Demand Generation and Revenue Growth Blog

Define Your ICP Like a Pro (Using AI and Actual Buyer Logic)

Written by Ian Spencer | Jul 11, 2025 2:19:11 PM

Let’s cut to it — most “Ideal Customer Profiles” are complete fiction.

They’re written in pitch decks, not pipelines.
Built on hope, not evidence.
Reviewed once, then forgotten.

And when founders come to me saying their outbound doesn’t work, this is usually why.

Because if your ICP is wrong — everything downstream breaks:

  • Your messaging won’t land.

  • Your sequences won’t convert.

  • Your sales team will waste time chasing low-fit leads.

You don’t need a fancy tool. You need a way to define your ICP based on real-world signal — and yes, you can use AI to help with that.

Let’s get into how.

First, Here’s What an ICP Isn’t

Before we fix it, let’s debunk a few myths.

Your ICP is not:

  • “Any business in [industry] with over 10 employees”

  • Just a job title

  • A rehash of your total addressable market

  • A one-pager full of buzzwords

It’s not meant to make you feel good.
It’s meant to help you say no — to the wrong accounts, the wrong conversations, the wrong leads.

Because clarity isn't about who you can help. It’s about who you help best.

Why Defining a Precise ICP Matters So Much

A clear, evidence-based ICP acts like a cheat code for every part of your growth motion:

  • Sales velocity improves — because reps aren't wasting cycles on bad-fit leads.

  • Marketing gets sharper — targeting and messaging becomes focused and relevant.

  • Product strategy tightens — you solve for the right segment, not everyone.

  • Revenue becomes repeatable — because you know who, why, and how you win.

Founders who skip this step?
They end up building for noise, not signal.
They chase meetings instead of momentum.

Here’s What a Strong ICP Actually Includes

A good ICP isn’t just a demographic profile. It’s behavioural. It’s psychographic. And, where possible, it’s backed by evidence.

Here’s what to include:

Attribute Examples
Industry & vertical RetailTech, B2B SaaS, DTC fashion
Company size 11–50 FTEs, or £2M–£5M in revenue
Tech stack Shopify, HubSpot, Salesforce
Buyer titles VP Growth, Head of Ops, Brand Director
Buying triggers Hiring SDRs, new funding round, launching new market
Internal language & pain “Struggling with conversion drop-off” or “churn in post-trial phase”
Reason they buy Reduce CAC, speed up time-to-value, solve one critical pain
Why they might say no Budget constraints, internal team pushback, not a current priority
And most importantly — the confidence level that you can win.

If you wouldn't bet your next 3 months of runway on that segment, they shouldn’t be your ICP.

How to Use AI to Define (and Validate) Your ICP

This is where things get interesting.

Most people use AI tools like ChatGPT to write copy.
But its real power is in accelerating critical thinking — if you feed it the right prompts and context.

Step 1: Give It Your Raw Data

List 5–10 of your best customers:

  • What industry are they in?

  • How big is their team?

  • What problem did they come to you with?

  • What do they say they like about your product?

This is your foundational signal. Paste this into ChatGPT and ask:

“Based on these customer profiles, identify shared traits, buying triggers, common pain points, and what language they use when describing their problems.”

Let it synthesise patterns you may have missed.

Step 2: Run a Contrarian Check

Next, ask:

“What assumptions might we be making that aren’t supported by this data? Are there anti-personas we should avoid?”

This helps you define who not to chase — just as important as who to target.

Step 3: Build Segment Hypotheses

Now test:

“Based on this, what 3 ICP segments would you recommend we test outbound messaging against? For each, define pain points, job titles, buying signals, and reasons they would buy now.”

You now have a prioritised list of segments, based on your actual customers — not market guesswork.

Bonus: Score Your ICPs Before You Scale

Before committing fully to a segment, you should run a simple scoring model. Here's how I’ve done it inside Revnuu for clients scaling outbound:

Factor Weight
Revenue potential 30%
Ease of access (data) 20%
Pain visibility (in messaging) 20%
Speed to close 15%
Referral potential 15%
Score each segment out of 5 for each factor.

Multiply by the weights.
Pick the top one or two to focus your GTM efforts.

AI won’t decide this for you — but it can accelerate your ability to reason through it.

What Most Founders Get Wrong About ICPs

Let’s talk mistakes — I see these all the time:

❌ They define it once and never revisit it

Your ICP isn’t static. If your product, pricing, or GTM changes — so should your ICP.

❌ They skip qualitative insight

Analytics tell you what happened. Conversations tell you why. Use both.

❌ They confuse TAM with traction

Just because there’s a market doesn’t mean you can win it now. Be ruthless.

❌ They write it in pitch-deck language

No one talks like this:

“Mid-market retailers seeking scalable omnichannel enablement”
Your ICP doc should sound like the customer, not an investor memo.

Real Talk: ICP Clarity is What Unlocks Everything Else

Without it, AI is just a content toy.
Your CRM is just an expensive spreadsheet.
Your sequences are just noise.

But when you get this right — when you know exactly who you help, how, and why they buy — everything sharpens.

  • Your cold messages become conversations

  • Your sales calls become faster to qualify

  • Your team knows where to focus, not flail

This is what we mean by building a revenue engine, not just running marketing campaigns.

Final Thought: Get Obsessed With Fit, Not Volume

As a founder, it’s easy to get caught up in the numbers game — more leads, more reach, more impressions.

But early-stage growth rarely comes from more.
It comes from better.

The better your ICP definition, the better your messaging, targeting, and close rate.

And yes — AI can absolutely help you get there faster.
Just don’t ask it to do the thinking for you.