We won a Product Hunt hackathon

With Tinkso, we have been building digital products for clients for over three years. They make life easier for our clients and help improve their bottom line. Those seem decent indicators of making a successful product.

If we can do it for our clients, why not for ourselves? How hard can it be to build a successful product from the ground up? What would it cost and when do we make a profit, if at all?

In this series of articles, we’ll try and be as open as possible about our attempts to create a digital product that would provide at least enough revenue to cover the cost of building, marketing and maintaining it. In other words, we set the bar really low.

The opportunity: a no-code hackaton by Product Hunt

Being a web application design agency mostly working with low and no-code solutions (and primarily building apps using Bubble.is) we were busy with client work and never found the time to build our own ideas.

When Product Hunt announced a contest for apps developed using little or no code, we realized this was the perfect incentive to build an app and showcase the level and quality of no-code solutions nowadays.

The idea: an app that uses crowd sourcing to help make decisions

In design and copy writing you’ll often find yourself in situations where you’d like to validate an idea or a specific design. As a freelancer you don’t always have a team available to review different versions of a logo, illustration or icon. As a startup you would want to validate a company or domain name, figure out the best email subject line, header or call to action for your website. Getting feedback quickly is crucial in making the right decisions and keeping the design process moving. At least, that’s the idea…

So, we wanted to build an app that would make it easy for designers and entrepreneurs to get feedback from a large group of people instantly and at low cost. We decided we needed a crowdsourcing platform.

My partner Matt did research on Amazon’s Mechanical Turk while working at Xerox years ago and was intrigued by the concept ever since.

MTurk allows anyone to outsource a task to a very large group of people and get almost immediate results. You create and post a Human Intelligence Task (HIT) that could be anything from answering a simple survey to identifying objects in images or translating texts.

There is a bit of a moral issue with Mechanical Turk (and other crowd source platforms) and it’s something we acknowledge. Some HIT’s ask people to complete a 45 minute task for just $0.80 as this article rightfully points out as a very dangerous trend in the largely unregulated world of online crowdsourcing. You may think twice about becoming a Turker.

While some people rely on MTurk to make ends meet, this research (interestingly, the research was done as HIT on MTurk) shows that most MTurkers are young and for a large part student. Recognizing the low payout per HIT and the over representation of young but highly educated people, we figured that the type of HIT we had in mind was a perfect fit for the MTurk demographic.

This one or that one?

So, we would let people upload two illustrations, icons, logos, photos or write two texts. The Turkers give their opinion on both and select the best one. In the end, they have produced a ‘winner’. In other words; easy to setup A/B user tests at low cost that give immediate results. Who wouldn’t want that!?