Outbound resource

An ideal customer profile template for B2B outbound research

Define which companies fit your offer, record the evidence for each account and agree when a research list is ready. Copy the four blank worksheets below into your team's document or CRM. No email or download is required.

Editorial update:

Use the template ↓Discuss your outbound plan →

Copy the B2B ICP worksheets

Start with the segment brief, repeat the evidence record for each account, then use the acceptance record before handover. Replace every bracketed field. These are original planning conventions, not a universal qualification standard.

  1. 1. Segment criteria brief

    Segment / version: [name, date, owner] Offer and supported workflow: [what we can deliver] Account scope: [business model and supported markets] Must-have criteria: [requirement, reason, acceptable evidence] Preferences: [useful attributes that are not hard requirements] Hard exclusions: [condition and delivery reason] Buying roles to research: [workflow owner, user, decision participant] Unknown-data rule: [what requires review and who resolves it] Approval: [delivery owner, sales owner, date]

  2. 2. Account evidence record

    Account / website: [name and URL] ICP version: [approved brief] Criterion checked: [repeat this entry for each requirement] Observed fact: [what the source actually states] Source / observation date: [URL or permitted internal record, date] Result: [met / not met / unknown / conflicting] Open question and next check: [question, action, owner] Account decision: [include / exclude / review, reason] Reviewer / decision date: [name, date] Contact research: [role evidence, separate from account fit]

  3. 3. Research-list acceptance record

    List / campaign / ICP version: [identifiers] Reviewed accounts: [count and selection method] Decision totals: [include, exclude, review; reconcile to reviewed count] Missing or conflicting evidence: [accounts, criterion, next action] Duplicate accounts checked: [method and exceptions] Contact restrictions checked separately: [owner and record] Unresolved accounts: [held for research; not silently approved] Reviewer disagreements: [rule clarified and accounts rechecked] Handover decision: [accepted / returned, reason, owner, date] Next review trigger: [offer change, evidence change or scheduled review]

  4. 4. Reviewer disagreement and rework log

    Review batch / ICP version: [references] Sample selection method and limits: [how accounts were chosen] Account and disputed criterion: [reference] Reviewer A decision / evidence: [include, exclude or review; source] Reviewer B decision / evidence: [include, exclude or review; source] Cause: [missing source / stale source / ambiguous rule / entity mismatch / other] Resolution owner and next check: [person, action, date] Final decision and reason: [supported finding or still unresolved] Rule clarification / new version: [if needed] Other records affected: [search rule, count or unknown] Rework completed and checked by: [person, evidence] Release decision: [accept stated scope / hold affected records / return batch]

Copyable worksheets
4
Research outcomes
3
Illustrative example
1

Copy these fields into your ICP worksheet

Scroll the table sideways to compare all columns.

Copy these fields into your ICP worksheet
FieldWhat to recordHow to check it
Offer and use caseThe workflow and supported changeConfirm with product or delivery owner
Account scopeRelevant business model and operating marketsCompany website and current public material
Required capabilityIntegration or operational prerequisiteVerified source or discovery question
Fit evidenceObserved fact, URL and dateRead the original source
Buying rolesUser, process owner and decision participantRole research, then confirmation
ExclusionsConditions that prevent useful deliveryWritten reason linked to the offer
UnknownsInformation still neededNamed research or discovery action
Decision and ownerInclude, exclude or review; reviewer nameCheck evidence against the agreed criteria

Define an account, then identify people

The ideal customer profile describes an organization that could benefit from your offer and that you can serve. A buyer persona describes a person involved in that decision. Keep the two separate: a matching job title does not rescue an account outside your service territory, and a fitting company can still have the wrong contact.

Start from delivery fit

List the prerequisites for a successful engagement before adding demographic filters. A product might require a particular integration, transaction volume or implementation owner. Use only prerequisites that your delivery team can explain. Size and funding can help narrow research, but neither proves that the organization has the problem you solve.

Let unknown stay unknown

For each account, store the source and observation date beside the finding. Missing data should produce a research action rather than a made-up value. Choose one of three outcomes: include, exclude or review. This makes the list auditable and prevents an AI-generated score from concealing a critical missing requirement.

Calibrate reviewers before accepting the research list

Have two reviewers independently apply the same ICP version to a small batch before comparing decisions. Include ordinary records as well as records with missing facts, possible exclusions and ambiguous company identities. Record how the sample was selected: a deliberately difficult sample is useful for finding rule problems, but its error rate is not an estimate for the whole list. Agree the scope and review effort with the owner rather than treating a fixed sample size as universally sufficient.

