Niching down: who should you focus on first?

So you should now have a customer profile, and potentially a slightly awkward collection of people around it. You might find that one person fits perfectly yet has ignored three messages, another sits well outside the profile and paid immediately, two free users work in completely different industries, and the ten people on your list look promising but haven't done anything yet.

This is normal, and is why we should now start thinking about niching downChoosing a smaller, more specific customer group to concentrate on first, so your product, message and attempts to reach people can all address the same problem. It doesn't set the permanent limit of the business..

Choosing one group can feel like volunteering to make a small business even smaller. You're trying to GET MORE CUSTOMERS, right?! Removing perfectly plausible sources of cash from the plan seems the opposite of productive.

Not true: and here's where we need to think of your nicheThe specific group of customers receiving your attention now. Members should share a problem, buying situation and enough similarities for you to serve and reach them in a consistent way. as the group receiving the majority of your attention first. It gives you a period in which the next five conversations, the page and the product all address the same people and problem. It doesn't decide the largest market you'll ever serve, and it doesn't require you to refuse good money from somebody who arrives from elsewhere.

And why? Simple. If you try to speak to everyone at this stage - with no market reputation, perhaps a quickly-pivoting product - you'll end up speaking to no-one.

The customer profile you wrote in the previous chapter gives you the first group to consider. This chapter checks whether focusing on them makes sense and records what would make you look again, so you're choosing where to apply the majority of your efforts and energy.

You may also see this group called a customer segmentA group of customers who share relevant needs, behavior or characteristics. Different segments may need different products, messages, prices or ways to buy., target marketThe people or organizations a business has chosen to sell to. A starting market is the smaller part receiving attention first., wedgeThe first specific customer group, use case, or problem a business uses to get started. A good wedge is small enough to focus the work and gives the business a sensible route into another market later., or beachheadA business term for the first market a company aims to establish itself in before moving into related markets. This playbook usually calls it a starting market.. They aren't perfect synonyms, although all three seem to be applied around this stage in go-to-market. Call it whichever suits you best.

Your niche changes what you do next

“Small businesses” can be accurate and still leave you writing a different page for a bookkeeper, a dog groomer and a small factory. “Women aged 25 to 40” may describe the people buying a product without explaining the problem they share, why they want it now, or where you could find them.

A clear starting marketThe first customer group you choose to concentrate on while learning how to sell and deliver the product. You can review and widen it when customer behavior supports the change. changes six pieces of your work:

Psychologically, a narrower group also makes it easier for a possible customer to recognize that the page is for them. They see their own problem reflected back to them, a result that'd provide a true solution, and proof from people or businesses like theirs; they have less work to do before deciding whether to keep reading, try it, or buy it.

If your chosen group changes none of these, the label still gives you nowhere to focus. An “independent appointment business,” for example, could lose $30 when somebody doesn't turn up or lose half a working day; those businesses won't necessarily pay the same price, believe the same proof, or need the same product.

Start with the people getting the best result

The strongest customer group from your ideal customer profileA working description of the person or company most likely to get a strong result from your product. In this playbook it uses three clear details: their situation, what they already do, and what has changed recently. is the first candidate here, although it doesn't win by default. Bring forward the customer actions you recorded and compare up to three possible groups in the same order:

1. The result they got;
2. Whether they came back or kept using it (retentionCustomers continuing to use or pay for a product over time. Early on, repeat use is often the first way retention becomes visible.);
3. The referralsA customer introducing or recommending the product to somebody else. It costs them some reputation, which makes it stronger than a private compliment. they made;
4. What it cost to find and serve them (the cheaper the better);
5. The quality of their feedback;
6. How straightforward they were to sell to (top points to those who take the least time/effort).

Keep payment on a separate line, and that's because although somebody paying is still the strongest proof that people will buy, there are too many variables to make it a solid contributing signal.

Once you've compared the customers, three questions decide whether this group will support the business you want:

• Is the problem pressing enough that people take action? (product/market fitWhen a recognizable group repeatedly chooses, uses, and pays for a product because it solves a problem they care about. One payment or one successful use is evidence to investigate; it doesn't establish product-market fit on its own.)
• Are there enough plausible buyers at a workable price? (market sizeAn estimate of how many plausible buyers exist and how much they might spend. At this stage, a sourced range is more honest than a precise forecast./serviceable obtainable marketThe part of a market a business could realistically reach and serve with its current product, price, resources, and ways of finding customers. At this stage, it should be a sourced range rather than a precise forecast.)
• If you serve this group well, is there another sensible group or product you could move into later? (land and expandStarting with one customer group, use case, or part of an account, then moving into related groups, needs, or accounts after the first one works. The possible next market explains where the business could go; it doesn't prove the business will get there.)

Let this help you decide who deserves closer attention.

Narrow enough = consistency

We can test a possible group without adding another score. People in the group should share:

• The same urgent problem;
• They want roughly the same result from the product;
• Similar proof that makes the result believable;
• One realistic way for you to find and reach them;
• Roughly the same product experience.

