- Build the ICP Around Evidence You Can Check Before Outreach
- Keep the Company Profile Separate From the Market and the Person
- Turn the Customer Sample Into Company Facts You Can See
- Test Which Customer Traits Can Become Pre-Contact Fields
- Let the Usable Profile Shape Targeting and Routing
- Frequently asked questions
-
Explore OKKI Go
An ideal customer profile becomes usable when its account attributes can be checked before first contact, or when a team has named a reliable proxy for each one. Traits visible only after purchase may help explain past wins, but they cannot decide which accounts enter a prospecting list.
Build the ICP Around Evidence You Can Check Before Outreach
An ideal customer profile starts from a timing constraint. A team must choose which companies to study before it knows them as customers, so the information doing that work must exist before contact. Salesforce describes an ICP as a description of best-fit prospective companies and points to firmographic, technographic, environmental, and behavioral attribute classes, checked September 2026. That gives us the company-level object. It does not make every attractive characteristic usable. A characteristic becomes part of a pre-contact profile only when someone can find it on the company before the choice to pursue that company has already been made. The same principle holds whether the profile lives in a research note, a CRM, or OKKI Go.
The next step follows from that constraint. Customer history can suggest what to look for, but it cannot change the past moment at which a new company must be understood. Begin with characteristics of good customers, then ask whether each characteristic was visible while the company was still an unknown prospect. If it was, it may help form a usable profile. If it was not, it remains a description of a past relationship until a reliable pre-contact stand-in can be named.
A Past Win Is Evidence, Not Yet a Screening Rule
Existing customers are the right place to form hypotheses. Look across the accounts that produced the kind of work the business wants more of, then distinguish their company conditions from the qualities of the relationship that unfolded later. A high-value renewal, a smooth implementation, or an unusually collaborative champion may explain why an account was valuable in hindsight. None automatically tells a team what it could have known before first contact. The productive question is narrower: which company facts were already visible, and which later observations merely help us understand the sample?
Keep the Company Profile Separate From the Market and the Person
A target market names the broad territory in which the business may look. An ICP asks a more specific question about a company within that territory. A buyer persona asks about a person. HubSpot's ICP template treats the target as a company and uses attributes such as industry, size, budget, and buying process, while placing the buyer persona at the individual level, checked September 2026. The distinction is not terminological housekeeping. It tells us what kind of evidence a profile may use before contact: company evidence can define the account; a person's motivations and goals cannot be assumed from it.
- Target market: the broad commercial space in which a company may belong.
- Ideal customer profile: the observable company conditions that make a prospect comparable to the customer sample.
- Buyer persona: a person-level understanding of motivations, needs, and behavior.
Keep Person-Level Questions Outside the Company Profile
This boundary also gives us a useful counterexample. A rich description of a person, such as a preferred title, a motivation, or a likely objection, can be useful later without becoming a company condition. Salesforce's buyer-persona guidance focuses on the motivations, goals, pain points, and behavior of people, checked September 2026. Paired with HubSpot's company-level ICP definition, it shows why a vivid persona is not proof that a company belongs in a pre-contact profile. The same is true of a rich description of a customer relationship. It may illuminate what happened after purchase, but it does not automatically supply a fact that was available before the relationship began.
Turn the Customer Sample Into Company Facts You Can See
Once the company and person are separated, the customer sample becomes easier to use. Look for company facts that recur across the accounts you want to understand, then ask whether those facts were visible at the earlier point when the companies were still prospects. Industry, operating model, geography, public technology signals, and information in a company's own materials may be relevant candidates. Salesforce lists firmographic, technographic, environmental, and behavioral attributes in its ICP framing, checked September 2026. The source supplies attribute classes, while the team still has to decide which facts are meaningful and findable in its own market.
Translate each promising pattern into a company fact, rather than preserving it as a customer story. If a pattern is a particular industry, describe the industry. If it is a visible operating model, describe the public sign that would reveal it. If it is a technology condition, state what can be checked. The aim is not to make every field a universal predictor. It is to make the profile intelligible at the only time it can guide targeting, when a company is still being understood from the outside.
Make Customer Patterns Comparable Across New Companies
Comparison is the discipline that prevents a profile from becoming a collection of flattering adjectives. A company fact should mean the same thing when two researchers look at two different prospects. Name the fact, name where it can be seen, and preserve the customer observation that first made it interesting. That leaves the team able to revisit the pattern when the market changes or when new customer evidence makes the original interpretation less convincing.
