An account can be a perfect fit and show no current interest. Another can generate a lot of activity while being impossible for you to serve.
That is the basic distinction between ICP fit and intent:
- Fit describes whether a company and person resemble the customers you can help successfully.
- Intent describes evidence that they may be paying attention to a relevant problem, category or product now.
Treating them as one score too early creates bad priorities. Activity can hide poor fit, and firmographic similarity can be mistaken for readiness.
What is ICP fit?#
ICP fit measures how closely an account matches the conditions under which your product creates value and your business can serve it well.
Company-level criteria may include:
- Industry or business model
- Geography and regulatory environment
- Company or team size
- Revenue motion and operational maturity
- Technology or integration requirements
- Problem frequency and severity
- Ability to implement and sustain the product
Person fit adds role, responsibility, seniority and involvement in the problem.
A useful ICP is not a description of everyone who has ever paid you. It identifies repeatable conditions for success and meaningful exclusions.
What is intent?#
Intent is inferred from behaviour or change.
Examples include:
- Relevant category research
- Product, pricing or trial activity
- A direct request for information
- Public discussion of a problem
- Repeated engagement with category content
- An organisational event that creates a new evaluation window
Intent varies in strength and identity resolution. Account-level topic research is different from a known stakeholder requesting implementation detail. The guide to buyer intent signals explains those levels.
The fit-intent matrix#
| Low intent | High intent | |
|---|---|---|
| High fit | Monitor, educate and build familiarity | Prioritise for evidence-based action |
| Low fit | Ignore or exclude | Investigate briefly, then suppress if the mismatch is real |
High fit, high intent#
This is the obvious priority, but still inspect the evidence. Confirm the relevant person, timing, existing ownership and appropriate next action.
High fit, low intent#
These accounts belong in market education, relationship-building or monitoring. Forcing immediate outbound may burn a good future account for no present reason.
Low fit, high intent#
This quadrant creates many false positives. The activity could come from a student, consultant, competitor, tiny team, unsupported geography or use case you cannot solve.
High activity may justify checking whether your ICP assumption is wrong. It should not automatically override known delivery constraints.
Low fit, low intent#
Exclude the account. More enrichment and messaging will not create relevance.
Keep separate component scores#
Instead of one mysterious lead score, expose the components.
Company fit: 0–5#
- Serviceable market
- Problem prevalence
- Operational maturity
- Commercial suitability
- Exclusion checks
Person fit: 0–3#
- Relevant function
- Appropriate responsibility
- Plausible role in the decision or workflow
Intent: 0–5#
- Relevance of observed behaviour
- Strength and specificity
- Person- or account-level resolution
- Recency
- Independent corroboration
Relationship and timing: -3 to +3#
- Existing customer or opportunity
- Previous champion
- Recent contact
- Opt-out or suppression
- Change window or deadline
A routing rule can then require minimum fit before intent affects priority.
Fit is not static#
Company fit can change.
A business may enter a serviceable market, adopt a required platform, create a new team or cross a maturity threshold. It may also shrink, change strategy or remove the workflow your product supports.
Store the evidence and effective date behind fit fields. “Employee count: 200” without a source or timestamp is less useful than it looks.
The same applies to person fit. A contact may change role or company. Revalidate identity at the point of action.
Intent is not a personality trait#
Intent expires. Do not permanently label an account “high intent” because of behaviour from last quarter.
Use event-level records with:
- Signal type
- Effective time
- Source
- Entity level
- Expiry window
- Evidence strength
Then calculate the active intent view from current evidence. Keep old events for history without presenting them as present readiness.
Example decisions#
A perfect account with no signal#
A 300-person B2B software company has the right team, systems and use case, but no recent activity.
Decision: monitor key changes, provide useful content and let account strategy—not fabricated urgency—determine contact.
Strong research from a poor-fit account#
A five-person consultancy consumes several pages and downloads a guide, but your product requires a substantial internal sales team.
Decision: confirm whether a relevant use case exists, then route to self-serve, partner or suppression rather than enterprise outbound.
Moderate fit with strong person-level evidence#
A company is near the lower size boundary, but its RevOps leader describes the exact workflow problem and is hiring a team to handle it.
Decision: human review. The evidence may reveal that the practical ICP is broader than the current firmographic rule.
High fit with negative timing#
A target account shows topic interest but recently renewed an incumbent contract and requested no sales contact.
Decision: suppress. Fit and positive intent do not override explicit boundaries.
How to operationalise the model#
Define fit from customer evidence#
Study successful, unsuccessful and churned customers. Identify conditions that affect value and delivery, not merely attributes that are easy to query.
Create hard gates and soft preferences#
Unsupported geography may be a hard exclusion. Preferred company size may be a soft score. Document the difference.
Map signals to entity level#
Label activity as person, account or anonymous. Do not attribute account behaviour to a named buyer without evidence.
Apply recency and negative rules#
Expire intent and carry opt-outs, ownership, opportunity state and contradictory events.
Route by quadrant#
Create different destinations for prioritised review, monitoring, nurture, self-serve and suppression.
Learn from overrides#
When a rep accepts a moderate-fit account or rejects a high-score one, capture why. Review patterns rather than arguing about isolated examples.
Metrics that improve the model#
Track outcomes by fit and intent bands:
- Reviewer acceptance
- Positive reply
- Meeting creation
- Opportunity conversion
- Win rate and sales cycle
- Implementation success and retention
- False-positive and suppression rate
A scoring model that creates meetings but produces poor customers may be optimising intent while ignoring fit.
Use intent to order fit accounts#
The cleanest rule is:
Fit determines eligibility. Intent and timing determine priority.
There will be exceptions, and those exceptions should teach you. But keeping the concepts separate makes the system easier to explain, measure and correct.
The goal is not to find the most active account. It is to find the strongest intersection between a customer you can help and evidence that attention is useful now.
TWL Signals
Turn market attention into a qualified prospect feed.
Monitor the people and companies your buyers already follow, keep the original context, enrich the profiles and route only the strongest matches into outreach.
Start building your signal workflow