Framework library · Innovation and product
Lean Startup
The Lean Startup treats a new venture as a set of beliefs to be tested rather than a plan to be executed. Each turn of the build-measure-learn loop builds just enough to test one hypothesis, measures what customers do, and ends in a decision: persevere, pivot to a new hypothesis, or stop. Progress is measured in what has been learned, not in what has been built.
Use it when
- You are launching something new to the business, where customer demand, price or the route to market is unknown.
- A venture is consuming money and the sponsor wants evidence of progress other than features shipped.
- The team disagrees about what customers want and needs a way to settle it with behaviour rather than argument.
- You can reach real customers quickly and cheaply enough to run several experiments before the main investment.
Avoid it when
- The product extends a business you already know well and the uncertainty is in execution, not demand. Use a business case and a project plan.
- Regulation, safety or network integrity rule out exposing customers to an unfinished product. Test with prototypes and partners instead.
- You want to test one specific assumption in depth. That is the job of a single minimum viable product. This register is for the whole set of hypotheses and the decisions they lead to.
- Nobody has authority to stop the venture. A register whose decisions cannot be acted on is theatre.
How to run it
Write the leap-of-faith hypotheses
Value hypotheses (customers will get value and pay) and growth hypotheses (new customers will find the product). Write each so a test could prove it wrong.
Order them by how much of the plan depends on them
Test first the belief that would sink the plan if wrong and has least evidence behind it.
Design an experiment and set the pass mark
Name the metric and the threshold before running anything, and phrase the metric so that higher is better. A target set after the result will always be met.
Run it and record what happened
Measure behaviour (sign-ups, payments, repeat use), not stated intent. Note the numbers and what customers said.
Decide to persevere, pivot or stop
Persevere if the evidence supports the hypothesis. Pivot to a new hypothesis if it points elsewhere, and stop if no credible hypothesis is left. Write down what was learned either way.
Review on a fixed rhythm
Hold a pivot-or-persevere meeting every month or two with the sponsor, using this register. Decisions that drift are decisions not made.
Work through it
Answer the questions below, or load the worked example to see a finished one. The drawing updates as you type. Export the result as a PowerPoint deck, a Word document, an Excel workbook, a PDF or plain text.
What you type stays in this browser, so you can close the page and come back to it. It is not sent to Blue Prysm or anyone else, and the exports are made here, on your device. Privacy policy.
Mistakes to avoid
- Measuring vanity metrics, such as page views or total sign-ups, that rise whatever happens. Use rates and cohorts that could fall.
- Running experiments without a pass mark, then reading every result as encouraging.
- Persevering after a failed test without saying why. The register should show the reasoning, or the method has been abandoned.
- Pivoting so often that no hypothesis is tested properly. A pivot changes one element of the plan, not all of it.
Where it comes from
Eric Ries coined the term on his blog Startup Lessons Learned in September 2008, building on Steve Blank's customer development method (The Four Steps to the Epiphany, 2005), and set out the method in The Lean Startup (Crown Business, 2011). Source.
Use it with
Further reading
Work through it with us
The frameworks here are free to use as they stand. If you would rather work through the question behind this one with us, these are the ways an engagement starts.