Ideal Client Profile
When To Use
You say yes to every client at any price, and the worst ones take the most time. This answers one question: in what situation does someone need exactly you, and what tells you early that an engagement will go wrong?
When Not To Use
If you already know who you are for and nobody can repeat why you over the alternatives, go to Positioning Statement. If you have fewer than three finished engagements, write a first guess and revisit it after the next few.
Inputs
- 3 to 5 of your best past engagements and 2 or 3 of your worst, in your own words: what was happening when they called, what you did, how it went. No client names needed.
- Anything the client said about why they hired you, from your notes or emails.
If you have none of this, I start from two engagements you remember well, one good and one bad, and mark the output as a first draft.
Approach
Jobs to Be Done, as the Christensen Institute describes it: people do not buy a service, they hire it to make progress in a particular circumstance. So the profile describes situations and the progress wanted, not a type of organisation. The failure it prevents is the profile written from imagination, a client you wish you had, which then filters out the people who actually pay you.
Workflow
- Ask at most three questions: which engagements count as best and worst for you (your own reasons, not mine), what you want more of in the next year, and any client you would never take again and why.
- For each best engagement, find the circumstance: what had just happened when they called you (a change, a deadline, a failure, a new person in a role). Write it as a trigger you could spot from outside, then look for the two or three triggers that repeat.
- Write the job in three dimensions, from what clients said and did: functional (the work that had to get done), emotional (how the buyer wanted to feel), social (how they wanted to be seen by their boss, board or team). Quote your notes where you have them; mark anything inferred as "my reading, to check".
- Map who decides as roles in the decision you observed: who felt the problem, who signed, who could block, who used the work. Roles only, never a judgement of the people in them.
- From the worst engagements, turn what went wrong into observable red flags: things you could notice in a first call or brief, such as "no one owns the budget" or "wants the outcome by a date nobody can give". You mark which flags mean walk away and which mean ask more.
- Test the draft against your best engagement and one you lost or declined. If it would have screened out a client you loved, it is wrong; adjust it with you.
Output Format
# Ideal Client Profile
Drawn from: [number] best and [number] worst engagements, described by [your name]
## Situation and trigger
| Trigger you could spot | Seen in | What it costs them to wait |
|---|---|---|
| [what had just happened] | [engagement A, engagement C] | [in their words, or "unknown"] |
## The job they hire you for
- Functional: [the work that has to get done]
- Emotional: [how the buyer wants to feel]
- Social: [how they want to be seen, and by whom]
## Who decides
| Role in the decision | What they care about | Seen in |
|---|---|---|
| [feels the problem / signs / can block / uses the work] | [from your notes] | [engagement] |
## Red flags
| Signal you can observe | Seen in | Walk away or ask more |
|---|---|---|
| [observable signal] | [engagement] | [your call] |
## Decision
[Your name] decides which red flags are walk-away rules and tests the profile on the next [number] enquiries, by [date].
Done When
- Every trigger, job and red flag traces back to a named past engagement of yours.
- The profile describes situations and roles, never a size, sector or type of organisation.
- Each red flag is something you could observe in a first call.
- The test against your best engagement passes.
Quality Bar
- Situations, not demographics: "a new leader inherited a stalled project" beats any label.
- Inferences are marked "my reading, to check"; nothing is presented as what a client said unless they said it.
- Decision roles describe a part in a decision, never a person's character or worth.
- You set which flags mean walk away; I only list what the engagements show.
- Built from your real past clients, never imagined ones.
Next
Run win-positioning-statement (Positioning Statement) to say what makes you the obvious choice for these clients.
About the makers
This pack is made by 90n (90n.ai): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).
