How to choose between customer segments in a Lean Canvas

A quotation tool could serve freelance designers, small agencies and renovation contractors. All three send proposals. Each group can describe a frustrating afternoon spent preparing one. Putting all three into the same customer box, however, leaves the founder with conflicting demands about pricing, approvals and the information a quote must contain.

This guide uses an explicitly fictional quotation tool to compare three candidate groups. The numbers are teaching examples, not market research or GenerateBizPlan customer results. You can replace them with your own observations without turning a guess into a fact.

Define the decision before describing the customer

Start with the decision you need to make. For this example, the founder wants to choose one group for a small trial of a tool that creates proposals from reusable service descriptions. The founder can support standard service packages but cannot yet calculate construction materials or manage approval chains.

For a definition of the block, use our customer segments glossary. Here the task is narrower: choosing between plausible groups once you understand what a segment is. Name the chosen group, supporting evidence and conditions for reconsidering.

Write candidate segments at the same level of detail

Compare groups that can be observed and approached in similar ways. 'Freelancers' beside 'five-person agencies selling monthly website maintenance' creates an uneven comparison. The detailed option will appear stronger simply because somebody spent longer describing it.

Use four fields for each candidate: business situation, recurring task, current workaround and buying role. Confirm assumptions in conversations.

Separate the user, buyer and early adopter

In the fictional agency, an account manager prepares proposals, a director approves prices and the owner approves a subscription. One company can therefore contain several people whose answers matter. Record the role attached to each observation. A user who likes a prototype cannot necessarily authorise payment.

Ash Maurya, the creator of Lean Canvas, recommends identifying paying customers separately from other participants and showing their perspectives in the model. His guidance also distinguishes the wider customer segment from the early adopter subset. These distinctions are explained in his Lean Canvas guide.

For our comparison, the candidate groups are alternative starting markets. Within each group, early adopters are the people most likely to try the first version. An agency owner with an urgent proposal backlog could qualify; the owner of an otherwise similar agency with no current backlog might not.

Keep linked participants on the same canvas when they belong to the same buying arrangement. Separate canvases help when you are comparing genuinely different offers or business models. Splitting the user and payer automatically would hide the relationship you need to understand.

Collect comparable evidence for each candidate

Write the same questions for all three groups before reviewing their answers. Ask for a recent example of the task, the document or process used, the people involved and the consequences of delay. Then examine whether the proposed first version could handle that specific example.

Record existing alternatives in the customer's language. 'Word document copied from the previous project' contains more useful information than 'inefficient workflow'. It tells you what the new tool must replace and what the customer already knows how to use.

Use the same recruiting route where practical, and record differences where it is not. Six warm introductions to friends and six unanswered cold messages measure access differences as well as segment interest. Our guide to finding early adopters covers recruitment and conversations in more detail.

Read a comparison table without inventing certainty

Suppose the founder completes six conversations per group. The fictional notes produce this table. Each row records an observation from those conversations; counts in different rows can refer to the same people and must not be added together.

Agencies and contractors show more trial interest in this invented sample. Designers offer a clearer first test with the current product: the buyer was interviewed and neither agreed trial requires an unsupported capability.

Six conversations cannot establish market size, a reliable conversion rate or willingness to pay across a population. The table supports a limited decision about which trial to run next. Preserve that distinction when you copy the conclusion into a business plan.

Set the selection rule before choosing a winner

For this example, the founder sets two prerequisites: access to the buying decision and a trial that the current product can deliver. After those prerequisites, the founder considers evidence of recurring problems and concrete trial commitments. This is a practical decision rule for the example, not an official Lean Canvas scoring system.

Avoid adding every row into one total. A missing purchasing decision can disappear inside a high score for enthusiasm. Keep disqualifying conditions visible and explain why they matter for the next step.

A useful decision note would read:

Start with freelance designers selling fixed-scope website packages. In this fictional round, two agreed to test a real proposal, both control their purchase and both trials fit the current product. Reconsider the choice if those trials reveal a required capability we cannot deliver or if buyer discussions remove the uncertainty for agencies.

Keep the other candidates for review after the trial.

Check how the choice changes the rest of the canvas

A segment decision should change other boxes. For designers, the initial problem might be keeping scope consistent when reusing a previous proposal. For agencies, internal review could matter more. Contractors may need calculations that the first tool cannot provide.

Write the unique value proposition around the selected task. In the example, 'Create a proposal using your agreed service scope and reusable package descriptions' states what the trial attempts to deliver. Any time-saving promise needs measurement before it becomes a claim.

The channels box should name a credible route to more people matching the selected description. The revenue streams box should identify who pays and what they would pay for, with an untested price clearly marked as an assumption. Support requirements belong in costs, including time spent helping people complete the first proposal.

Keep a record that can survive the next conversation

Save the dated evidence and decision with your canvas. After the trial, update key metrics with what actually happened: completed proposals, support needed, purchase decisions or another measure tied to the original uncertainty. Keep trial sign-ups separate from completed use and payment. They answer different questions.

You can then revise the customer block and use the same evidence in GenerateBizPlan. Keep assumptions labelled in the generated document and review them against your notes. A polished paragraph should still show where the business depends on an unanswered question.

FAQ

How many customer segments should I compare?

Use a small set you can investigate with comparable effort. Three is the worked example here, not a required number. Add another candidate when there is a specific reason it may need a different offer, buying process or delivery model.

Can two segments share one Lean Canvas?

Yes, when their relationships are part of the same business model and remain clear. Comparing alternative markets with different problems and pricing may be easier with separate drafts. Label the roles so a buyer, a user and an alternative market do not get mixed together.

What if every segment has weak evidence?

Record that uncertainty and choose the smallest next investigation that can resolve it. You can refine the problem, improve access to buyers or test a narrower subgroup. An empty evidence field is an honest starting point; filling it with an assumed success rate does not improve the decision.

Related glossary terms

• Customer segments
• Early adopters
• Problem
• Existing alternatives
• Unique value proposition
• Channels
• Key metrics

See also

• Lean Canvas for a freelancer
• Lean Canvas for a consultant