How to Build an ICP for LinkedIn Outreach (The Three-Fit Method)

Most LinkedIn outreach fails before anyone writes a single message.

It fails at the list.

The wrong people get the right message. Or the right people get it at the wrong time. Either way, nobody replies, and the team decides "LinkedIn doesn't work for us."

But the channel was never the problem. The ICP was.

In this guide, you'll learn how to build an ideal customer profile that actually works for LinkedIn outreach. One you can search for, filter by and write to.

We call it the Three-Fit Method.

Let's dive right in.

Why Most ICPs Are Useless for Outreach#

Here's a typical ICP:

"B2B companies, 50 to 500 employees, in growth mode, that value quality."

Sounds fine.

But try typing "values quality" into a LinkedIn search bar.

Exactly.

Most ICPs get written for a slide deck. They describe a vibe, not a list. And a vibe gives your reps nothing to filter on, nothing to check on a profile and nothing to write about.

An outreach ICP has a different job.

It has to answer three questions, in this order:

1. Which companies?
2. Which people inside them?
3. Why now?

Miss any one of those and the whole thing wobbles.

So let's fix it.

The Three-Fit Method, in One Picture#

The Three-Fit Method splits your ICP into three layers.

Each layer narrows the one above it.

Company fit gets you a market.

Person fit gets you a name.

Moment fit gets you a reply.

Most teams stop after the first layer.

That's why their lists are huge and their reply rates are tiny.

Now let's build each layer properly.

Step 1: Start From Evidence, Not From Hope#

The biggest ICP mistake is writing down who you WANT to sell to.

Instead, start with who has already bought.

Here's how:

First, pull a list of your best customers.

Not all customers.

The ones who:

• Paid without a long fight on price
• Got value quickly
• Stayed, renewed or referred someone

Then, pull a second list: the deals that went badly. Churned fast. Demanded endless custom work. Never really used what they paid for.

Finally, put both lists side by side and look for patterns.

Ask things like:

• What industries show up again and again in the good list?
• How big were they when they bought?
• Who signed? Who championed it internally?
• What had JUST happened at the company before they reached out or said yes?

That last question is gold.

Keep it in your back pocket.

It becomes moment fit in Step 4.

What if you don't have customers yet?

Then use the closest thing you have: past employers, past clients from a previous business, or the 10 to 20 people who gave you the most useful feedback when you described your offer.

And treat the result as a hypothesis, not a fact.

You'll test it in the first few weeks of outreach.

Which brings us to the first layer...

Step 2: Define Company Fit#

Company fit answers one question: is this business the kind we help?

Keep every criterion concrete.

If you can't check it in under a minute, it doesn't belong here.

Here are the criteria that tend to matter most for B2B:

Industry. Be specific. "Software" is not an industry for outreach purposes. "Accounting software for dental practices" is.

Company size. Use a range, and know why the edges are where they are. Too small and they can't pay. Too big and they have a procurement process you can't survive.

Location. Time zones, languages and buying habits all matter. So does whether you can actually serve them.

Business model. Do they sell to businesses or consumers? Do they sell through a sales team or online? This changes what they care about.

Growth signals. Are they hiring? Raising money? Opening offices? Stable companies and growing companies buy for very different reasons.

Here's a worked example.

Say you run a small agency that builds outbound systems for service businesses.

A fuzzy company fit looks like this:

"Service businesses that want more clients."

A sharp company fit looks like this:

"IT services and software development firms. 20 to 150 employees. Headquartered in the US, UK or Canada. Selling project work to other businesses. Founder still involved in sales."

See the difference?

The second version is something a rep can check on a LinkedIn company page in seconds.

Next, let's find the actual humans.

Step 3: Define Person Fit#

Companies don't reply to messages.

People do.

So person fit answers: who inside the company feels the problem AND can do something about it?

For most B2B offers, you're looking for two roles:

• The person who feels the pain every day. They're your best first conversation.
• The person who can say yes to the money. Sometimes that's the same person. In smaller companies, it usually is.

For each role, write down four things:

1. Titles. List the real titles people use. "Head of Sales", "VP Sales", "Sales Director" and "Chief Revenue Officer" can all mean the same seat.
2. Seniority. Founder, director, manager, individual contributor.
3. What a bad week looks like for them. Be specific. "Pipeline is thin" is generic. "Two reps hit quota and four didn't, and nobody can say why" is real.
4. What they'd brag about if your thing worked. This is what your messages should point at.

