Fail faster – the reason why you should try to crash early

Louise Fritjofsson

Louise Fritjofsson


I know it's hard to try to put your beautiful idea up for failure. But it has to be done – sooner or later.

Fingers are tingling. You’re barely sleeping at night, and you never feel tired. You talk faster than normal and feel frustrated things are not moving as fast as you want. It’s brilliant, your idea. It’s your baby.

Recognize that feeling?

92% out of start-ups crash and burn within 3 years after launch. If you’re an entrepreneur these numbers will not frighten you. You see, you probably lack ability to understand the brutal reality that the odds are against you. By a lot.

This means we spend a lot of time on things that don’t work. And that’s OK. We need to be willing to fail a 1,000 times to find the one thing that truly works.

But one thing that we can do is to become better at testing our ideas as early as possible to understand which ones have a future and which ones are simply worthless. I simply want you to MVP the shit out of everything you do.

I bet you’ve heard of it: Minimal Viable Product. It means creating the least thing of a product or feature to test if your thesis about it. Some immediate problems to consider:

  • the MVP is often far from the product you want to offer
  • People are often not as technical as yourself, so make sure to build something for your mom to make sure you’re not creating something too difficult to understand in the first place

Maybe you're already doing it. But I want you to challenge that thinking. I was a part of this trap. I thought I MVP:d everything – and I did – but only to a certain extent. It’s hard to try to put your beautiful idea up for failure. It’s like killing your baby before it’s even born.

I wish I had gone further in validating ideas in several occasions looking back. We spent a lot of time building features in Vint that later had little or no impact. We could, and should, have been more aggressive in testing different takes on our product overall. I’ve been thinking about this a lot and as I swore to become better at really setting my ideas up for the fire-test before leaping out, I’ve now created a check-list to help me be brave enough to take it all the way. As I hope the list can be of use for others I wanted to share this list with you.

Before starting, make sure to have your objectives clear. Are you looking to make a consumer happy? Are you looking to see how fast you can build traction? Is retention important? Is the consumer supposed to create brand recognition? How fast is the consumer supposed to understand your value proposition? Are you supposed to offer customer service to make this work? Think twice about the questions you need to add on this list to create a MVP test that will give you a honest answer to your questions.

Now, KingLui’s quick & dirty guide to MVP testing:

1. Don’t do all the hard work 

Evaluate if you can test your thesis on already existing technology. Do everything you can to save money and time without compromising the test. Companies like Squarespace, Shoppify, Tictail and more is making it easy for us to get something out quick and without heavy investments #win

2. Collect data as efficiently as you can

Create a landing page or a few to start collecting data on how the users are behaving. Remember to build enough to collect data, but not more than that. It’s completely OK that buttons, filters and “buy now” actions don’t work – as long as you can measure the data from these attempts. 

2.1. Please remember to use this to upgrade your current products by adding buttons and action triggers to measure interest before building up entirely new features.

2.2. Drive traffic to your landing page through email campaigns. Get enough volume to dare to A/B test different copy and/or offers. Remember to measure clicks, time and re-opens to use this as a source of data as well. 

3. Sell it before it’s a product

Think you have something brilliant? Prove it! Starts with the sales process to make sure people/companies are ready to pay for it. Got clients that’s ready to invest and can’t provide the product? Not a problem – you have the result you needed, and you would of course never use the most important prospects for the MVP trials.  

4. Be careful 

Analyze your data carefully and make sure you have enough scale in the tests to make sure they are pointing you in the right direction. Remember that people wanted a faster horse and not a car – so talk to potential clients with respect that they have no idea what they want.

5. Ask the hard questions

When talking to people about your product, make sure you really get to the core of their ideas and suggestions.

- why is this needed

- what problem does it solve

- how would your behavior change if we solved this problem for you

6. Protect your brand

Afraid to build landing pages that don’t work or sell something you can’t deliver? Simply use another company name and problem solved.

Remember that an MVP is something completely different than a product everyone loves, kudos and thanks to brilliant Daniel Swenson @Blogglovin for bringing that truth to my attention.

Even with MVP testing you may still fail. It’s hard to understand all levels of data and analyze this correctly but I can assure you at least you will know if it’s a proper shit you have as an idea. From there, you can continue to tweak your tests or simply come up with something better to spend your time on. Horray!


Läs mer