  • Compare the criterion and source behind each disagreement, not just the final labels. Matching decisions can still share the same unsupported assumption.
  • Separate a research mistake from an unclear rule. Correct the source or entity match for the former; ask the delivery owner to clarify the latter.
  • Keep unresolved accounts out of the accepted handover. Record who will check the missing fact and when.
  • If a rule changes, identify other records classified under that rule and recheck the affected set. Preserve the previous decision and ICP version for traceability.

Worked disagreement: a preference mistaken for a requirement

Fictional example using the dispatch-product brief below: Reviewer A includes a single-depot operator because all required capabilities are confirmed. Reviewer B excludes it because the company has only one depot. The brief makes multiple sites a preference, so the delivery owner confirms inclusion and clarifies the wording. Search the reviewed batch for other single-depot exclusions and recheck them. This resolves a rule-interpretation problem; it does not establish buyer interest or show that the remaining list is error-free.

How to decide: include, exclude or review

Use these definitions consistently across your research team. Account fit is not proof of interest, budget, contact permission or a qualified sales opportunity.

  • Include: evidence supports every must-have criterion and no hard exclusion applies. Research the buying roles separately before choosing a contact.
  • Exclude: a verified fact fails a must-have criterion or meets a hard exclusion. Record the failed rule and source; do not reject an account just because a preferred attribute is missing.
  • Review: a required fact is missing, conflicting or too old for the decision. Name the next check and its owner. Keep the account out of the accepted list until that check resolves the decision.

Worked ICP example: a hypothetical dispatch product

Fictional planning example only. The accounts, product constraints and evidence below are invented; they describe no LeadFlow client or measured market demand.

Assume a dispatch product supports UK delivery operators with a central dispatch team and a CSV order import. Multi-site operations are a preference. A mandatory integration the product cannot support is a hard exclusion.

  • Account A — include: an imaginary discovery record confirms UK operations, central dispatch and acceptance of CSV import; no mandatory unsupported integration is reported. One depot does not disqualify the account because multiple sites are only a preference. Buying interest is still unknown.
  • Account B — exclude: an imaginary requirements record says a live integration unavailable in the product is mandatory. A fitting company size cannot override this delivery constraint.
  • Account C — review: an imaginary website lists several UK depots but says nothing about dispatch ownership or order import. Assign the researcher to confirm both requirements; the location count alone cannot justify inclusion.

AI prompt for checking an ICP brief

Review this ICP and supplied evidence only: [paste brief and sources]. Return a table of criterion, supporting evidence, unknowns and proposed next check. Separate account criteria from person-level criteria. Flag contradictory rules. Do not invent revenue, headcount, technology, pain points or purchase intent. Do not turn missing information into an exclusion unless the written rule requires it. End with questions for the human owner, not a fabricated confidence score.

Who this helps

  • Founders narrowing their first outbound segment
  • RevOps teams standardizing research fields
  • Agencies agreeing list acceptance criteria with a client

Work through the process

  1. 01

    Write the offer boundary

    Ask delivery what must be true for the engagement to work, including conditions that make an account unsuitable.

  2. 02

    Review known accounts

    Compare accounts you served successfully with poor-fit accounts. Record the reasons rather than copying every attribute of a favorite client.

  3. 03

    Trial the worksheet

    Have researchers independently classify a small sample. Discuss disagreements and clarify the ambiguous criteria.

  4. 04

    Version the ICP

    Record the date, owner and reason for each change. Keep campaign cohorts tied to the criteria used when the list was approved.

Questions and answers

Is an ICP the same as a buyer persona?
No. This worksheet describes account fit. Map people and responsibilities inside the fitting accounts separately so your message reaches someone who owns or influences the workflow.
Do I need a numerical fit score?
Not to start. Include, exclude and review decisions with explicit reasons are easier to audit. If you later add scores, retain hard exclusions and show how each input affected the result.
Can I use this ICP template in a spreadsheet or CRM?
Yes. Copy the segment brief into a shared document and use the evidence-record fields as spreadsheet columns or CRM fields. Keep repeated criterion checks linked to the account. Preserve the source, observation date and ICP version when you export or hand over a list.
What if we have no customers yet?
Label the first segment as a hypothesis. Start with delivery constraints and what you learn in research conversations, rather than inventing a best-customer pattern. Keep assumptions separate from observed facts and revise the brief as evidence arrives.
When should we change the ICP?
Review it when delivery constraints, the offer or evidence from real conversations changes. Do not widen filters solely because a small campaign has not produced meetings.

Sources and editorial notes

Original AI-assisted editorial templates and planning conventions prepared for LeadFlow. Examples are illustrative; they are not independently tested results, customer evidence or performance benchmarks. Check facts and adapt the material to your own workflow before use.

Keep reading