Whether the profile is maintained in a spreadsheet, a CRM, or OKKI Go, its usefulness comes from that shared meaning. No tool makes a trait observable simply by storing its name. The work is in distinguishing a company fact from a retrospective interpretation, then retaining enough context that a later reader does not mistake a loose label such as good fit for a fact about the company.
Test Which Customer Traits Can Become Pre-Contact Fields
Now the starting premise can be applied to each trait in the customer sample. The company-level definitions from Salesforce and HubSpot establish the object of an ICP, while Salesforce's person-level persona definition marks the boundary. They do not publish the three labels below. Those labels are an editorial field audit for turning the boundary into a usable profile. They ask only whether a trait is available when the company must be understood before contact. Phrases such as easy to implement, likely to expand, or values partnership may describe a successful relationship. Without a pre-contact company fact behind them, they cannot become pre-contact selection conditions merely because they sound relevant.
- Observable before contact: a researcher can verify the company condition from a defined, repeatable source. Keep the field and state the inspection rule.
- Needs a proxy: the desired trait is not visible directly, but a documented company-level indicator may stand in for it. Name the proxy and mark the inference as provisional.
- Known only after purchase: the trait depends on implementation, relationship quality, realized value, or another post-sale outcome. Keep it as a learning note, not as an account-list filter.
Use a Proxy Openly, or Remove the Field From the Filter
A proxy is useful only when the team can say what it stands in for and where its meaning may break down. Do not hide an inference inside a firmographic label. If a public operating signal may point to a relevant workflow, call it a proxy and keep the uncertainty visible. If no credible proxy exists, the trait belongs with what the team has learned from customers, not with what it can know about a future company. That is not a loss of sophistication. It is the boundary that keeps a rich customer description from pretending to be a usable pre-contact profile.
Let the Usable Profile Shape Targeting and Routing
After the field audit, the profile can shape targeting without pretending to know more than it does. It can help a team focus on companies whose observable conditions resemble the customer sample, while separating them from companies that only resemble it through a hopeful interpretation. This is not a grand score. It is a way to preserve the difference between a visible company fact, a reasonable but unproven proxy, and a trait that could only be known after a customer relationship exists. Routing becomes clearer because the profile carries those distinctions forward instead of collapsing them into a single fit label.
Imagine a team looking at companies in a target market after noticing that its strongest customers share an industry and a visible operating condition. It can look for those same company facts in the new set of prospects. A company with both facts is comparable to the sample in a way the team can explain. A company with only a possible proxy remains less certain. A company for which the relevant information emerges only after implementation may still be worth understanding, but the profile cannot honestly use that later knowledge to characterize it in advance. The point is not to produce a mechanical ranking. It is to let the quality of available evidence change how confidently the team treats the company.
Keep the Evidence Boundary When the Profile Enters a Workflow
When the profile enters a prospecting or routing workflow, preserve the distinction that created it. A visible company fact can support a confident targeting judgment. A proxy should remain visibly provisional. A post-sale trait should remain a lesson from the sample rather than being written back as if it were a known prospect fact. Before documenting the profile in OKKI Go or any other system, make sure this boundary survives the transfer. If available sources cannot show a condition reliably for a particular market or account type, the profile should acknowledge that limit instead of turning absence of evidence into a confident description.
A strong ideal customer profile is modest about what it knows. Keep the company conditions that can be checked, label proxies as inferences, and leave post-sale outcomes in the customer-learning record. That gives the next account judgment a reason someone else can inspect.
Frequently asked questions
What belongs in an ideal customer profile before first contact?
Use company-level conditions that can be observed from a defined source, or a clearly labeled proxy. Keep post-sale outcomes as learning notes until the team can translate them into an inspectable company condition.
How does an ICP differ from a target market and a buyer persona?
A target market is the broad area worth considering. An ICP is the company-level filter within that area. A buyer persona concerns the person inside a selected account, including motivations and pain points, so it should not replace the account-fit rule.
When should an ICP field use a proxy instead of a direct observation?
Use a proxy only when the desired trait cannot be observed directly but a repeatable company-level indicator can be named. Keep the inference visible, and keep the trait in the customer-learning record rather than the pre-contact profile if no credible proxy exists.
How should an ICP change prospecting and routing?
Let observable company facts support a more confident targeting judgment, while preserving proxy-only cases as provisional and post-sale traits as customer learning. The profile is useful when it changes how clearly the team can explain why a company resembles the sample before outreach.
