An ideal customer profile is not a fictional company everyone likes. It is a shared rule for deciding which accounts deserve time, budget, and personalized execution.
Salesforce's 2026 ICP guidance recommends using real CRM data:revenue, sales cycles, product usage, retention, and customer feedback:rather than assumptions. HubSpot similarly defines an ICP as the company or customer type most likely to buy, succeed, and generate long-term value.
The output of an ICP development workshop should therefore be testable. A new account either meets the rule, does not meet it, or lacks evidence the team must collect.
Who should attend
Include people who see different parts of customer reality:
- sales;
- marketing;
- customer success;
- product;
- revenue operations or data;
- finance when margins or service cost vary materially.
Send the source pack before the workshop. It should contain customer-level outcomes, not only logo lists and anecdotes.
The source pack
Prepare a consistent window of evidence:
- closed-won and closed-lost accounts;
- time to first value and product adoption;
- retention, expansion, contraction, and churn;
- gross or contribution margin where available;
- sales cycle and implementation effort;
- support or service intensity;
- original buying trigger and use case;
- disqualification and loss reasons.
Remove test records and accounts whose results are not comparable. Label missing fields; do not convert missing evidence into a zero.
A 90-minute ICP workshop agenda
1. Define value: 10 minutes
Agree on what "ideal" means for this product and stage. Possible dimensions include customer outcome, retention, expansion, margin, sales efficiency, and reference value. Rank them so the team cannot redefine success for each logo.
2. Review the evidence: 20 minutes
Compare top-performing and poor-fit cohorts. Look for characteristics that explain outcomes, not merely attributes that are easy to buy from a data provider.
Useful characteristics include:
- operating problem and urgency;
- company stage and change event;
- technical or process prerequisites;
- budget and buying authority;
- geography or regulation;
- implementation capacity;
- evidence of the desired outcome.
3. Write positive and negative fit: 20 minutes
For each dimension, write both sides.
| Dimension | Positive fit | Exclusion or risk |
|---|---|---|
| Problem | measurable pipeline gap the offer solves | generic desire for "more leads" with no owner |
| Readiness | systems and team can use the result | no capacity to act on qualified demand |
| Economics | expected value supports acquisition and service cost | customization destroys contribution margin |
| Timing | observable trigger creates a decision window | no change event or committed date |
Negative fit is not an afterthought. It protects customer outcomes and prevents the team from forcing weak accounts through the funnel.
4. Separate account, buying group, and signal: 15 minutes
The ICP defines the kind of account that can succeed. A persona describes the people involved. A signal describes why the account may care now.
Keep those layers separate:
- account fit: durable characteristics and prerequisites;
- buying group: roles, responsibilities, objections, and authority;
- signal: current event or behavior that changes priority.
This separation follows Tenbound's Market, Signal, Message, and Motion standard.
5. Build the score: 15 minutes
Turn the profile into a transparent decision rule. Use a small number of fields with documented sources. Weight only factors with an evidence-based reason.
Include three outcomes:
- target now: fit and timing evidence are present;
- nurture or research: fit may be present but timing or evidence is
missing;
- exclude: a documented disqualifier applies.
Do not let an opaque score overrule an explicit exclusion.
6. Assign validation: 10 minutes
Choose owners for:
- CRM definitions;
- enrichment source quality;
- customer interviews;
- quarterly cohort analysis;
- score changes;
- exception approval.
Record the date, version, and evidence behind every change.
The workshop deliverable
End with one page containing:
- target outcome and economic definition;
- positive account characteristics;
- exclusions and risks;
- buying triggers;
- buying-group roles;
- required evidence fields and sources;
- score and routing outcomes;
- review owners and next validation date.
The document is not finished when everyone agrees. It is finished when the profile can select a cohort and the team can compare its accepted pipeline, retention, and customer outcomes against a control.