What is an ICP? Ideal Customer Profile

What is an ICP? Learn how to define an ideal customer profile, how it differs from a buyer persona, and how to build a lead list with a template and example.

What is an ICP? An ideal customer profile is a description of the companies best suited to your product, based on the problem they need solved, their ability to buy, and their ability to succeed with it. In B2B sales, it guides account selection using criteria such as industry, company size, geography, and exclusions.

Why an ideal customer profile matters

An ICP gives a founder, sales rep, or agent a shared rule for deciding which companies belong on a prospect list. Without one, each search can reflect a different interpretation of the market. With explicit criteria, you can repeat a search and explain why an account was included or excluded.

It also separates account fit from timing. A company may fit your product but have no reason to buy today. A recent funding round may justify research, but it does not establish a need. Use funding signal outreach to explore that distinction rather than treating funding as automatic qualification.

ICP vs buyer persona

The ideal customer profile describes the account: its business model, industry, size, location, needs, and constraints. A buyer persona describes the individual inside that account: their job title, responsibilities, buying authority, objections, and priorities.

For example, “UK software companies with 51–200 employees that need to coordinate customer onboarding” describes an account segment. “Head of Customer Success responsible for onboarding handoffs” describes a buyer persona within it. You need both to turn a company list into a useful contact list. A matching title alone does not establish company fit.

How to build an ideal customer profile: a step-by-step template

1. Start with evidence of customer fit

Review customers that adopted the product, stayed, and received value. Compare them with lost deals and customers that struggled. Record the problem, buying context, adoption requirements, and reasons for exclusion. If you do not yet have customers, use interviews to form a hypothesis and label it as untested.

2. Define the company criteria

Write down industry or business model, employee range, geography, and any technical or operational requirements. Separate requirements from preferences. Avoid using a broad label such as “B2B” when the actual constraint is a particular sales process or implementation need.

3. Record the problem and buying context

Describe the problem in the buyer's terms. Note which events might make it urgent, who owns it, and what must be true for adoption. Revenue, technology stack, or funding stage can be useful criteria only when they explain fit and you can obtain reliable evidence for them.

4. Set exclusions and an unknown-data rule

Specify accounts you will not pursue and why: unsupported regions, incompatible requirements, or a business model your product does not serve. Decide which missing fields require research. Missing evidence should remain unknown rather than becoming an assumed match.

5. Add the buyer persona and test the segment

Choose the roles to contact within matching accounts. Review a small sample before expanding the list. Track which criteria predicted useful conversations and actual adoption, then date the next revision. Keep targeting changes separate from changes to the message so the results remain interpretable.

Use this copyable ICP template:

  • Target company: industry, business model, employee range, and geography.
  • Problem: the specific job the company needs to accomplish.
  • Required conditions: operational or technical requirements for a useful fit.
  • Buying context: events worth researching and evidence needed to confirm urgency.
  • Exclusions: company types or requirements you cannot serve.
  • Buyer persona: responsible role, priorities, and buying authority to verify.
  • Evidence: where each criterion comes from; what stays unknown.
  • Review: owner, version date, and evidence that would change the profile.

Ideal customer profile example

This is an illustrative ICP for a fictional onboarding software vendor, not a Coil customer or a measured result.

  • Target company: UK B2B software businesses with 51–200 employees.
  • Problem: customer onboarding requires handoffs between sales, implementation, and customer success.
  • Required conditions: a dedicated customer success team and a repeatable onboarding process.
  • Buying context: hiring implementation staff is a reason to investigate workload; an interview must confirm the problem and priority.
  • Exclusions: consultancies delivering bespoke projects and businesses without a customer success function.
  • Buyer persona: Head of Customer Success; confirm who owns budget during qualification.
  • Evidence: company and role information from prospect research; process and urgency from a conversation. Unconfirmed requirements remain open questions.
  • Review: the founder owns the profile and revises it when discovery or adoption evidence contradicts these assumptions.

