Ideas

How to pressure-test an idea before you commit to it

Most ideas don't fail because they were bad. They fail because one thing everybody assumed turned out not to be true. Here's how to find that thing first.

By the Offo teamUpdated 6 min read

The short answer

To pressure-test an idea, write down everything that has to be true for it to work, find the assumption that matters most and has the least evidence behind it, and test that one first, as cheaply as you can. Run a pre-mortem by imagining the idea has already failed and listing why. And when you ask people for feedback, ask about what they've done, not what they say they'd do.

  • List what has to be true.
  • Find the riskiest assumption: important and unproven.
  • Imagine it failed, and write down why.
  • Test cheaply before you spend.

Why ideas fail

Ideas rarely fail for lack of features, effort or polish. They fail on an assumption: that people have this problem, that they'd pay to solve it, that you could reach them, that it could be done for the money. Each one felt obvious at the start. Pressure-testing is how you find out which obvious thing isn't true before you've spent a year and your savings on it.

It works for anything with a bet at its heart: a business, a side project, a new job, a community event, a feature at work.

Step 1: Write down what has to be true

Take a page and finish the sentence “This works if…” as many times as you can. It helps to think in three groups, a lens popularised by the design firm IDEO:

  • Desirable: people want it. They have the problem, they notice it, and they care enough to act.
  • Viable: it adds up. It pays for itself, or earns enough to be worth your time.
  • Feasible: you can do it. You have, or can get, the skills, time, money and permissions.

Most people list five or six assumptions and stop. Push for fifteen. The dangerous ones tend to be the ones that feel too obvious to write down.

Step 2: Find your riskiest assumption

Not all assumptions are equal. Sort them on two questions: how important is it (if it's wrong, does the whole idea fall over?), and how much evidence do you have (have you seen it, or just believed it?).

Little evidenceStrong evidence
Critical to the ideaTest this firstKeep an eye on it
Nice to havePark it for laterDon't worry about it

Your riskiest assumption sits in the top-left corner: critical and unproven. That's the one to test first. Everything else can wait, because if it's wrong, nothing else matters.

Step 3: Run a pre-mortem

A post-mortem asks why something died. A pre-mortem, a technique the psychologist Gary Klein described in the Harvard Business Review (Klein, 2007), asks the same question before you start.

Imagine it's a year from now and the idea has failed. Not wobbled: failed. Now write the story of why. Give yourself five minutes and list every reason you can think of.

It works because of a quirk in how we think. Klein cites research on “prospective hindsight” which found that imagining an event has already happened increases people's ability to correctly identify reasons for future outcomes by 30% (Mitchell, Russo & Pennington, 1989). “What could go wrong?” invites optimism. “It went wrong: why?” invites explanations.

Step 4: Argue the other side

For centuries, the Catholic Church appointed an official whose job was to argue against making someone a saint: the original devil's advocate. You can do the same for your own idea.

  • Steelman the objection. Write the strongest case against your idea, the one a smart critic would make, not a straw man.
  • Flip it. Ask what would have to be true for the opposite approach to be right.
  • Find the uncomfortable question. The one you'd rather nobody asked. That's usually the one to answer.

Step 5: Test it cheaply

Once you know your riskiest assumption, design the cheapest test that could prove it wrong.

  • Ask about the past, not the future. Rob Fitzpatrick's The Mom Test (2013) makes the case well: people will say they love your idea to be kind. Ask what they did the last time they had the problem, and what it cost them.
  • Do it by hand first. Before building anything, deliver the thing manually for a few people. You'll learn more in a week than in a month of planning.
  • Pre-sell it. A one-page description and a way to pay, or to put a name down, tests desire better than any survey.
  • Run a small version. One evening, one street, one class, one product. Small enough to fail cheaply.
  • Set the pass mark in advance. Decide what counts as success before you look. Twenty sign-ups? Five paying customers? Otherwise every result will look encouraging.

Your pageHouseplant subscription box?
£15 a month, rare plants, care cards

Offo writes backHere's where I'd poke: people who love rare houseplants tend to love choosing them, and a box takes the choosing away. That's the shaky bit. Cheap test: post the three plants you'd send next month and ask who'd pre-order at £15. If the replies are all “can I just have that one?”, you might have a shop, not a subscription.

An illustrative example of asking Offo to push back: it names the shakiest assumption, and a cheap way to find out.

Twelve questions that pressure-test any idea

People

  1. Who, specifically, has this problem? Can you name three of them?
  2. What do they do about it today?
  3. What would make them switch?

Money and time

  1. What does it cost to deliver, including your time?
  2. Who pays, and when?
  3. What's the least you could spend to find out if it works?

Risk

  1. What has to be true that you haven't checked?
  2. If it fails in a year, what's the most likely reason?
  3. What would make you stop?

Momentum

  1. What's the smallest version worth doing?
  2. What could you test this week?
  3. Who could you ask who'd tell you the truth?

Getting honest pushback

The hardest part of pressure-testing your own idea is that you want it to work. Friends are too kind, strangers are too busy, and your own brain is on the idea's side.

An AI can be a useful sparring partner here, as long as you ask it to push back rather than cheer. Ask for the weakest assumption and a cheap way to test it, not for reassurance. (Our guide to using AI as a thinking partner has more on asking the right way.)

That's one of the things Offo, a notebook for iPad that writes back, is built for. Sketch or write your idea on the page and ask Offo to push back: it names the assumption most likely to be wrong, and a cheap way to find out, in a few handwritten lines you can argue with.

Questions people ask

How do I know if my business idea is good?

You can't know in advance, but you can find out cheaply. List what has to be true for it to work, find the most important and least proven assumption, and test it with real behaviour, such as pre-orders or sign-ups, rather than opinions.

What is a pre-mortem?

A pre-mortem is a planning exercise in which you imagine a project has already failed and list the reasons why. Described by the psychologist Gary Klein in 2007, it surfaces risks that people are reluctant to raise or wouldn't otherwise think of.

What is a riskiest assumption?

It's the belief your idea depends on most that you have the least evidence for. If it's wrong, the idea fails, so it's the one to test first.

How can I test an idea without spending money?

Talk to potential customers about what they've done rather than what they would do, deliver the service by hand for a few people, or describe it on one page and ask people to pre-order or sign up.

Sources

  1. Klein, G. (2007). Performing a project premortem. Harvard Business Review, September 2007. hbr.org/2007/09/performing-a-project-premortem
  2. Mitchell, D. J., Russo, J. E. & Pennington, N. (1989). Back to the future: temporal perspective in the explanation of events. Journal of Behavioral Decision Making, 2(1), 25–38. doi.org/10.1002/bdm.3960020103
  3. Fitzpatrick, R. (2013). The Mom Test: How to Talk to Customers and Learn If Your Business Is a Good Idea When Everyone Is Lying to You.

Get pushback before it costs you

Offo is a notebook for iPad and Apple Pencil. Sketch your idea, ask Offo to push back, and get the shaky bit named in a few handwritten lines.