Problem definition

Anti-ICP

Anti-ICP is licensed CC BY 4.0. Attribution: Katafacts (katafacts.com).

1 · What it is

What it is

An anti-ICP names who not to pursue, or who to pursue only with a specific countermeasure — the same one-page, multi-section discipline as the ICP canvas, deliberately not a mirrored version of it with the polarity flipped. Not every red flag deserves the same weight: a hard disqualifier (an account under a firm size floor, no one to own the tool post-purchase) should end a deal outright, while a softer red flag (a single-threaded champion, a vague trigger) predicts risk without being disqualifying on its own — proceed, but with a specific next step. A real anti-ICP also carries churn-pattern evidence: not just why a deal didn't close, but why a customer who did buy later left, which is often the clearest signal of all for what to watch for earlier.

2 · When to use it

When to use it — and when not to

Use it when

You have real win/loss and churn evidence and want to turn the patterns in it into concrete, actionable disqualifiers and red flags — not a rep's general sense of 'red flags.'
Reps keep spending real time on deals that predictably won't close or won't stick, and there's no written list of the specific patterns that predicted it before.
You want to separate what should end a deal outright from what should just trigger a specific countermeasure, instead of treating every red flag the same way.

Not when

You don't have real churn or loss evidence yet — an ICP evidence log first, then an anti-ICP built from what it actually shows, beats guessing at disqualifiers from scratch.
You're tempted to just invert the ICP canvas's own sections — a real anti-ICP is disqualification rules and churn evidence, not a mirrored canvas; forcing that shape misses the hard/soft distinction that makes this genuinely useful.
You already have a recent, accurate disqualifier list elsewhere — don't duplicate an existing source of truth.

3 · How to fill it in

How to fill it in

Hard disqualifiers: What firmographic or technographic floor should end a deal outright, no matter how it's trending?
Soft red flags — proceed with a countermeasure: What risk signal predicts trouble but shouldn't kill a deal on its own — and what's the specific next step when one shows up? Not a disqualifier, a trigger for a specific countermeasure.
Process red flags: What timing or procedural pattern, independent of firmographic fit, has predicted a bad outcome before?
Known churn patterns: Why have real customers actually left — not a guess, the real reasons on file.

4 · What good looks like

What good looks like

The example below is deliberately built from real deal evidence rather than assumption, splitting genuine firm disqualifiers from softer red flags that call for a specific countermeasure — and it shows the same single-threaded-champion failure pattern appearing on both sides of the sale, once as a pre-sale red flag and again as a post-sale churn pattern.

Same example, as a downloadable xlsx workbook.

Download .xlsx

Scoping canvas

Beacon Analytics — anti-ICP

Renee Okafor, VP Sales · 2026-03-19

Hard disqualifiers

Fewer than 20 employees on a self-serve-only motion — the ICP evidence log's win/loss review found this segment essentially never converts, and support cost outweighs any realistic deal size.
No revenue operations or sales operations function at all, not even a single person — the ICP canvas's own firmographic floor exists precisely because nobody would own the tool after purchase.
Already running a dedicated RevOps analytics competitor as the system of record — a colder, slower displacement motion the ICP canvas explicitly excludes, not the same buying process as this ICP.

Soft red flags — proceed with a countermeasure

Single-threaded champion with no identified path to an economic buyer by week three — not a disqualifier on its own (the ICP canvas is explicit about this), but proceed only with multi-threading as the very next qualification step, not by continuing to build the case around one contact.
'They filled out a form' is the only trigger on file — qualification standard work already flags this as interest, not urgency; treat it as a soft flag to dig for a real trigger before advancing, not as grounds to disqualify outright.

Process red flags

A security review surfaces unplanned at week six, after the proposal was already built around a week-two close — the exact pattern that delayed Calder Industries' close in the evidence log. Not disqualifying, but replan the timeline the moment it appears instead of treating it as a minor delay.

Known churn patterns

Bought to answer a one-time board request, not a recurring need — usage never became habitual, and the renewal conversation had nothing real to point to.
The champion left the company within two quarters and no second stakeholder had ever been onboarded — the same single-threaded-champion pattern from the soft red flags above, showing up again after the sale instead of before it.
Never integrated past the initial CRM connection — the second data source that was supposed to grow the value case never got added, so the account never outgrew its original single report.

Narrative

The clearest throughline here is the single-threaded champion pattern showing up on both sides of the sale: pre-sale it's the soft red flag most likely to stall a deal, and post-sale it's the single largest driver of the churn patterns on file — a champion who leaves with no one else onboarded takes the account's only real relationship with them. That makes multi-threading the single highest-leverage intervention this list points to, worth building into qualification standard work as an explicit step rather than leaving it to sales judgement. The hard disqualifiers stay genuinely rare and firm — sub-20-employee self-serve and no RevOps function at all are floors, not scores, because the evidence behind them is this consistent.

5 · Common mistakes

Common mistakes

Treating every red flag as an automatic disqualifier, including soft ones like a single-threaded champion.
Over-disqualifying kills winnable deals that just need a specific countermeasure (multi-threading, a firmer trigger) — flattening hard and soft red flags into one list throws away exactly the distinction that makes this artifact useful.
Only capturing pre-sale loss reasons and skipping churn evidence entirely.
Churn evidence is often the clearest signal of all for what to watch for earlier — a customer who bought and later left for a specific, findable reason is more informative than a deal that just didn't close.
Running the anti-ICP check from memory instead of the actual defined disqualifiers.
Anti-ICP patterns exist because specific past deals and accounts already failed this way — a rep's general sense of 'red flags' isn't the same as disqualifiers actually derived from real evidence.

6 · What it connects to

What it connects to

upstream

ICP canvas
The ICP canvas defines who to pursue; this defines who not to, or who to pursue only with a specific countermeasure — a deliberately separate list, not this canvas's mirror image.
ICP evidence log
The same real win/loss evidence that sharpens the ICP canvas is exactly what a disqualifier or red flag here should trace back to — not a rep's unstated impression of what goes wrong.

downstream

Qualification standard work
The disqualifier check in qualification standard work is only as good as this list — a stale or missing anti-ICP means that step has nothing real to run against.

7 · Where AI helps

Where AI helps

Judgement — stays yours

Deciding whether a red flag is genuinely a hard disqualifier or a softer one worth a specific countermeasure
Deciding whether a churn pattern is specific enough to act on, not just a vague impression of why an account left

Analysis — AI helps

Drafting each section's bullets from a rough description or real loss/churn notes
Flagging a section that's too thin to mean anything yet, rather than padding it

Drudgery — automated

Formatting the canvas and keeping every section on one page
Exporting to xlsx in the house format

9 · Rate this kata