Defining Your Ideal Customer Profile — and Sizing the Market Behind It
On Tuesday I argued that marketing should be judged like a capital investment. If you’re going to invest in a machine, you need to know what it’s going to produce, and if you’re going to invest in a pipeline, you need to know exactly who it’s for.
That’s the ideal customer profile. It’s the first thing we build in the Pipeline Sprint, because every other part of the build depends on it.
ICP from Margin, Not Persona
Most businesses have done some version of this exercise already, usually a workshop with a whiteboard and a persona called “Operations Olivia”, complete with a stock photo and a list of hobbies.
That isn’t an ideal customer profile, it’s a guess with a name.
A proper ICP isn’t built from opinion. It’s built from your sales data, and it starts with margin. The question isn’t “who do we like selling to?” but “which customers make us the most money, and what do they have in common?”
That’s why it follows on from the margin work in Month One. Once you know which services and customers sit at the top on gross margin, you have the raw material, and the ICP is the pattern behind them.
The Five-Step Build
This is how we build it.
1. CRM Mining
Pull every customer from the last two to three years and rank them on what they’re actually worth to the business, looking at:
- Total revenue
- Gross margin contribution
- Average deal size
- Sales cycle length
- Repeat purchase rate
- Whether they’ve referred anyone
The top 20% by margin contribution is your starting point, so study them closely, because they’re the customers your marketing should be finding more of.
2. Firmographics
Next, look for what those top customers have in common. For a B2B business, that usually means:
- Sector, which SIC codes from Companies House make precise rather than approximate.
- Company size, in terms of headcount and revenue band.
- Location, whether that’s the region or proximity to your sites if delivery matters.
- The buyer, meaning the job title and seniority of the person who signed and who else was involved.
- The trigger, or what happened that made them need you, such as a new contract, a failed supplier, an expansion or a compliance deadline.
The trigger is the one most businesses miss, and it’s often the most useful, because it tells you when to reach a prospect as well as who to reach.
3. Market Intelligence
With the pattern defined, you can use tools like Apollo.io or LinkedIn Sales Navigator to find other companies that match it. This is the point where the ICP stops being a description and becomes a list.
4. Search Language
Keyword research shows the words your ideal customers actually use to describe their problem, which is rarely the language the business uses internally. That gap matters, because it shapes every message we write during the build.
5. Document and Share
The ICP is written up as one clear document and shared with sales, marketing and whoever manages the website, because if sales and marketing are working from different pictures of the ideal customer, the pipeline suffers. It’s then reviewed every quarter as new sales data comes in.
Sizing the Total Addressable Market
Once the ICP is defined, the next question is how many businesses actually fit it.
Apollo.io TAM Analysis
In the Sprint, we build the total addressable market in Apollo.io by filtering on the ICP’s firmographics, including sector, size, location and decision-maker job titles. The result is a count of real, named companies that match your best customers, along with the people inside them who make the buying decision.
It’s one of the most clarifying numbers a business can have, and most owners have never seen it.
Sometimes it’s bigger than expected, which shows there’s far more room to grow than the current pipeline suggests. Sometimes it’s smaller, which means the business needs to win a bigger share of a tight market or widen the profile carefully.
Either way, the plan changes once the number is on the table.
What the Numbers Change in Outbound
The TAM figure turns outbound from a hopeful activity into a piece of arithmetic.
Here’s an illustrative example. Say the analysis finds 1,200 UK companies that match the ICP. That number tells you:
- How much outreach is realistic. 1,200 companies is a finite market, so you can’t treat it like an endless list, and every message has to be relevant.
- How to batch it. Rather than contacting everyone at once, you can work through the market in sensible batches and learn from each one before the next goes out.
- What the targets should be. If a reasonable share of those companies become qualified opportunities over a year, you can work backwards to how many conversations you need each month.
- What the budget should be. Once you know the size of the market and the value of a customer, the spend needed to reach it is no longer a guess.
It also explains why vague targeting fails so badly with AI. If the ICP is loose, AI outbound will happily contact thousands of businesses that were never going to buy, burning budget, damaging your sender reputation and putting your name in front of the wrong people, all at speed. A tight ICP is what makes automation safe to use.
ICP Red Flags
These are the signs that your ICP isn’t doing its job:
- “We sell to everyone.” If no customer is excluded, there’s no profile.
- Messaging that could apply to any business. If the copy doesn’t speak to a specific sector or role, the ICP isn’t being used.
- Sales and marketing describe different customers. That disagreement shows up as poor lead quality and blame.
- The website treats every visitor the same. There are no sector pages, no tailored routes and no landing pages for the segments that matter most.
- It was written once and never revisited. Markets change, so an ICP from three years ago may be pointing at customers who no longer fit.
If any of these sound familiar, fix the ICP before spending anything more on outreach or ads.
Next week, the build moves on to outbound, starting with why AI outbound isn’t spam at scale, unless it’s done badly.
Frequently Asked Questions
What is an ideal customer profile?
A data-led description of the customers who generate the most revenue, margin and lifetime value.
How is an ICP built?
By ranking customers on margin contribution, identifying firmographic patterns, then layering in market intelligence.
What is total addressable market analysis?
Counting how many companies match your ICP, which sets realistic targets for outbound and budget.
Is the ICP included in the Pipeline Sprint?
Yes. A defined and validated ICP and an Apollo.io TAM analysis are Month One deliverables.
How do I start?
Book the £495 AI Pipeline Audit.
Book the £495 AI Pipeline Audit
If your ideal customer is still a persona rather than a pattern in your sales data, that’s the first thing to fix. The AI Pipeline Audit starts with margin and ICP, taking 90 minutes and costing £495, which is deducted in full if you continue to the Pipeline Sprint.
