong! Testing the UI is part of the job. Same with the content. A prototype is not here just to test the core mechanic. The User Experience (UX) is a whole, from the layout to the design, the colors and the content. All these things should converge to one point: a Minimum Viable Product (MVP). Very few features, but perfectly executed. Think Apple. iPhone 1 with iOS1 didn’t even have copy/paste capabilities. But everything was so perfectly polished. Nobody complained. Now think Microsoft. At the time, Pocket PCs were out there with hundred of features poorly executed. Who won?
You have to make every single detail perfect. And you have to limit the number of details. — Jack Dorsey
First time experience
Last but not least, spend time on the First Time Experience (FTE). Throwing a on-boarding with 5 slides explaining what your app is all about is not the right solution. People don’t read. Instead, you should look at what famous games are doing. Don’t force the user to open an account before having tested your app, don’t ask them to invite their friends before knowing if they like your product. Common sense?