What goes in an ideal customer profile template?
An ideal customer profile (ICP) template describes the company that buys fastest and stays longest, in five parts: firmographics, trigger events, problems, buying group and disqualifiers. Fill each line from evidence (won and lost deals), and every line should be something you can later turn into a search filter. Copy this:
If you have no customers yet, fill it as a hypothesis and test it with our two-week ICP test for founders with no customers. The rest of this post assumes you have at least 20 to 30 closed deals, won or lost.
How do you build an ICP from your won and lost deals?
Export every deal that closed (won or lost) in the last 12 months, tag each with the same few fields, and compare win rate and revenue per deal across groups. The group with the best mix of win rate, deal size and cycle length is your ICP; the group you never win in becomes a disqualifier.
- Export closed deals with company, industry, employee count, deal value, days from first contact to close, and the loss reason.
- Add two columns by hand: what happened at the company in the 90 days before first contact (the trigger), and the title of the person who signed.
- Group by industry and size band, then count deals, wins and revenue for each group.
- Divide won revenue by deals worked in each group. That number tells you what a deal in that group is worth before you know if it will close.
Worked example (invented company, invented numbers). Northwind Ltd sells project scheduling software. It closed 40 deals in 12 months: 12 won, 28 lost.
| Group | Deals | Won | Win rate | Won revenue | Revenue per deal worked | Median days to close |
|---|---|---|---|---|---|---|
| A. Agencies, 20 to 100 staff | 15 | 8 | 8 / 15 = 53% | 8 x $6,000 = $48,000 | 48,000 / 15 = $3,200 | 24 |
| B. Software firms, 100 to 500 staff | 10 | 3 | 3 / 10 = 30% | 3 x $9,000 = $27,000 | 27,000 / 10 = $2,700 | 61 |
| C. Any industry, 1 to 19 staff | 9 | 1 | 1 / 9 = 11% | 1 x $2,400 = $2,400 | 2,400 / 9 = $267 | 15 |
| D. Any industry, over 500 staff | 6 | 0 | 0% | $0 | $0 | no wins |
| Total | 40 | 12 | 12 / 40 = 30% | $77,400 | 77,400 / 40 = $1,935 |
Group B has bigger deals ($9,000 against $6,000), but Group A is worth more per deal worked ($3,200 against $2,700) and closes in less than half the time. So A is Tier 1, B is Tier 2, C is Tier 3, and D goes into disqualifiers.
The trigger column adds the why now. In the example, 6 of Group A's 8 wins (75%) had posted a job ad for a project or operations role in the 60 days before first contact, against 2 of its 7 losses (29%). That becomes the first trigger event in the template. The 28 loss reasons split into no decision (10), no budget (8), chose a competitor (6) and poor fit (4), which add up to 28; the 4 poor-fit losses tell you which features to list as disqualifiers.
What is the difference between an ICP and a buyer persona?
An ICP describes a company; a buyer persona describes a person inside it. You need both, and they are filled in separately: the ICP decides which accounts to go after, the persona decides who to contact there and what to say.
| Ideal customer profile | Buyer persona | |
|---|---|---|
| Describes | A type of company | A role in that company |
| Main fields | Industry, size, location, triggers, disqualifiers | Title, seniority, goals, objections, where they look for help |
| Built from | Won and lost deals | Call notes and interviews with buyers and users |
| Used for | Account lists, tiers, territory | Messages, sequences, call scripts |
The tools draw the same line. LinkedIn's Sales Navigator filter definitions (checked 5 Nov 2026) define a Persona filter by function, seniority, job title and geography, all attributes of a person. HubSpot's account-based marketing setup page (updated 6 October 2026) puts the Ideal customer profile tier on the company record, in three tiers, and puts Buying role on the contact record, with example options Decision Maker, Budget Holder and Blocker. HubSpot says "Companies that are Tier 1 should be a great fit for your products/services". For the persona half, see our buyer persona definition and example.
How do you turn an ICP into a Sales Navigator search?
Map each template line to one filter, using account filters for the firmographics and lead filters for the buying group. These are the filter names on LinkedIn's lead and account filter definitions page, checked 5 Nov 2026:
| Template line | Sales Navigator filter | Filter type |
|---|---|---|
| Industry | Industry (for accounts, child industries are included) | Account and lead |
| Employees | Company headcount | Account and lead |
| Headquarters | Headquarters location (account) or Company headquarters location (lead) | Account and lead |
| Revenue | Annual revenue (Min-Max) | Account |
| Trigger: growing team | Company headcount growth or Department headcount growth (last 12 months) | Account |
| Trigger: hiring | Job opportunities (open jobs posted on LinkedIn) | Account |
| Trigger: new in role | Changed jobs (in the last 90 days) | Lead |
| Buying group | Function, Seniority level, Current job title | Lead |
| Active on LinkedIn | Posted on LinkedIn (in the 30 days) | Lead |
| Lookalikes of won deals | Past customer (leads who worked at a company that was your customer) | Lead |
For Northwind's Tier 1, that is: Industry set to the agency industries, Company headcount set to the bands that cover 20 to 100 staff, Job opportunities on, then a lead search with Function, Seniority level and Current job title set to the roles that signed Northwind's won deals. The filter page lists no technology filter for accounts (checked 5 Nov 2026), so check tool use on the company's site or in a data tool.
