The test for a real segment
Most segmentation is filing. Splitting a list by industry, headcount, and country produces neat groups and identical emails, which means it produced nothing.
The only test that matters
Write the first sentence of the email you would send to each segment. If two of them are the same sentence, those are not two segments. Merge them and stop maintaining a distinction that changes nothing.
A segment exists when the group has a distinct problem, or the same problem showing up in a distinct way. Everything else is a filter, which is useful for building lists and useless for deciding what to say.
The four splits worth making
Start here rather than with a taxonomy. In most B2B businesses these four do the majority of the work.
By the problem they have, not the industry they are in
"Teams whose customer information is spread across a CRM, a shared inbox, and someone's notebook" is a segment. "Professional services" is a filter.
This is the split that changes messaging most, and it is the one most teams skip because it requires knowing customers well rather than reading a firmographic field.
The highest-value splitBy what they are doing today
Using a competitor, using spreadsheets, using nothing, or using an internal tool someone built. Each needs a completely different message.
Someone replacing a competitor needs a comparison. Someone using spreadsheets needs permission to believe a change is worth the disruption. Sending either message to the other lands badly.
Four groups, four messagesBy trigger event
New funding, a new hire in a relevant role, opening an office, a compliance deadline, rapid growth.
Triggers beat attributes because they carry timing. A company that just hired its first operations manager has a reason to care this month rather than in principle.
Timing beats fitBy who you are talking to
The person who feels the pain, the person who signs, and the person who has to implement it care about three different things.
Same company, same product, three genuinely different emails. This split is often more valuable than any company-level split and is the one most often ignored.
Role, not seniority
Building segments you can maintain
Define each in one sentence, written down. If the definition takes a paragraph, the segment will be applied inconsistently by two different people within a month.
Use fields you already capture reliably. A segment defined on a field that is 40% populated will silently exclude most of the people who belong in it. Our guide to CRM data hygiene covers which fields survive contact with a busy week.
Write the message before you build the list. If you cannot write a distinctive first sentence, the segment is not real. Building the list first means you discover this after the work.
Keep an unsegmented default. People who fit nothing should still be reachable, with a general message. Otherwise a chunk of your list quietly becomes unmailable.
Put the segment on the record, not in an export. A segment that exists only in a spreadsheet is a snapshot, and it is wrong within weeks.
Segmentation that wastes time
Splitting by attributes that do not change the message. Country, company age, and job title seniority are often in this category. If you would send the same thing, do not split.
Segments of one. If a segment has three companies in it, write to them individually rather than building a segment definition and a message template.
Nested sub-segments. Industry, then size, then region, then role produces twenty-four groups, of which you will maintain three. The rest go stale and start receiving messages written for a market you no longer sell to.
Segmenting before you know your customers. Under a couple of hundred contacts, this is theatre. Write individually, notice which framings earn replies, and build segments from that evidence.
Never revisiting them. Segments describe a business that is changing. Reviewing them quarterly takes twenty minutes and prevents the slow drift into messaging aimed at the company you were two years ago.
Segmentation is upstream of everything else in outreach. The structure of a sequence, covered in cold email sequences that get replies, only works if the list underneath it is coherent, and no amount of copywriting rescues a message sent to people who do not have the problem. It also feeds lead scoring, since a segment that consistently converts is a strong fit signal.
Kin holds lead groups on the same records as pipeline and outreach, so a segment stays current as deals move rather than becoming a static list that needs rebuilding every quarter.
Frequently asked questions
- How many segments should a small team have?
- Three to five. Each one needs its own messaging, which is real work to write and maintain. Ten segments in a ten-person company means nine of them are stale within a quarter.
- Should I segment by industry or by company size?
- By whichever changes the problem. In some businesses a 20-person law firm and a 20-person agency have the same problem and the industry split is decorative. In others industry changes everything. Test it against your own closed deals rather than assuming.
- What is the difference between segmentation and personalisation?
- Segmentation changes what you say to a group. Personalisation changes details for an individual. Segmentation is more valuable and far less work, because writing four good messages beats merge-fielding four hundred mediocre ones.
- How do I segment when I have very few leads?
- Under about 200 contacts, formal segmentation is premature. Write to people individually and pay attention to which framings get replies. Those patterns become your segments once you have enough volume for them to matter.
- How often should segments be reviewed?
- Quarterly. Segments drift as your product and market change, and a segment defined eighteen months ago is often describing a customer you no longer serve well.