Narrow your first customer segment until buyers share the same trigger, consequence, buying role, route to market and price without making the pool too small.
Short answer: Narrow your first segment until ten qualified prospects can hear the same problem, result, price and sales message without material changes. They should share a buying trigger, consequence, decision-maker, route to reach them and viable price. Broaden only if the resulting reachable pool cannot support the customer count your economics require. Demographics alone do not define a useful segment unless they change how or why the customer buys.
“Everyone who needs this” is too broad because need does not create one buying process. “Independent interior design studios handling 8 to 20 live residential projects” can be useful if those firms share the same operational problem and owner.
Narrowing is not about making the description impressive. It reduces variation so early evidence is interpretable and delivery can repeat.
Use the Segment Coherence Test
The Segment Coherence Test checks five buying conditions. Interview and offer to ten prospects who appear to fit. The segment is coherent when at least eight share four or more conditions without you rewriting the offer.
- Trigger: Question: What event makes action timely?; Evidence of coherence: The same deadline or operational change recurs
- Consequence: Question: What gets worse if they do nothing?; Evidence of coherence: Buyers recognise a comparable cost, delay or risk
- Authority: Question: Who approves spending?; Evidence of coherence: The same role can reach a decision
- Access: Question: Where can you reach them credibly?; Evidence of coherence: One or two repeatable routes exist
- Price: Question: What scope supports buyer value and your contribution?; Evidence of coherence: One price band works without large customisation
Eight out of ten and four conditions are working thresholds for an early diagnostic, not statistical proof. A high-value or regulated market needs deeper evidence. The point is to expose variation before it enters the business.
Narrow by buying mechanism, not description
Age, gender, company size and location can matter, but only when they change the problem, purchase or delivery. A 35-year-old and a 55-year-old may buy the same bookkeeping result for the same reason. Two firms of equal size may have completely different approval processes.
Useful narrowing variables include:
- The event that creates urgency
- The workflow where the problem occurs
- Volume or complexity that makes the cost material
- The role carrying responsibility
- Existing alternative and switching constraint
- Geography when delivery or trust is local
My view is that behavioural and operational boundaries should come before demographic ones for most small-business offers. They tell you what to sell and how to reach the buyer. Demographics often tell you only how to describe the person.
Listen for offer drift
Your segment is too broad when every conversation changes the promise. One buyer wants speed, another wants compliance, a third wants strategic advice and a fourth expects low-cost execution. Those may be separate markets sharing a category label.
Track material changes in a simple table: If the same columns repeat, you have coherence. If every row differs, split by the variable that explains the largest change. Do not average the requests into a service that promises everything.
Some variation is ordinary. Customers use different words and have exceptions. Split only when the difference changes value, sales process, price or delivery enough to require a separate commercial model.
Check whether the narrow pool can support the business
Calculate required active customers: Required customers = monthly contribution target divided by contribution per customer, rounded up
Then build a defensible count of reachable prospects that meet the segment conditions. A narrow segment fails when your required customer count consumes an implausible share of that pool or when acquisition must constantly expand beyond it.
Do not use a national industry total if your route reaches one region. Do not count firms too small to feel the consequence or too large to use your delivery.
A segment can begin small if purchases repeat and retention is credible. Use observed buying frequency, not hoped-for subscriptions. Run a low case in which customers buy once.
Worked example: Cara's sample-tracking service
Cara offers a sample-tracking service to “small creative firms”. Her first ten conversations include graphic designers, architects, stylists and interior designers. The physical items, deadlines, buyer roles and willingness to pay differ. She rewrites the offer six times. The category is not coherent.
She narrows to independent interior design studios managing 8 to 20 live residential projects. Eight of ten prospects describe the same trigger: samples move among a studio, client and supplier during approval. Owners carry the cost of lost items and delayed decisions. All can be reached through one industry network, and seven accept the same £420 monthly price as plausible.
Direct customer-specific costs are £35 a month. Delivery and support take eight hours, and Cara values her time at £30 an hour.
Economic surplus per customer is: £420 minus £35 minus (8 × £30) = £145 a month.
Cara wants £1,740 monthly economic surplus at the first stage. Required active customers are: £1,740 divided by £145 = 12 customers.
She identifies 180 studios that meet the volume and ownership conditions within routes she can legitimately use. Twelve customers equal 12 divided by 180 = 6.67 per cent of the reachable pool.
That percentage is not automatically attainable, but it is testable. Cara should make ten fixed-price offers and measure paid response. If most studios need fewer than eight support hours, economics improve. If data entry takes 14, the shared price may fail even though the segment is coherent.