The group is still broad when you need several products, pages, prices or sales processes to serve it. A self-employed developer who buys a tool on a card and an enterpriseA large organization where a purchase may involve users, a champion, a budget holder, senior approval, procurement, legal, security and a longer sales cycle. engineering team that needs security approval may contain similar users; but you'll have to contact different people, use different language, provide different proof, and wait a very different length of time for payment.

You can also go too far. “Independent dog groomers with exactly 83 customers in towns beginning with B” is obviously ridiculous; at least the prospect list would be short. Details in your niche should only matter when they change the problem, purchase, product, or route to the customer.

The rough market size and unit economicsThe money made or lost on one customer, sale, order or transaction after the direct costs of winning and serving them are included. work from ValidationThe work of checking whether a problem is pressing enough for people to take action, using what they've already done and spent before relying on what they say they might do. can sit beside this comparison. For each group, the early calculation can stay as:

Number of matching buyers × plausible price × how often they buy

Add the likely customer acquisition costThe money spent to win a new customer. For a solo founder, the time spent on sales and marketing should also remain visible even when no cash was spent., or the money and time needed to win one customer, plus the cost of serving them; where delivery or transaction costs are substantial, use marginThe money left from a sale after the costs directly connected to delivering it have been removed., the money left after those costs, instead of revenue. This is a rough range, because a precise spreadsheet still can't tell you how many people will buy.

Five strangers

Research can give you three reasonable groups, but how do you prove them? Try this.

Pick the strongest of the three and test one group at a time; more than one and you'll have no idea what's working or why.

Five personal invitations to try your product are enough for the first check. The people should match your customer description and, where possible, shouldn't know you. Each one receives the same offer and the smallest real version you can provide: a by-hand service, a lightweight minimum viable productThe smallest working version of a product that lets a customer try to get the intended result. It can be manual or partly manual while the founder is learning., a working trial, a paid pilot, or an enterprise conversation that can produce a letter of intentA written statement that a person or organization intends to proceed with a purchase or partnership, usually subject to later conditions. It shows more effort than a compliment, while falling short of completed approval or payment..

One person taking a real action gives you enough reason to continue investigating. Depending on what you sell, they might:

• Pay;
• Begin the trial or pilot and use it;
• Share the time, access, or data needed to get a result;
• Take the request through an internal approval step;
• Sign a letter of intent during a long enterprise purchase.

Five compliments give you five compliments. “I'd buy this” still doesn't move any money, and an enthusiastic sign-up that never returns hasn't tested the result.

If nobody acts after five messages, the group may be wrong; you may also have contacted the wrong people, described the problem badly, asked for too much, or used a route they ignore. Review those four possibilities before replacing the whole market.

The appointment tool with three markets

Imagine you've built a simple product that collects a deposit and sends a reminder before an appointment. Your early conversations produce three groups:

• Independent dog groomers with long appointments and recurring customers;
• Beauty salons with several staff and an existing booking platform;
• Private tutors arranging lessons directly with parents.

All three deal with cancellations and no-shows. They experience the loss differently, use different systems, and reach the decision to buy through different people.

The dog groomers become your starting group. A missed appointment can remove a large block from a working day; many of the businesses you found already chase deposits through messages, you can identify them through public local listings and specialist groups, and the first product can work alongside their diary without replacing a larger booking system.

Choosing them changes what happens next:

This example ends with a decision and a test, because choosing a market hasn't proved that the market will buy. The five groomers still have to act.

An unexpected customer ≠ new direction

A beauty salon may find the product and offer to pay while you're concentrating on dog groomers. Take the sale when the existing product can serve them well; note who they were, why they bought, what result they got, and how much help they needed.

Serving that salon doesn't require a new page, campaign, or product plan. If it needs staff rotas, treatment-room scheduling and a migration from its booking platform, the work could easily pull you away from the simpler problem you're testing. If four salons arrive with the same request and budget, you've got a second group worth investigating - but don't let yourself get distracted before this happens.

The time to widen comes when sales and successful use in the first group become increasingly predictable, or when a recognizable second group repeatedly appears with the same problem. They also need money to spend, proof they'll believe, a way for you to reach them, and buying and support needs you can handle.

A different kind of focus

Some products only work when two groups, or enough people in the same group, are active at once. Choosing who to focus on is only the first decision; you also need to decide where to concentrate them and which side needs to arrive first.

A marketplaceA product that connects two or more customer groups so they can buy, sell, book or exchange something. Supply and demand need separate customer descriptions. may need one geography, category or use case on both sides so buyers find enough relevant supply. Start with supply when buyers would otherwise arrive to an empty marketplace; concentrated demand can come first when suitable suppliers already exist and want proof that buyers will turn up.

If your product only works when users can find or interact with one another - the network effectThe product becomes more valuable as more relevant people join and participate. The number of active people and what they do with one another both affect the result. - choosing a group isn’t enough; those people also need to be concentrated in the same place. “University students” sounds like a niche, but 200 students spread across 200 universities would leave each person with nobody relevant to connect with. Starting with students at one university would give the product a much better chance of working.

