Business assumptions map: choose the uncertainty to investigate first

A business assumptions map separates the beliefs behind an offer by their importance and supporting evidence. Use this guide to write five distinct assumptions, show what you actually know and choose one uncertainty to investigate first. The output is a prioritised worksheet, not a declaration that your business idea has been validated.

Content last updated: 21 October 2026.

What decision does a business assumptions map support?

A team can be busy testing an idea without learning about the uncertainty that matters most. Somebody wants a landing page, somebody proposes interviews and somebody starts comparing software. Before choosing any activity, identify which unsupported belief could make the offer unworkable. This worksheet helps choose that question; it does not conduct the experiment.

An assumption is something you currently take to be true without enough evidence to rely on it. Some assumptions concern customer demand. Others concern delivery, access to resources or whether the proposed work fits your operating constraints. Keeping them separate prevents evidence about one issue from being used to claim that everything else is settled.

Strategyzer describes mapping assumptions by importance and evidence. Its guidance on strong hypotheses adds that a useful claim should be testable, precise and focused on one thing. Those are the source concepts behind this exercise. The worksheet, fictional example and checking procedure below are an original practical adaptation, not a report of performed client work or an official certification method.

Choose one offer and one intended customer situation. If your document covers several unrelated services, split them before mapping. A business proposal may help define the offer’s boundaries. Today’s task is smaller than writing the proposal: expose the beliefs you would otherwise carry into it unnoticed.

Turn a confident pitch into five separate claims

Consider a hypothetical independent analyst offering a monthly inventory review to small retailers. The service would use a retailer’s existing stock records to prepare a short exception report. This is a constructed exercise. No retailer has been interviewed, no sample has been measured and no sales result is implied by the example.

The sentence “Retailers will want this, share their records and pay enough to make it worthwhile” hides several different claims. A positive response to the idea would not establish access to usable records. A usable spreadsheet would not establish willingness to buy. Write each issue separately so later evidence can be attached to the claim it actually addresses.

1. The intended retailers experience a recurring problem that an inventory review could address.
2. A reachable person at the retailer can authorise this type of service.
3. The retailer can provide the required records in the agreed format.
4. The analyst can produce the promised report from those records.
5. The proposed workload fits the delivery time allowed for the offer.

These are starting claims rather than finished experiment rules. Refine ambiguous words such as recurring, required and allowed using your actual offer. If the relevant boundary is unknown, record that gap instead of inventing a precise number. The purpose of precision is shared understanding, not making an uncertain claim look more scientific than it is.

Do not force every business into these exact five rows. Your critical belief might concern distribution access, a partner’s availability or a particular customer workflow. Keep the same discipline: one row should contain one claim whose evidence can be examined separately.

Judge importance by the consequence of being wrong

For each claim, ask whether the offer could continue in essentially the same form if the claim were false. A belief that supports the whole delivery route deserves more attention than a detail that can be changed without altering the offer. Importance is relative to this proposal, not a permanent label attached to the subject.

In the hypothetical review service, access to suitable stock records is fundamental to preparing the report. The colour of the internal worksheet is probably not. That comparison does not mean presentation never matters. It means the current proposal relies on one condition more directly than the other. Another offer might produce a different ordering.

Write a short consequence next to every importance rating. “High, because no suitable records means no agreed report” is more informative than “high, because essential.” A colleague can now challenge the reasoning or identify an alternative delivery route. Without the reason, the rating is merely a preference expressed in a box.

Use the broader business risk plan for hazards such as equipment failure or account loss. This map is not a replacement for that plan. It highlights unsupported conditions behind a particular offer before you choose the next learning activity.

Record the evidence and what it cannot establish

Replace a vague evidence label with an identifiable observation or document. Include its date and relevant context. If somebody has previously supplied a stock file, note which file, under what conditions and for which task. That record might support access to a file without demonstrating that the monthly review can be delivered from it.

If you have no supporting material, write none yet. Do not manufacture interview notes or sample results to complete the exercise. An empty evidence field is a useful discovery when the business has been treating a belief as a fact. The worksheet should expose that gap rather than conceal it behind confident language.

Partial evidence deserves a partial description. A retailer’s comment that the idea sounds useful is a comment about the proposal. It does not establish purchase, permission to share records or repeated use. Similarly, an analyst’s experience with a different spreadsheet may be relevant background without proving that the new customer’s data will support the promised output.

Keep a field titled “not established.” For example: “Existing sample shows the column names; it does not show whether recent records are complete.” This makes the next question easier to write and reduces the chance that the same evidence will be stretched across several claims.

The business idea validation guide goes further into conversations, offers and response tracking. Do not repeat that entire process here. Your present result is an evidence gap linked to a specific assumption, ready to inform the choice of a later test.

Place the claims and select one investigation candidate

Draw a simple grid. One dimension separates higher from lower importance. The other separates weak or absent evidence from more relevant supporting evidence. Place each assumption using the notes you have just written. These labels are a temporary comparison, not a calculated probability of commercial success.

If two assumptions occupy the higher-importance, weak-evidence area, write why you choose one first. A dependency can help: you may need to establish access to a representative file before examining whether the report can be produced from it. That is a proposed order of work, not a finding that either assumption is true.

Offer and intended customer:
Assumption ID and one clear claim:
Consequence if the claim is false:
Importance and reason:
Existing evidence reference and date:
What that evidence does not establish:
Position in the map:
Selected for next investigation: yes / no
Selection reason:
Next question to answer:
Person responsible:

Give every assumption a stable reference such as A-01. If you substantially change a claim later, preserve the earlier wording in a dated note. Otherwise, evidence collected for the old claim may appear to support a new and different one. A stable ID helps only when its meaning remains traceable.

Save a useful first map in fifteen minutes

On Windows, open Notepad from Start and create a new document. On Mac, open TextEdit from Finder → Applications, choose File → New and, if needed, Format → Make Plain Text. Write the offer in one sentence, separate five claims and add an importance reason to each. Attach real evidence references where available and mark the remaining gaps. Place the claims in the grid, then choose one next investigation question. No advertising account or paid experiment is required to complete this preparation.

Choose File → Save As in Windows Notepad or File → Save in Mac TextEdit. Save business-assumptions-map-v1.txt in Documents; add a date if that name exists instead of overwriting it. Close the note, then reopen that exact file from Documents in File Explorer or Finder. Check all five IDs and map positions, the selected investigation candidate, evidence limits and unanswered question. Fifteen minutes is a planning target with available inputs, not observed performance.

Read the result for misleading conclusions. Does selected sound like proven? Change it to selected for investigation. Does no evidence sound like disproved? State that the claim remains uncertain. The distinction matters because an unsupported belief is not automatically false, and an important belief is not automatically worth an expensive test.

Connect later work to your SWOT analysis or digital roadmap when relevant. Keep today’s completion condition narrow: one offer, five separate claims, visible evidence limits and one justified investigation candidate. The map should make the next question clearer, not create the appearance of validation without any new observation.