1---2name: win-ideal-client-profile3description: Builds an Ideal Client Profile from your best and worst past engagements, naming the situation and trigger that makes clients need you, the job they hire you for, who decides and the red flags to walk away from. Use for "run win-ideal-client-profile", "who is my ideal client", "I say yes to everyone", "write my ideal client profile", "which clients should I turn down", "who am I really for", "red flags in a prospect", part of the Claude Guide for Finding Clients Pack by 90n.4metadata:5 source: https://skillmd.com/skills/polar-bear-org/win-ideal-client-profile6---78# Ideal Client Profile910## When To Use11You say yes to every client at any price, and the worst ones take the most time. This answers one question: in what situation does someone need exactly you, and what tells you early that an engagement will go wrong?1213## When Not To Use14If you already know who you are for and nobody can repeat why you over the alternatives, go to Positioning Statement. If you have fewer than three finished engagements, write a first guess and revisit it after the next few.1516## Inputs17- 3 to 5 of your best past engagements and 2 or 3 of your worst, in your own words: what was happening when they called, what you did, how it went. No client names needed.18- Anything the client said about why they hired you, from your notes or emails.19If you have none of this, I start from two engagements you remember well, one good and one bad, and mark the output as a first draft.2021## Approach22Jobs to Be Done, as the Christensen Institute describes it: people do not buy a service, they hire it to make progress in a particular circumstance. So the profile describes situations and the progress wanted, not a type of organisation. The failure it prevents is the profile written from imagination, a client you wish you had, which then filters out the people who actually pay you.2324## Workflow251. Ask at most three questions: which engagements count as best and worst for you (your own reasons, not mine), what you want more of in the next year, and any client you would never take again and why.262. For each best engagement, find the circumstance: what had just happened when they called you (a change, a deadline, a failure, a new person in a role). Write it as a trigger you could spot from outside, then look for the two or three triggers that repeat.273. Write the job in three dimensions, from what clients said and did: functional (the work that had to get done), emotional (how the buyer wanted to feel), social (how they wanted to be seen by their boss, board or team). Quote your notes where you have them; mark anything inferred as "my reading, to check".284. Map who decides as roles in the decision you observed: who felt the problem, who signed, who could block, who used the work. Roles only, never a judgement of the people in them.295. From the worst engagements, turn what went wrong into observable red flags: things you could notice in a first call or brief, such as "no one owns the budget" or "wants the outcome by a date nobody can give". You mark which flags mean walk away and which mean ask more.306. Test the draft against your best engagement and one you lost or declined. If it would have screened out a client you loved, it is wrong; adjust it with you.3132## Output Format33```markdown34# Ideal Client Profile35Drawn from: [number] best and [number] worst engagements, described by [your name]36## Situation and trigger37| Trigger you could spot | Seen in | What it costs them to wait |38|---|---|---|39| [what had just happened] | [engagement A, engagement C] | [in their words, or "unknown"] |40## The job they hire you for41- Functional: [the work that has to get done]42- Emotional: [how the buyer wants to feel]43- Social: [how they want to be seen, and by whom]44## Who decides45| Role in the decision | What they care about | Seen in |46|---|---|---|47| [feels the problem / signs / can block / uses the work] | [from your notes] | [engagement] |48## Red flags49| Signal you can observe | Seen in | Walk away or ask more |50|---|---|---|51| [observable signal] | [engagement] | [your call] |52## Decision53[Your name] decides which red flags are walk-away rules and tests the profile on the next [number] enquiries, by [date].54```5556## Done When57- Every trigger, job and red flag traces back to a named past engagement of yours.58- The profile describes situations and roles, never a size, sector or type of organisation.59- Each red flag is something you could observe in a first call.60- The test against your best engagement passes.6162## Quality Bar63- Situations, not demographics: "a new leader inherited a stalled project" beats any label.64- Inferences are marked "my reading, to check"; nothing is presented as what a client said unless they said it.65- Decision roles describe a part in a decision, never a person's character or worth.66- You set which flags mean walk away; I only list what the engagements show.67- Built from your real past clients, never imagined ones.6869## Next70Run win-positioning-statement (Positioning Statement) to say what makes you the obvious choice for these clients.7172## About the makers7374This pack is made by 90n ([90n.ai](https://90n.ai)): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).