An enterprise product may have similar users in two industries and a completely different route to purchase through security, legal, procurement and final approval. The starting group needs a buying process you can support as well as people who want the product.

For a consumer product, the situation and recent behavior may tell you more than a job title or company; for a service business, the time required to deliver the work should live within the calculation, because a high price can be wiped by highly personal delivery.

Your starting-market record

The starting-market record puts the three possible groups, customer behavior, payment, rough numbers, and next test in one place. It also keeps the groups you've left outside the next steps, so one slow week doesn't send you back to selling to everybody.

The record can end in one of three answers:

• Choose: the customer actions and rough numbers support one group receiving your attention now;
• Test: one group looks strongest, although nobody has taken a real action yet, so the five invitations come next;
• Reject: the result, urgency, buyer count, cost, or route makes this group a poor place to concentrate.

“Everyone for now” leaves you with the same decision next week.

A prompt to compare three possible groups

Purpose: Compare the customer groups you're considering without inventing market numbers or turning a neat description into proof.

When to use it: Once you've completed the customer behavior and rough number sections of the starting-market record.

I need to choose the customer group I'll concentrate on first.

My context
- Product or service: [what I'm offering now]
- Product stage: [idea / prototype / lightweight MVP / pre-launch / live]
- Business model: [how I expect it to make money]
- Result I want the customer to get: [the change I want them to experience]
- Time or delivery limits I have to work within: [my current limits]
- Validation record: [paste or attach]
- Ideal customer profile record: [paste or attach]
- Earlier research and synthesis: [paste or attach, with source links]

The customer groups I'm considering

For each of no more than three groups, I'll provide:
- Group name in ordinary language
- The three clear details from my customer profile
- Customer or user cases and source dates
- Results achieved
- Repeat use or retention
- Referrals
- What it cost to find and serve them
- Feedback in their words
- How straightforward the sale was
- Payment, including whether they knew me before buying
- What makes the problem urgent
- Rough number of matching buyers and where the number came from
- Plausible price and purchase frequency
- Current way I can find and contact them
- Product changes they would need
- Other groups or products I could serve later

[Paste each group separately. Write “I don't know yet” where information is
missing.]

Your task
1. Review each group separately before comparing them.
2. Keep facts, my assumptions, your interpretation, and missing information
   separate.
3. Compare customer behavior in this order: result, repeat use, referral, cost
   to find and serve, quality of feedback, and ease of sale. Keep payment on a
   separate line as the strongest evidence that somebody will buy.
4. Check whether each group shares the same urgent problem, wants roughly the
   same result, believes similar proof, can be reached in one realistic way,
   and can receive roughly the same product experience.
5. For each group, show what would change about who I contact, what my page
   leads with, the proof I need, what I build next, where I find people, and
   what they compare me with.
6. Review my rough buyer count, price, purchase frequency, and cost to serve.
   Identify missing numbers without supplying invented ones.
7. Check whether the group has enough urgency and plausible buyers for the
   business I described, and whether success could lead to a sensible next
   market.
8. Recommend one outcome for each group: choose, test, or reject. Explain which
   customer actions and supplied numbers support the answer.
9. Recommend one starting group. If nobody has taken a real action, produce a
   five-person test using the same offer for five matching strangers.
10. Name the smallest customer action that would justify continuing the test,
    what would make me review the choice, and which out-of-group behavior I
    should watch.

Rules
- Don't invent customer behavior, quotations, payments, results, repeat use,
  referrals, costs, prices, market size, companies, people, communities,
  source links, or willingness to buy.
- Don't treat compliments, survey interest, free sign-ups, active use, letters
  of intent, and payment as the same result.
- Don't hide conflicting cases or average unlike customers together.
- Don't recommend several starting groups to avoid making a choice.
- Don't use a numerical score or total.
- Use “I don't know yet” where evidence is missing.
- Keep any uncertainty attached to the claim it affects.

Return exactly these headings
## Group 1
## Group 2
## Group 3 (only if supplied)
## Customer behavior comparison
## Payment
## Facts
## My assumptions
## Model interpretation
## Rough number check
## What choosing each group would change
## Choose, test, or reject
## Recommended starting group
## Five-person test (only if nobody has acted yet)
## Out-of-group behavior to watch
## Review trigger
## Missing information

Read the response beside the records you supplied. The model can compare your information and point to a weak assumption; it can't make a market appear or turn five polite replies into demand.

Good enough for now

You've finished when one customer group is receiving your attention; you can explain the choice through customer actions and rough numbers; it changes your target list, page, proof, product priorities and route to conversations; and you've written down the event that will make you review it.

If nobody has paid or used the product, your next step is smaller: five named strangers, one consistent offer, and one real action to look for. The next chapter, PositioningHow you explain what the product is, who it's for, which alternative it replaces and why the chosen customer should prefer it., will use the group you've chosen to work out how to explain the product so they quickly understand why it's for them.

Sources

• Dunford, April. “A Product Positioning Exercise.” 24 November 2019. Accessed 12 September 2026.
• Stripe Atlas. “Pitching Your Early-Stage Startup.” Accessed 12 September 2026.