Lesson 2: Ideal Customer Profile and the Painful Problem

Why this level matters

In Level 1 you saw that product-market fit means a specific group of people pulling your product out of your hands. This level covers the two choices that come before any building. You decide who that group is (your ideal customer profile, or ICP) and which problem you will solve for them.

A weak choice here is expensive. If the customer is too broad, every signal you collect later will be noisy. If the problem is mildly annoying, people will praise the idea and never pay for it. The rule for this level:

Start with a problem people already spend time or money to solve.

We will use the running case throughout: an AI product for small accounting firms.

Task 1: Write a narrow ideal customer profile

An ICP describes the customer you serve best. It is not everyone who could possibly buy. "Accounting firms" is a market. "US tax-and-bookkeeping firms with 3–15 staff, run by an owner-partner, on QuickBooks Online" is a profile.

An ICP is like a rifle sight rather than a floodlight. A floodlight lights everything faintly. A sight puts all your effort on one target.

Do this: fill in the five fields below for your case.

1. Firm size: headcount and client count. Example: 3–15 staff, 150–600 clients.
2. Role: the person who feels the pain, and the person who signs the cheque. Example: the owner-partner does both; a senior bookkeeper feels the pain daily.
3. Tools: the software already in use. Example: QuickBooks Online or Xero for client books, a practice-management or client-portal tool (such as TaxDome) for requests, plus email and spreadsheets.
4. Workflow: the recurring work and when it happens. Example: monthly bookkeeping and bank reconciliation for small-business clients, plus individual returns that peak before the US federal filing deadline of 15 April. Extended returns are due 15 October.
5. Buying trigger: the event that makes them look for a solution now. Examples: losing a staff member before tax season, a backlog of unreconciled accounts, or a decision to take on more clients without hiring.

Check your ICP against this list:

• Could you find 20 of these firms by name this week?
• Would two firms that match the profile describe their work the same way?
• Does the buying trigger point to a specific moment, not "whenever"?

If any answer is no, narrow the profile further.

Task 2: List candidate problems

Write down at least three problems your ICP has. Keep each one specific and observable.

• Month-end reconciliation: matching bank transactions to the books for every bookkeeping client, every month.
• Client document chasing: repeatedly asking clients for statements, receipts and tax forms.
• Tax-season intake: collecting, sorting and checking each individual client's documents before preparing the return.

Avoid vague items such as "accountants are overworked". You cannot score those, and you cannot test them.

Task 3: Score each problem on four factors

Do this: rate each problem from 1 (low) to 5 (high) on each factor.

1. Frequency: how often does the pain occur?
2. Cost: how many hours or how much money does it consume, including errors and rework?
3. Urgency: is there a deadline or penalty that forces action now?
4. Current workarounds: how much do they already do or spend to cope? Examples: paid tools, extra staff, overtime, home-made spreadsheets. A high score is good news. Effort spent on workarounds proves the pain is real.

An example scorecard (your own scores will come from evidence, not guesses):

The total is a starting point, not a final answer. Read the pattern behind it:

• High frequency, moderate urgency (reconciliation) means steady daily use, which helps retention later.
• Low frequency, high urgency (tax-season intake) means intense demand for a few months and a quiet product for the rest of the year.
• Strong existing workarounds (document chasing) prove the pain is real. They also mean competition. Portal tools such as TaxDome already offer automated client requests and reminders, so your product has to beat that, not just beat email.

If two problems tie, choose the one with the higher frequency. Frequent pain creates habits.

Task 4: Choose one problem and write your kill criteria

Pick one problem to test. Testing two at once splits your attention and blurs your evidence.

Do this: before you collect any evidence, write down what would make you drop the problem. Doing this first stops you from explaining bad news away later.

Example kill criteria for month-end reconciliation:

1. Fewer than half of the ICP firms you contact report spending more than a set number of hours per client per month on it.
2. Most firms say their accounting software's built-in matching already handles it well enough.
3. No firm can name money or staff time it currently spends on the problem.
4. The owner-partner, who signs the cheque, ranks it below other problems.

Record this in one line: "We will test [problem] for [ICP]. We will drop it if [criteria]."

The experiment

1. Write your five-field ICP.
2. List three or more candidate problems.
3. Score each one on frequency, cost, urgency and workarounds.
4. Choose one problem and write your kill criteria.

How to read the result: a strong candidate scores high on frequency or urgency and on workarounds. If your top problem scores low on workarounds, people are not yet spending anything to fix it, so treat it as risky. If you cannot write kill criteria, the problem is still too vague. Go back to Task 2.

Key takeaways

• An ICP covers firm size, role, tools, workflow and buying trigger. Keep it narrow enough to name 20 real firms.
• Score problems on frequency, cost, urgency and current workarounds.
• Existing workarounds are evidence of real pain, and also of competition.
• Frequent problems build habits; urgent, seasonal problems create spikes.
• Test one problem at a time, and write kill criteria before you collect evidence.