Market Choice, Customer and Economic Case · Lesson 1

Define a purchasable segment, not merely a category

11 min read

Lesson objectives

  • Define an eligible account denominator and separate reachable demand from biased expressions of interest.
Reading tools

A segment must be buyable, not merely presentable

Suppose a software company says it serves “small businesses.” That category is too broad for an entry decision. It may include a shop owner who buys alone and a service firm that needs four approvals, with different budgets, sales routes, integration demands and service costs. A useful entry segment is a group of accounts that can be identified before selling, shares a consequential problem, can be reached through a feasible route, and can buy a repeatable offer. We call this a “purchasable segment” in this lesson. It does not imply every account will buy. It means the team can estimate the eligible account set, contact it, and test a specific purchase hypothesis at an affordable cost. This is our operational decision definition, not an official industry classification.

Fix the decision unit before estimating size. For an enterprise service, do not mix the number of firms with employees or end users. If a parent company buys for its branches, is one parent contract or each branch the account? That choice changes the apparent market and the conversion denominator. Define the period and the geography you can actually serve. A global market figure is not helpful to a team that can sell in two regions. The US Small Business Administration’s market research guidance asks about demand, market size, location, saturation and pricing, and distinguishes existing data from direct customer research. Yet a general source does not establish demand for this particular offer. That remains a hypothesis to test with potential buyers.

Start with exclusions before ranking. Write eligibility rules observable before investing a salesperson’s time: the operation the product serves, a minimum usage level if integration economics require it, a reachable contact route, an identifiable problem owner, and service and support capability. Do not use “has budget” as a gate if you have no honest way to observe it in advance; make it a later qualification question. Separate a hard gate from a ranking signal. Inability to support a customer’s system language may block a sale today. A slow approval cycle may reduce attractiveness without making the account inherently ineligible. If the product team could close the language gap in two months, record when that segment should be reconsidered rather than burying the constraint in an average score.

After eligibility, compare segments on common criteria. Problem intensity is not company size: ask what happens if the problem persists, who owns its resolution, and how often it recurs. Reachability asks whether the team can name accounts and find a permitted, testable contact channel. Profit potential looks at expected contribution after service cost, not headline revenue alone. Delivery fit checks setup burden, support demand and whether a solution can be repeated. Record evidence and uncertainty beside each rating. Five interviews from a founder’s network are not representative market evidence, though they can reveal the words buyers use. Three genuine requests for proposal more strongly support purchase intent in a narrow group, but they still do not prove that hundreds of similar accounts are reachable.

Do not treat unlike ratings as precise scientific measurements. A transparent decision matrix can expose sensitivity: state weights, apply the same rating definitions to each option, and test whether a reasonable change to weights changes the order. If a small change flips the result, gather more evidence about the decisive criterion or run a smaller trial. Balance expected value against speed of learning. A large segment with slow procurement may be less useful for a first entry than a smaller segment that gives evidence of payment and use within weeks, when the first wave aims to test the message and delivery rather than maximize immediate revenue.

Build an account list and defend its denominator

“People who might benefit” and “accounts we can reach now” are different populations. Begin with an account list carrying a stable identifier so branches and databases do not double-count the same buyer. Record where each entry came from, when it was updated, why it meets the gates, and whether a contact route exists. If a decisive field is missing, mark it unknown rather than eligible. An industry report can help frame an upper bound, but a launch plan needs a reproducible lower-level denominator: perhaps 800 accounts in scope, 320 on a supported system, 210 of those above the usage threshold, and 150 with an appropriate contact route. Do not multiply percentages from unrelated samples automatically. Supported-system users might be more likely than others to exceed the usage threshold. If the intersection is unknown, state a range or collect the cross-tabulation before publishing a false point estimate.

Those 150 are not forecast sales. They are a list on which reachability can be tested. Thirty may turn out to be duplicate accounts, leaving 120; half the email addresses may bounce because contact data is old. Define stages: eligible account, reached account, conversation with a problem owner, offer-qualified opportunity, purchase request, contract. Each transition has its own denominator. Conversations divided by sent messages cannot be compared directly with contracts divided by qualified opportunities. Keep acquisition cohorts by entry date and channel, because an aggregate may hide a channel that consumes the team with little result. Ask whether the purchase is once-only or renews; that matters to the unit economics taught in module two, although it does not redefine eligibility for this first wave.

Design the segment test to distinguish at least two explanations. If buyers describe scheduling as painful but do not request a trial, perhaps the pain lacks priority, perhaps the message does not connect its promise to delay cost, or perhaps the purchase owner was absent. Rather than switch segments immediately, divide qualified conversations between two clear messages: one promises fewer coordination hours, the other more reliable technician arrival times. Hold price, channel and the trial-request definition as constant as feasible. Stronger response only to the second message suggests a value-framing problem more than a market-selection problem. If rejection is repeatedly tied to an integration unavailable to most accounts, the eligibility gate or product readiness needs revision. Write an alternative explanation before seeing the results, so every outcome cannot be retrofitted as support for the prior choice.

Segment selection is a resource commitment, not a slide. If the team can onboard ten accounts a month and a successful campaign yields forty requests in a week, customer waiting time and trust may deteriorate. The entry decision therefore needs a temporary acceptance cap, priority for accounts representing the repeatable use case, and a truthful route for the rest that does not promise an impossible date. If an attractive segment takes six months to close, cash-flow capacity and team financing precede expansion. Management may choose a segment with lower long-term profit for faster learning, but should name that tradeoff and a date to reconsider it. An interest signal alone does not authorize higher spending; need, purchase, delivery and profit evidence must connect in one inspectable chain.

Lesson summary

The denominator is the serviceable account; an interest sample remains a signal to test.