For job titles, Sales Navigator supports Boolean search in the Company, title and keyword fields, with a limit of about 15 operators. Operators must be in capitals, only straight quotation marks work, and wildcard * searches aren't supported. An example title string: ("project manager" OR "head of operations" OR "operations director") NOT assistant. Disqualifiers go in as search exclusions: the help page lists Company, Geography, Seniority, Title and Industry for leads, and Geography, Industry and Company headcount for accounts.
Can you search for your ICP without Sales Navigator?
Free LinkedIn search covers part of the template. Its people search help page (checked 5 Nov 2026) lists All Filters such as Locations, Current company, Past company, School, Industry, Profile language and Keywords, plus Actively hiring as a Premium feature. That list has no company headcount filter, so firmographics like size have to be checked by hand or in another tool. Our guide to finding leads on LinkedIn without Sales Navigator shows workarounds.
Data tools use their own names for the same fields. In Apollo's people search API reference (checked 5 Nov 2026), the template lines map to these API query parameters: person_titles, person_seniorities, organization_num_employees_ranges (written as ranges like 1,10), organization_locations (headquarters), q_organization_job_titles (titles in the employer's active job postings, which is a hiring trigger) and currently_using_any_of_technology_uids, which Apollo says covers 1,500+ technologies. The same page says that endpoint displays at most 50,000 records, so a tight ICP also keeps searches usable.
Siela's Lead Finder starts from the other end: you paste your website, it suggests roles, industries, company sizes and locations, and you check that form against your template before pressing start. Its exclusions cover recruiters, freelancers, consultants, students and Open to Work profiles, plus competitors or keywords you name, which is where your disqualifiers line goes.
How often should you update your ICP?
Re-run the deal analysis every quarter, or after every 20 new closed deals, whichever comes first, and change the template only when the numbers move. Keep three tiers, as HubSpot's model does, and send outreach effort in that order.
This week: export and tag
Pull the last 12 months of closed deals and add the trigger and signer columns. Expect it to take an afternoon for 40 deals.
Fill the template from the top group
Write only what the won deals show. Anything you believe but can't count goes in a separate hypotheses line.
Build one saved search per tier
Turn Tier 1 into filters, save it, and note how many results it returns. Too many results usually means a missing trigger.
Score new leads against it
Give points for each template line a lead matches; our lead scoring framework for small teams shows a simple points model.
Frequently asked questions
How many deals do you need to build an ICP?
Twenty to thirty closed deals, won and lost, is enough to see a pattern if you keep the groups broad, such as three or four size and industry bands. With fewer, treat the profile as a hypothesis and test it with outreach. Lost deals matter as much as won ones, because they show where you should stop spending time.
What are the 5 parts of an ideal customer profile?
Firmographics (industry, size, location, stage), trigger events that make buying urgent now, the problems in the customer's own words, the buying group with the decision maker and budget holder named by title, and disqualifiers. Add the evidence behind it, meaning how many deals and which dates, so anyone reading it knows how much to trust it.
Can a startup have more than one ICP?
Yes, but rank them. Run outreach to one Tier 1 profile at a time, because messages, case examples and channels differ by segment. Keep a second profile as Tier 2 and revisit it once the first is producing deals predictably, or when won deals keep arriving from it without effort.
What is an ICP in B2B sales?
In B2B sales the ICP is a description of the type of company that gets the most value from your product and is most likely to buy, used to decide which accounts sales and marketing go after. It sits above personas, which describe the people to contact inside those accounts.
Should firmographics or triggers come first in an ICP?
Firmographics define who could buy; triggers tell you who is likely to buy now. Start with firmographics to set the boundaries, then rank accounts inside them by trigger. A matching company with a fresh trigger, such as a new job ad or a leader who just changed roles, should be contacted before a perfect fit with nothing happening.
Sources
- LinkedIn Help: Sales Navigator lead and account search filter definitions (checked 5 Nov 2026)
- LinkedIn Help: Using Boolean Search on Sales Navigator (checked 5 Nov 2026)
- LinkedIn Help: Search Exclusions in All Filters search (Sales Navigator) (checked 5 Nov 2026)
- LinkedIn Help: Search for people on LinkedIn (checked 5 Nov 2026)
- Apollo API: People API Search (checked 5 Nov 2026)
- HubSpot Knowledge Base: Set up account-based marketing in HubSpot (updated 6 Oct 2026, checked 5 Nov 2026)