Skip to content

Prototype versus Beta

Erik Larson Apr 5, 2013 10:44:20 AM

A few people have asked what we mean by 'prototype.' Why we don't just call it a 'beta' like so many startups do? Because it isn't. Here is a quick explanation.

Wikipedia's definition of a prototype is "an early sample or model built to test a concept or process or to act as a thing to be replicated or learned from." In our case, we built this prototype to test only a few of the most important features to see if they actually work. If it turned out it actually helps people with the big decisions they face, then we could go on to building a complete beta product.

The Cloverpop prototype works better than we expected after thorough testing, so we decided to leave it up for people to use while we move on to building the product. We're also still using it to research how well it helps people with their decisions in the longer term, too.

The Cloverpop prototype is actually the 27th prototype we've created. It took a month or two to build, test, iterate and then set aside each of those early prototypes. And we would have set the this prototype aside, too, if the results hadn't been good.

We are doing this following lean startup principles. Or at least the core of the philosophy, before it became obfuscated with some of the business-speak you will find at that link. The core is this: a startup exists to learn what customers want by building something and measuring what happens while investing as little as possible. Stay lean and mean so you can iterate faster and keep at it longer.

So a prototype is different from a beta because it is lacking many obvious and important features that people will need before they can use the product as intended, but it still has enough features to prove that it works. When the product idea is new and the customer problem is big, the prototype should have hardly any features at all. Why? Because the big question is whether people want what it does, not whether it does everything they want.

To test that question, we gave the Cloverpop prototype four features:

  1. A simple questionnaire that adapts slightly to each person based on their responses
  2. An algorithm that analyzes their responses and provides an answer using a large database of other decisions
  3. A personalized analysis and summary of the decision
  4. A simple forum with anonymous short stories of the decisions people have used it for

Curiously, Edison's lightbulb also had four features: a globe, a filament, a stem and a base.

We had a few hundred people use the prototype, measured how well it all worked statistically and subjectively, and iterated a half-dozen times while a few hundred more people used it. After a thousand or so people flowed through, we finally got to something that seems worth taking to the next stage and turning into a beta.

The beta will still have rough edges, but it will have more 'cool' and needed features that build on the prototype so it does most of what people expect it should. It will have a fun and interactive user experience with unique integrated collaboration features.  It will have an enhanced algorithm and bigger database to give more advice and give you predictions about how your decision will turn out.

Making all that happen will take more work and you can see why building a prototype first is so important. Why build all that if it turns out no one cares?

Are you and your team facing a important business decision at work? Give Cloverpop a try and let us know if you care.