ICP Analysis for B2B Companies | Oren Greenberg
Growth audit module · Strategy

ICP strategy analysis

Most B2B companies define their ICP in a workshop and use it as a guide. This analysis builds it from closed-won data - so you know which companies you actually win, not which ones you think you should.

Request a sample

Sound familiar?

  • Sales says some leads are great and others are a waste of time - but no one has put numbers on the difference.
  • Your ICP exists as a slide, but every team interprets it differently in practice.
  • Budget is spread across segments that have never converted at scale.
  • New market versus existing segment decisions are made on instinct and internal politics, not evidence.
  • You're generating leads but the deal sizes are all over the place - and no one knows why.

What the analysis tells you

Who you actually win

The firmographic and technographic characteristics that distinguish your closed-won companies from the ones you lost - derived from your own data, not benchmarks.

The real disqualifiers

The signals that predict a loss before a conversation gets started - the ones worth filtering out before they consume sales time.

What drives deal value

Which company characteristics predict large deals versus small ones - so effort goes to the accounts with the highest ceiling, not just the fastest to close.

Distinct customer profiles

Where your market splits into genuinely different buying segments - each with its own win conditions, deal profile, and what to say to them.

A scoring formula

An operational ICP score your team can apply to inbound leads and outbound targets - so the ICP is used, not filed.

Where to concentrate next

Which profiles represent the highest-confidence expansion opportunity versus the ones that look promising but haven't been validated by your own history.

Found in real analyses

Recent findings from delivered work, anonymised.

5,131 → 639

The company's ICP had been built on deal-level data that counted long-term clients up to 80 times each. Collapsing to one row per company halved the apparent win rate and invalidated 18 of the 21 factors that had been treated as significant signals.

ICP analysis, B2B professional services, 2026
2.4×

Companies with an interim CTO were 2.4× more likely to lose - it was the single strongest disqualifier in the logistic model, and it had never been tracked. Adding it to the qualification checklist immediately reduced wasted discovery calls.

ICP analysis, B2B professional services, 2026
28×

The deal value gap between the top and bottom ICP quartile - top-quartile accounts averaged £151k; bottom-quartile averaged £5.4k. Company size (median 2,894 vs 114 employees) explained most of the variance. The company had been treating both quartiles the same.

ICP analysis, B2B professional services, 2026

Want to see it first?

Request a sample and see what an evidence-based ICP analysis looks like in practice - actual methodology, actual output structure, client details removed.

Request a sample

Frequently asked questions

What is an ICP analysis?

An evidence-based analysis of who your company actually wins business from - derived from your closed-won data, enriched with firmographic and technographic signals, and validated statistically rather than built from internal opinions. The output is a set of scored customer profiles that tell sales and marketing who to prioritise, and why, backed by the numbers.

How is this different from an ICP workshop we've already done?

Workshops produce consensus documents. This produces evidence. The analysis starts with your actual closed-won deals, collapses them to company level to remove repeat-client inflation, enriches each company with external data, and runs statistical tests to find which firmographic and technographic factors genuinely predict conversion and deal value - not which ones your team agrees sound right.

What data do you need from us?

A closed-won deal export from your CRM - ideally with company names, deal values and close dates. We enrich the company data externally from there. If your CRM data is incomplete or inconsistently tagged, we can work with what exists and flag where the gaps limit confidence.

What does the output look like?

A set of distinct customer profiles, each with its defining firmographic and behavioural characteristics, the factors that predict wins within it, the factors that predict losses, and the deal-value profile. Each profile comes with a scoring formula your team can apply to new prospects - so the ICP is operational from day one, not a slide deck that gathers dust.