Assumptions Mapping

What it is

A team exercise for finding the beliefs a plan depends on and deciding which to test first. David J. Bland, founder of the consultancy Precoil, developed it and set it out in Testing Business Ideas (Wiley, 2019), co-written with Alexander Osterwalder. The team asks "what has to be true for this idea to work?" and writes each answer as a testable statement starting "We believe that...". Each is tagged by type of risk: desirability (does the market want it?), viability (can it make enough money?) and feasibility (can we deliver it?); Bland's 2020 Strategyzer excerpt of the book adds adaptability (can it survive a changing market?). Each statement is then placed on a 2x2 grid with importance on one axis (would the plan fail if this were wrong?) and evidence on the other (do we have observed proof, or only opinion?). The important assumptions with the least evidence, in the top-right corner, are tested first. Bland learned the 2x2 from Jeff Gothelf and Josh Seiden, the authors of Lean UX, and from Giff Constable, then added design thinking's risk types to it; Gothelf's own Hypothesis Prioritization Canvas is a related variant that plots hypotheses by risk and perceived value instead.

For a PMM, the method applies most directly to the go-to-market layer of a plan: who the buyer is, which message lands, what price they will pay and which channel reaches them. Each test is designed on a Strategyzer Test Card (hypothesis, test, metric, success criteria) and recorded on a Learning Card (observation, insight, decision). It does not replace Concept Testing or Message Testing; it decides which experiments to run, and in what order. It also differs from a Pre-mortem: a pre-mortem imagines the plan has already failed to surface risks, while Assumptions Mapping ranks existing beliefs by how much evidence supports them and designs a test for each risky one.

When to use it

• Entering a new segment, geography or buyer role, where the existing playbook's buyer, message and price may not carry over.
• A Tier 1 or Tier 2 launch is being planned around a new pricing model, channel or packaging, and the plan cites internal conviction rather than customer evidence.
• A Business Model Canvas or launch brief has been drafted and its customer-side blocks are still marked "assumed".
• The team is already running experiments but cannot say why these ones, such as a landing page test chosen because it was easy rather than because it addressed the biggest risk.
• Leadership asks for a go or no-go recommendation on an expansion bet and PMM needs to show what is proven, what is not, and what it would take to find out.

How to run it

1. Frame the bet in one sentence. For example: "Launch our scheduling product to UK dental clinics in Q3."
2. List the assumptions with the people closest to each risk. Run a 60 to 90 minute session with product, sales, finance and engineering. Have each person write assumptions silently first, one per sticky note, colour-coded by type. PMM should add the go-to-market assumptions the others tend to miss: who signs the contract, which pain the message should lead with, the price point and price metric, and which channel produces pipeline.
3. Rewrite each as a testable statement. Replace "clinics need this" with "We believe that practice managers, not owners, choose scheduling software."
4. Plot every statement on the importance and evidence grid. Ask of each one, "If this is wrong, does the plan fail?" and "What have we observed, not just heard, that supports it?" Opinions, including senior ones, count as no evidence.
5. Pick the top-right assumptions, usually three to five. These are important and unproven; test them before spending launch budget.
6. Design each test on a Test Card before it runs. State the hypothesis, the test, the metric and the success criterion, for example "at least 4% of landing page visitors join the waitlist". Choose the cheapest test that gives enough evidence: interviews, a landing page smoke test, a feature stub (the book's name for a button inside the product for a feature not yet built, which practitioners often call a fake door), a concierge service run by hand, or a pre-sale.
7. Prefer what people do over what they say. Testing Business Ideas grades evidence by strength: a signed pre-sale outweighs a click, and a click outweighs an interview answer. Where a weak test passes, follow it with a stronger one before committing budget.
8. Record each result on a Learning Card. Write what you observed, what you now believe, and the decision: persevere, change the plan or stop.
9. Re-plot the map. Move tested assumptions to the evidence side and add any new ones the tests raised, then pick the next top-right set.

Cadence & ownership

Bland describes Assumptions Mapping as a team exercise, and in practice whoever leads the bet facilitates it: a product manager or innovation team for a new product, a founder at a start-up. For a launch or segment entry, PMM commonly owns the go-to-market assumptions and runs the map outright, because PMM holds the segment, messaging, pricing and channel evidence and runs most of the cheap experiments that test them. The go or no-go decision stays with the launch's executive sponsor or the product lead. At a solo or founding-PMM stage, the PMM often maps the whole bet, desirability through viability, with the founder in a single session; at a scaled company with a product discovery practice, product owns the desirability and feasibility map and PMM contributes the go-to-market layer to it. Run the first session as soon as the bet is named, before launch planning locks; then re-plot every two to four weeks, or after each Learning Card decision, until the top-right corner is empty or the bet is stopped.