Market segmentation research: how to find segments you can sell to

Market segmentation research splits a market into groups of buyers whose needs, behavior or buying process differ enough that you would sell to them differently. Customer segmentation research does the same with people who already buy from you. For most small teams, the work has three parts: interviews to discover which differences matter, data to count how many buyers fall into each group, and a short written description of each segment you can act on. Large statistical segmentation studies exist, but you usually do not need one to start.

If you already have candidate segments and need to pick one, see how to choose your first customer segment. This page is about the research that produces the segments.

A worked example: software for veterinary clinics

Suppose you sell appointment and reminder software to independent veterinary clinics. Sales are uneven: some clinics sign quickly, others trial it and leave. You suspect the market is not one group. This example is hypothetical and used throughout the page.

The obvious split is clinic size. The research question is whether size is what actually changes the buying decision, or whether something else, such as who answers the phone or how many clinics share an owner, matters more.

List the differences that could change what people buy

Jeff Sauro of MeasuringU suggests thinking in terms of who, where, what, when and how: who customers are, where they are, what they have done and do and think, when they buy, and how they buy (for example, direct or through a reseller). MeasuringU Under "what they have done," he suggests recency, frequency and money spent.

For the example, a candidate list might be:

• Who: number of vets, single site or group-owned.
• What they do: whether a receptionist or the vets handle scheduling.
• What they think: attitudes to new software.
• How they buy: whether a practice manager or the owner signs.

Keep only the variables you think could change the product, the price, the message or the channel. A difference that changes nothing you would do is description, not segmentation.

Use interviews to find which differences matter

MeasuringU recommends starting small, with a few interviews and short surveys with open-ended questions, before a larger data collection. MeasuringU Interview people across the groups you suspect, including customers who left. Ask about the last time they changed how they book appointments, who was involved, and what almost stopped them.

In the example, interviews might show that clinic size matters less than ownership. Group-owned clinics buy through a central manager who compares vendors, while independent owners decide alone after a staff member complains. That difference changes who you sell to and what you say. Our guide to interviewing B2B buying roles covers how to separate those people.

Count and value each segment

Once you have candidate segments, estimate how many buyers are in each and what they are worth.

• Your own data: tag existing customers by segment and compare revenue, retention and time to sign. MeasuringU suggests estimating each segment's size, then its value, then crossing variables (for example, company size by industry) to look for patterns. MeasuringU
• Public data: the U.S. Small Business Administration points to Census Bureau and Bureau of Labor Statistics data for segmenting a consumer population, and to NAICS codes, the Census Business Builder and Statistics of U.S. Businesses for counts of businesses. SBA

Public counts rarely match your segment exactly. They might tell you how many veterinary businesses exist by size, but not which are group-owned. Treat them as a ceiling and use interviews or directory checks to estimate the share that fits.

Decide whether you need statistical clustering

Cluster analysis and factor analysis can find segments that are hard to see by eye, but MeasuringU notes they need specialist software and statistical skill. MeasuringU As one example of scale, MeasuringU says most of the segmentation studies it runs have more than 1,000 participants, lists a typical engagement at $50,000 to $100,000 over seven to nine weeks, and describes a "typing tool": a short set of questions that predicts which segment a new prospect belongs to. MeasuringU segmentation service

That is a large investment. It tends to make sense when you already have many customers, a large survey budget and decisions that depend on precise sizing. For a small team with a few dozen customers, tagging accounts and running a round of interviews is usually a more proportionate first step. The typing-tool idea is still worth borrowing: write two or three screening questions that tell you which segment a new lead belongs to.

Write a one-page sheet for each segment

For each segment, record:

1. A plain description, such as "group-owned clinics that buy through a central operations manager."
2. How to recognize one from the outside.
3. What they need that others do not.
4. Who decides and who uses.
5. Roughly how many exist and what they are worth.
6. The evidence behind each line, with gaps marked.

Find people in each segment to talk to

Segmentation interviews need people from every group, including groups you do not sell to yet. Instant Expert can help: you describe the experience you need, review the people it finds, and it sends your invitations; you pay for the calls that get booked. Directory pages such as operations professionals in veterinary services and finance professionals in veterinary services show the kind of profiles you might shortlist. For how many conversations to plan, see how many customer interviews you need.

If you have a segment in mind but no one to ask, start a search with a one-sentence description of the people in it.