Know when the segment is too narrow
Warning signs include fewer reachable prospects than your required customers, buying tied to one annual event, dependence on one platform or association, and a description that customers do not recognise.
“Dentists who opened in the past 30 days and use one named supplier” may be precise but commercially brittle. Remove a qualifier and test whether the buying mechanism remains the same.
Do not broaden merely because ten prospects refuse. They may be the right segment rejecting a weak offer. Diagnose value, timing, trust and price. Broaden when evidence shows an adjacent group shares the mechanism and can use substantially the same sale and delivery.
Keep the first segment temporary
Your initial segment is a learning boundary, not a permanent corporate identity. It helps you collect comparable objections, contribution and results. Once repeatability exists, you can test an adjacent segment deliberately.
Expansion should preserve at least three conditions, such as trigger, consequence and buyer. Moving to a group that changes all five creates a new market test.
Where segmentation uses personal data or protected characteristics, ensure research and targeting meet applicable privacy, equality and sector requirements. Rules vary by jurisdiction, so obtain qualified local advice for material decisions.
Related guides
Define the segment in ten days
Write one sentence naming the workflow, trigger, scale and buyer role. Identify ten prospects that appear to fit and make the same offer to each.
Then act in this order:
- Record the five Segment Coherence Test conditions for every prospect.
- Count material changes to result, price and delivery.
- Split on the variable that explains those changes.
- Calculate required customers and the genuinely reachable pool.
- Keep the narrowest segment that supports both a consistent offer and viable customer count.
If you cannot use the same price and promise eight times, narrow again. If the pool cannot support the arithmetic, remove one qualifier and retest coherence.
Frequently asked questions
Is “small businesses” a customer segment?
Usually not. Small businesses differ by industry, workflow, owner involvement, volume, budget and buying trigger. The label may set an upper size limit, but it rarely creates one problem or sales process. Narrow to businesses sharing a consequence and decision mechanism, such as owner-operated wholesalers processing a specified volume of supplier changes.
The exception is an offer created by a rule that genuinely applies across small businesses, but sector and structure may still change implementation. Test ten prospects with the same offer before treating the category as coherent.
Can my segment include more than one industry?
Yes, when the same trigger, consequence, buyer, access route and price hold across them. A workflow-based offer can cross industries more naturally than a sector-specific one. Compare actual buying conversations rather than assuming similar language means similar value.
If regulation, terminology or delivery changes materially, separate the industries for the first test. The exception is an established network or platform that brings several industries through one buying route. Even then, track contribution and objections by group so a strong segment does not conceal a weak one.
How many prospects should I interview before narrowing?
Ten qualified prospects are a practical first coherence test, not proof of a market. Look for repeated mechanisms, not a majority opinion. If the first five reveal two clearly different buying processes, split early and recruit comparable prospects for each. High-value, regulated or highly varied work needs deeper investigation.
The exception is a concentrated market with very few buyers, where each conversation carries more weight and direct paid offers may be more useful than broad interviewing. Keep the offer consistent so additional conversations improve evidence rather than add new variables.
What if customers do not recognise my segment label?
Use language they recognise, but preserve the operational boundary internally. Customers may not call themselves “mid-volume residential studios”, yet they can confirm project count, sample flow and ownership. A segment name is useful only if it helps identification and communication.
Do not force jargon into marketing. Test the plain description in real buying conversations. The exception is a regulated or professional category where the formal label determines permissions or buying rules. Verify definitions with current authoritative guidance and qualified professionals rather than inventing your own classification.
Does a narrow segment limit future growth?
It limits the first test, not necessarily the business. A coherent segment helps you learn one buying and delivery mechanism. Growth can come from deeper penetration, higher value, additional problems or adjacent segments. The risk is becoming dependent on a tiny pool or one economic cycle, which your reachable-customer calculation should reveal.
The exception is an offer whose infrastructure requires a very large market immediately. A narrow test may not support that cost structure, so redesign the model or secure stronger demand evidence before building capacity.
When should I add a second customer segment?
Add one after the first has repeatable sales, known contribution and a clear reason the adjacent group should buy substantially the same offer. Test the second separately and compare acquisition, price, delivery and retention. Do not combine results until coherence is demonstrated.
The exception is early evidence that the first pool is too small to support required customers. In that case, broaden one condition deliberately, keep the other four stable and rerun the test. Adding unrelated buyers to avoid a no creates confusion rather than diversification.
Comments (0)