Most Founders Get ICP Wrong (Here’s How to Fix It)

What I’d Do Today If I Were Defining My ICP from Zero

Think Cold Email Isn’t Working?

You’re probably right.
But not for the reason you think.

Founders blame the wrong things:

• “Our email copy sucks.”

• “People don’t reply anymore.”

• “Buyers are overwhelmed.”

• “Email is dead.”

And so they try to fix it with:

• ChatGPT-written sequences

• More tools

• Higher volume

That’s not solving the problem. That’s hiding from it.

The real issue?
You’re sending good messages to the wrong people.

Because you haven’t actually defined a clear outbound ICP.

The Hidden Cost of a Flimsy ICP

Your ICP isn’t just a bullet-point persona in your deck.

It’s the foundation of:

• Who do you reach out to

• How you position your product

• What pain do you solve

• Why people reply

• Whether people buy

When you get it wrong, here’s what happens:

• You chase false signals: “Someone said they liked our product” (but never paid for it).

• You build for non-buyers: “They asked for a feature!” (but never onboarded).

• You waste time optimizing copy for a prospect who was never qualified in the first place.

In early-stage outbound, 90% of failures come from targeting the wrong people.

If I had to start all over again as a founder, here’s exactly what I’d do to fix that.

Step 1: Flip Your ICP Thinking: Problem First, Persona Second

Forget the LinkedIn filters for a moment.

Outbound doesn’t start with who you want to talk to.

It starts with one question:
“What painful, recurring problem do we solve that someone wants fixed today?”

Examples:

• “Sales teams can’t see which users are about to churn.”

• “Product leaders have no clean way to prioritize bugs vs. feature requests.”

• “Growth teams are flying blind because attribution is broken.”

This sounds simple, but most founders skip it. They jump straight to messaging before they even know if the pain they’re solving is urgent, frequent, or valuable.

You don’t sell software.
You sell relief from pain.

And if the pain’s not obvious, you’re not selling. You’re begging.

Step 2: Hunt for Signals of Pain, Not Just Personas

Now that you know the problem, go look for companies that probably have it.

But don’t just rely on firmographics. Use “signal-based targeting”:

1. Their Website
   What do they sell? How do they position themselves? What language do they use?
   Look for clues that relate to the problem you solve.
   
   

2. Their Product (if public)
   Sign up. Use it. Break it.
   Ask: Is there friction that your software/solution could remove?
   
   → Selling an API observability tool? See if their docs are stale or incomplete.
   → Selling UX tooling? Sign up and track the onboarding experience.
   
   

3. Job Postings
   These are gold.
   
   

   • Are they hiring someone to manage the problem your product solves?

   • Are they trying to build in-house what you already offer?

→ If they’re hiring a lifecycle marketer, they probably care about onboarding and retention.
→ If they’re hiring a DevOps engineer, maybe observability or infra costs are pain points.

Funding or Headcount Changes
A company that just raised $5M and hired 10 SDRs?
Their GTM engine is probably still full of friction, and your tool might fit right in.

Step 3: Find the Person Who Feels the Pain Most

ICP isn’t just about the company. It’s about the individual who:

• Lives with the problem daily

• Owns the KPI you improve

• Feels urgency (not just interest)
  
  

In outbound, you’re not just qualifying buyers. You’re qualifying pain-owners.

Examples:

• Head of Support, if you solve the ticket response time.

• Product Manager, if you improve user onboarding.

• Revenue Ops, if you touch pipeline attribution.

And here’s the key:

Don’t just find people with the title.
Find people with the motivation.

If someone’s posting about a challenge your product solves, or recently shared something related, that’s a clue.

If they’ve been at the company less than 3 months, they may still be shaping their stack. That’s a window.

If they’ve recently switched jobs, they may want to bring in tools they liked before.

These are the humans who reply because your message speaks directly to the problem they’re under pressure to fix.

Step 4: Build 2–3 Insightful Hooks Per Prospect (Yes, Per Prospect)

This is where most founders flinch.

“I can’t research every prospect that deeply.”
Well, then, stop complaining about low reply rates.

You don’t need 1,000 prospects. You need 10-20 who actually care.

For each target, write down:

• 1 thing from their website that connects to the pain

• 1 thing from their product, docs, or UX

• 1 thing from the person’s profile, post, or job description

This becomes your cold message scaffolding.

Example:

If I were selling a UX analytics tool to a fast-growing B2B SaaS:

“Hey Lea, I was checking out your onboarding flow after seeing your team just launch a new feature. Noticed a couple of friction points that might be impacting early activation. I saw the same thing when I was at [past startup]. Happy to share what we learned?”

Specific. Relevant. Earned.

What Great ICP Feels Like in the Field

When you dial in your outbound ICP, things feel different.

Here’s what you notice:

Early Signs of Fit

• People respond with “Yes! We’ve been struggling with this.”

• Your open rates don’t drop because your emails don’t feel cold.

• Calls aren’t exploratory, they’re fast and focused.

Stronger Conversations

• Prospects ask detailed questions because they’ve felt the pain.

• Objections are predictable and solvable.

• You hear “We’ve tried X, but it’s not working,” not “We’re just exploring.”

Momentum You Can’t Fake

• Prospects introduce you to teammates unprompted.

• You start getting inbound from similar companies.

• You’re not chasing. You’re getting pulled in.
  

Outbound becomes a feedback engine, not just a booking engine.

The Founder Trap: Outbound as an Activity, Not a Strategy

Too many founders treat outbound like a checkbox:

• Build a list

• Write a sequence

• Hit send

But outbound done right is not a volume game.
It’s a learning game.

It’s how you discover:

• Who really feels the pain

• What language they use to describe it

• What alternatives they tried

• What triggers action

It’s a structured way to do customer discovery at scale.

But only if you treat it like research, not mass marketing.

Final Word: Be Ruthlessly Precise

You don’t need fancy automation.

You need a tight lens on:

• What problem do you solve

• Who feels it

• When they feel it

• What tells you they’re feeling it

If your outbound isn’t working, don’t write a new email.
Write a better prospect list.

Talk to fewer people. Learn more from each one.
That’s how you find your wedge.

Stay ruthless. Stay real.
— Guillaume