Points 3 and 4 are where your messaging comes from.

And they're the parts most teams skip.

Don't skip them. If you can't describe a bad week for your buyer, you don't know them well enough to message them yet.

A quick tip:

Read 20 LinkedIn profiles of people who match your person fit.

Not skim.

Read.

Look at their headlines, their About sections and what they post about.

You'll pick up the exact words they use for their own problems.

Use those words in your outreach.

Not yours.

Now for the layer that changes everything...

Step 4: Define Moment Fit#

Moment fit is the most important layer.

It's also the one almost nobody writes down.

Here's the idea:

The right person at the right company can still ignore you, simply because nothing is happening that makes your message relevant TODAY.

Moment fit is a list of triggers.

Events that make a problem urgent.

Some of the most useful triggers for B2B outreach:

Go back to the evidence you pulled in Step 1.

What happened at your best customers right before they bought?

Those events are your triggers.

Rank them by how often they showed up.

Here's why this matters so much:

A message that starts from a trigger doesn't feel cold.

It feels like you noticed something.

"Saw you just opened a London office" is a reason to talk.

"I help companies like yours grow" is not.

Next up: making sure all of this actually works inside LinkedIn.

Step 5: Run the LinkedIn Search Test#

This step takes 20 minutes and saves you weeks.

Take every criterion from Steps 2 to 4 and ask:

"Can I filter for this, or see it on a profile in under a minute?"

Sort each criterion into one of three buckets:

"Values quality" is invisible. Drop it.

"Struggling with sales consistency" is invisible. But "hiring their second or third salesperson" is visible on the company's jobs tab. Swap it in.

Then run a real search with your filterable criteria and look at the first 25 results.

Ask two questions:

1. How many of these would you be genuinely happy to talk to?
2. Is the total result count sensible for your team's capacity?

If fewer than half look right, your filters are too loose.

Tighten them.

If the total is tiny, your filters are too tight, or your market is small.

A small market is fine.

It just means every message has to count.

Step 6: Write Your Disqualifiers#

This is the step that saves your reputation.

Disqualifiers are the reasons you will NOT contact someone, even if they match everything above.

Some common ones:

• Current customers, and anyone in an open deal
• Competitors and their staff
• Companies you've lost to recently (give it a few months)
• People another rep on your team is already talking to
• Anyone who said "not interested" in the last six months
• Companies below a certain size, no matter how good they look

Write these down.

Share them with everyone who builds lists.

Because the fastest way to annoy a market is to message the same person three times from three different reps.

Step 7: Turn the ICP Into a One-Page Brief#

An ICP that lives in someone's head is not an ICP.

Put it on one page.

Here's a simple template you can copy:

Company fit

• Industry:
• Size (employees):
• Location:
• Business model:
• Growth signals:

Person fit

• Primary titles:
• Secondary titles:
• Seniority:
• What a bad week looks like:
• What they'd brag about if this worked:
• Words they use for their problem:

Moment fit

• Top three triggers, ranked:
• Where to spot each one:

Disqualifiers

• Never contact:

Search setup

• Filters used:
• Expected list size:

That's it.

One page.

Everyone on the team uses the same one.

When a rep is unsure about a contact, they check the brief.

When messaging feels off, you check the brief.

When results drop, you start by questioning the brief.

How to Keep Your ICP Honest Over Time#

Your first ICP is a guess.

A good guess, if you followed the steps above.

But still a guess.

So treat it like one.

Every few weeks, look at your outreach results by segment:

• Which industries replied most?
• Which titles booked calls?
• Which triggers led to real conversations, and which led to polite silence?

Then adjust.

Drop the segment that never replies.

Double down on the one that does.

Add a trigger you keep seeing in good conversations.

Small, regular edits beat a big rewrite once a year.

And keep a simple rule: change one thing at a time. If you change the industry AND the titles AND the message in the same week, you'll never know which change worked.

Wrapping Up#

A sharp ICP makes everything downstream easier.

Your lists get smaller and better.

Your research gets faster.

Your messages get more specific, because you actually know who you're writing to and why now.

Here's the Three-Fit Method one more time:

1. Company fit: is this the kind of business we help?
2. Person fit: who inside it feels the problem and can say yes?
3. Moment fit: what's happening that makes now the right time?

Now I'd like to hear from you:

Which trigger shows up most often right before your best customers buy?

Let us know.

We read every reply.