Start with the work
“Small businesses” is too broad to help you decide what to research next. “Founders who need growth” is little better. Both describe an enormous number of businesses with different problems, budgets and ways of buying.
You need a smaller decision: for this particular service, which of two plausible segments is worth investigating first? The answer does not have to become your permanent niche. It needs to be specific enough that a small research exercise can confirm something useful or reveal a mismatch.
Before naming an industry, write one sentence about the work you can deliver. Include the deliverable, the situation it addresses and the limits you need to respect. A service description is not a promise that clients will achieve a financial result.
Illustrative offer: “I review the handoff between an enquiry and a proposal, then deliver a documented follow-up process that a small consulting team can operate.”
That is more useful than “I help B2B businesses grow.” You can ask whether a business has that handoff, whether someone owns it and whether your deliverable fits the way it operates.
Now name the constraints. Can you work only in one language or time zone? Do you need access to a certain type of operational record? Can you serve one project at a time? Are there regulated sectors or technical systems you are not equipped to handle? An attractive-looking segment that violates a hard delivery constraint is not a good starting point.
Choose two real options
Compare two candidates that could plausibly buy the same service. Changing the offer, geography, customer size and delivery model all at once makes it difficult to understand the trade-off.
Describe each segment as a business situation, rather than a string of flattering adjectives. Include the type of business, a relevant operational characteristic and the boundaries of your service.
- Weak: “Ambitious founders.”
- More useful: “Owner-led B2B training providers in the UK that sell scheduled programmes through enquiry calls.”
- A contrasting option: “Owner-led B2B consultancies in the UK that sell scoped projects through proposals.”
These descriptions are still hypotheses. Neither one tells you that a particular business has a follow-up problem, can afford your service or wants to speak to you. Company categories and LinkedIn profiles can help locate candidates; they do not establish willingness to buy.
Keep a third area on your page: not for this offer. Record exclusions before you encounter an impressive-looking company. Examples might include self-serve consumer products, businesses outside your working language, or projects that require an implementation you do not provide. Exclusions protect your time and make later research less subjective.
Separate observations from assumptions
For each candidate, make three short lists. Resist the temptation to give the segment a score out of ten. A neat number can hide the fact that one observation is recent and specific while another is merely something you read in a broad industry article.
| Record | What belongs here | What it does not prove |
|---|---|---|
| Observations | A dated conversation, a public service description, a process you have seen with permission | That every business in the segment has the same problem |
| Exclusions | Conditions that make your service unsuitable or impossible to deliver | That the remaining companies are qualified buyers |
| Unknowns | Who owns the process, whether the problem exists, what work is already in place | A hidden opportunity that the worksheet can discover for you |
An observation should be traceable. “Three people told me follow-up is a problem” is hard to evaluate later. “Two enquiry-handling conversations in September; both described proposal follow-up as an owner task” is clearer, if that is what actually happened. Do not put confidential names or client records into a public-facing research brief.
You may have much better access to one segment because of previous work, professional familiarity or a legitimate introduction. Record that as an access advantage. It is not proof of demand, but it can make a small test cheaper and easier to interpret.
Worked example
This is a fictional illustration. The businesses, observations and costs below are made up to show the method. They are not client results or market evidence.
Imagine a consultant offering the enquiry-to-proposal review described above. She is considering two segments, both in a language and time zone she can serve.
| Question | Segment A: B2B training providers | Segment B: B2B consultancies |
|---|---|---|
| Relevant situation | Enquiries about scheduled programmes | Enquiries that lead to scoped proposals |
| Illustrative observation | A public page describes an adviser call before enrolment | A public page describes a discovery call and proposal stage |
| Important exclusion | Fully self-serve course sales | Firms that require a full CRM implementation |
| Missing evidence | Whether follow-up is already automated and owned | Whether the owner would share enough process context for a review |
| Delivery constraint | The service must work around a programme calendar | The service must fit a short project review |
Neither column wins because it looks busier. The consultant might prefer B because she understands project proposals better. Or she might choose A because the calendar creates a clearer boundary for the service. Those are her reasons, not an algorithm’s verdict.
For a small research exercise, suppose A needs four hours at an internal time value of £40 per hour, plus a £30 external research quote. That is £190 in estimated research cost. B needs three hours at the same time value and a £50 quote: £170. Those figures only compare the stated work. They say nothing about market size, response rates, revenue or which segment is likely to buy.
The useful next step is to name the uncertainty the consultant will investigate. If she cannot identify what a sample would help her learn, collecting more profiles will probably not resolve the decision.
Define a small test
A test needs a question and a stopping point. “Build a large list and see what happens” is neither. Start with the minimum research that could change your decision.
For example: “Can I find a small set of businesses that appear to use an enquiry-to-proposal process, operate within my delivery boundaries and have a plausible role responsible for it?” You would still need to establish the actual problem through an appropriate conversation. A profile alone cannot do that.
Write down:
- What you will inspect. Business-level information, service descriptions and relevant professional roles, where available.
- What would rule a candidate out. Your written service exclusions, rather than a vague judgement that a company feels wrong.
- What remains unproven. Need, budget, authority, timing and interest usually require further evidence.
- The research limit. A small sample size, a time budget and a point at which you will stop and review.
- The decision the test can change. Continue with this segment, adjust the service boundary, or investigate the other segment.
If LinkedIn is an appropriate source for the roles you need to inspect, a profile sample can be one input. Check what the actual file contains before relying on it. A profile record is not a verified email address or confirmation of the company’s buying plans.
Make a provisional choice
Use the Service-to-Segment Decision Sheet to keep the two options on one page. Record the service, the evidence, the exclusions, the costs and your own decision. You can choose A, B, a further research step, or neither. “Not enough evidence yet” is a useful outcome when it names the exact gap.
A good decision note has three parts:
“I will investigate segment B first because it fits my delivery limits and I understand its proposal workflow. I still need to establish who owns follow-up. I will review a small profile sample before deciding whether to commission more research.”
That note does not promise clients. It gives your next hour of work a purpose.
If you are considering paying someone to do the research, read When Is Outsourced Prospect Research Worth Paying For? next. It explains how to separate research effort, quality checks and service delivery, so a low price per row does not become the whole decision.
Questions to settle before you move on
Can I choose more than one segment? You can serve several. This exercise deliberately compares two options for one offer so that the first research decision stays manageable.
Does a small sample validate demand? No. It can help you inspect whether the records and businesses fit your brief. It does not establish willingness to buy.
What if neither segment looks workable? Record why. Adjusting the service or your delivery constraints may be more useful than collecting another list.
Do I need to send the worksheet to the team? No. It runs locally in your browser. You can download or print your own brief without signing up here.