A UK software company with 80 employees and a Head of Customer Success passes the initial search criteria. It still needs research to establish its onboarding process. A consultancy with the same size and title fails the business-model criterion. This distinction keeps a search match from becoming a sales-qualified account by default.

How to turn an ICP into a lead list with Coil

Translate the searchable parts of your ICP into inputs, then keep the remaining qualification questions visible. Coil is an agent-operated outbound platform: humans use the web app, while agents use the REST API, JSON CLI, and runtime skills against the same organization-scoped state. Recipes are reusable workflows; runs record their durable executions.

1. Map criteria to supported search fields

The system recipe Prospect Search accepts job titles, locations, a company-size range, industry inclusion or exclusion, and a fetch count. It runs through Apify using Atomus Leads Finder. Use one finite size bucket or adjacent buckets forming one contiguous range. Business process, urgency, and budget need separate evidence.

2. Choose a source

Use Prospect Search for structured targeting, or the Sales Navigator Search system recipe for an existing LinkedIn Sales Navigator search URL. Set geography in that Sales Navigator search. Neither path turns every sentence of an ICP into an executable filter.

3. Review limits before running

Start with a bounded sample and review provider spend approval. Human admins manage credentials and recipe publication; an agent can draft a workflow and run permitted published work.

4. Inspect the imported leads

Review run status, imported records, and skipped duplicates. Use lead enrichment or direct research for missing evidence. Keep fit, excluded, and needs-review decisions distinct; a qualification label alone does not delete a lead or prevent delivery.

5. Review before handoff

Inspect the selected records before exporting or handing them to configured sending infrastructure such as SmartLead. Sending remains in the external provider.

For the broader process of connecting targeting, research, and execution, see agent-driven lead generation.

Common ICP pitfalls

Writing from intuition alone. Treat the initial profile as a hypothesis until customer evidence supports it. Include poor-fit accounts in the comparison so the definition explains exclusion as well as inclusion.

Making the profile too broad. A description that includes most of your market gives an operator little direction. Narrow it using the problem and adoption requirements, rather than an arbitrary target list size.

Filtering on unavailable data. If your source cannot return a required field, add a research step or mark the account for review. Do not claim a technology requirement has been checked when only industry and size were available.

Letting the definition drift. Save the criteria and revision date alongside your workflow inputs. An agent needs explicit criteria just as a human operator does; a changing prompt makes results harder to compare.

Related concepts

Cold outreach uses the account selection and buyer context defined by the ICP. RevOps can own the definition and its review process; in a small team, the founder often takes that responsibility. A clear profile gives both people and agent-driven outbound workflows a concrete basis for research and qualification.

Frequently asked questions

What is an ICP in sales?

An ICP, or ideal customer profile, describes the companies best suited to your product. In B2B sales, it sets account selection criteria such as industry, employee count, geography, business problem, and exclusions. Sales teams use it to decide which accounts to research and contact.

What is the difference between an ideal customer profile and a buyer persona?

An ideal customer profile describes the company. A buyer persona describes a person inside it: their role, priorities, buying authority, and objections. Use the ICP to select accounts and the persona to select contacts and shape the message.

How specific should an ICP be?

Specific enough that two people reviewing the same account can reach the same fit decision. Define required criteria, exclusions, and what evidence counts. If a field is unavailable, mark it for review rather than treating missing data as proof of fit.

How do you build an ICP without customers?

Start with a hypothesis based on buyer interviews, the problem your product solves, and the companies able to adopt it. Label assumptions, test a small prospect segment, and revise using qualification conversations and product usage. Do not present an untested segment as a proven customer pattern.

When should an ideal customer profile change?

Review it when customer evidence or your product changes. Compare retained customers, lost deals, and poor-fit accounts, then revise the criteria that no longer explain fit. Keep a dated version so you can distinguish a targeting change from a change in outreach performance.

Back to glossary