Problem Statement: Formula, 5 Steps and Four Examples
A problem statement is one to three sentences long. It names who has the problem, when it happens, what it costs them today and how you will know it is solved. It names no solution. Write it before you pick a solution, since you will test every later proposal against it.
Formula: [Who] has [problem] when [context]. Today they [workaround], which costs them [impact]. It is solved when a metric you choose moves from its current baseline to a target you set, by a fixed date. Filled in for our example: solo dog groomers in Austin lose about 3 of 30 weekly appointments to no-shows (assumption). In our example, it is solved when the no-show rate falls from 10% to 5% of booked appointments within 90 days; both the target and the deadline are assumptions.
What a problem statement is (and isn’t)
Wikipedia defines a problem statement as “a description of an issue to be addressed, or a condition to be improved upon”1. In practice your team and outside reviewers read this paragraph first, and it decides which work counts as progress.
A plain example with no figures: “Freelance bookkeepers lose billable time at every month-end chasing clients for missing receipts by email. It is solved when time spent chasing falls below a target they set from their own logs.”
Four kinds of sentence get mixed up with it:
| Type | What it says | Example |
|---|---|---|
| Problem statement | Who is stuck, at what cost, and the finish line | Bookkeepers lose billable time chasing receipts at month-end |
| Goal | The finish line alone | Cut receipt-chasing time this quarter |
| Solution | What you will build or do | A client portal that requests receipts each week |
| Hypothesis | A testable guess about cause or fix | Clients forget receipts because nobody asks until month-end |
In academic writing the term has a second meaning. Purdue OWL describes a problem statement as “a move that a document makes to help the reader realize why the document is important”2. In a thesis or paper, it names a gap in what is known, the consequence of that gap and the question your work answers. MIT OpenCourseWare teaches it as the third step of its D.I.S. method (Describe, Inquire, State)3.
Once your statement holds up, write the value proposition next. That is where the solution belongs.
When to write one, and when you can skip it
Write a problem statement at these points:
- Before you choose a solution, so the options compete on the same target.
- Before a round of customer interviews, so you know which answers would prove you wrong.
- When you open a project. PMI frames the statement as a move from the current state to a future state by a target date4, which fits the first page of a charter.
- For the need or significance section of a grant application (see the FAQ on federal grants).
- In the introduction of a thesis or research paper2.
- When a feature request is disputed.
- For the problem section of a business plan.
Skip it when the cause and the fix are already known and cheap. A broken checkout button needs a ticket.
The parts of a problem statement
A complete statement has six parts. Each answers one question, and each has a weak version that reads fine and says nothing.
| Part | Question it answers | Weak → strong |
|---|---|---|
| Who | Which role or group, in which market, at what moment? | “Small businesses” → “Solo groomers in Austin who rebook regulars at checkout” |
| Problem | What goes wrong for them? | “Scheduling is hard” → “Booked clients don’t show up” |
| Context | When, where and how often? | “Sometimes” → “Several times a week, on slots booked weeks ahead” |
| What they do today | What is the current workaround? | (missing) → “Texts reminders by hand the night before” |
| Impact | What does it cost in time, money or risk? | “Lost revenue” → “About $210 a week (assumption)” |
| Solved when | Which metric, from what baseline, to what target, by when? | “Fewer no-shows” → “From 10% to 5% of bookings within 90 days (assumption)” |
The five Ws map onto the first five parts. Who is Who. What is the problem. When and Where sit in the context. Why is the impact, the reason the problem is worth fixing. The five Ws give you no workaround and no finish line, so the formula adds both. The target date comes from project practice, where PMI makes it one of three required components4.
If “Who” still covers everyone, narrow the audience before you go on. For the workaround row, see what customers use today with a competitor analysis, including spreadsheets, manual texts and doing nothing.
Wikipedia’s four-part template ends in a “proposal”, the potential solution1. We leave the proposal out. A statement that contains its own fix can only be answered by that fix, and you lose the chance to compare options.
Our problem statement formula
The formula has one slot for each of the six parts. Fill the slots in order, then check the draft against the five rules below it.
[Who: a role or group, in a market, at a trigger moment] has [problem] when [context: where, when, how often]. Today they [current workaround], which costs them [impact: time, money or risk]. It is solved when a named metric moves from its starting value to a target by a set date.
The five rules:
- Who passes the three-part test. It names a role or group, the market it works in, and the moment that sets off the problem. If one of the three is missing, narrow down. Your Ideal Customer Profile (ICP) is where you work this out in depth.
- No solution words. Nothing names a product, a feature, a hire or a tool. Test: three different solutions must each be able to answer the statement.
- Every figure is confirmed or labeled “assumption”. A number with no data behind it gets the label or gets cut.
- One metric, one pass line. Give the baseline and the target, with a date. The target moves in the direction that means “better”. Shares are written as percentage points or as a relative change.
- Evidence sits under the sentence. Each evidence line records something you can check, such as a count or a log entry, and is marked confirmed or assumption.
The Problem Statement Canvas from Disciplined Entrepreneurship has eight sections, among them emotional impact and taking a step back5. We keep six parts because each one can be checked against data.
How to write a problem statement in 5 steps
Step 1. Name the problem and collect evidence
Before you have a product, use what exists:
- interviews with people who have the problem;
- reviews of the tools or services they use now;
- forum and community threads where they complain;
- your own logs, such as a calendar or invoices.
Label each evidence line “confirmed” if you saw it yourself, otherwise “assumption”. If you are unsure which pain is the real one, find the pains first and use questions that don’t lead the customer.
Output: one rough sentence, with evidence lines below it marked confirmed or assumption.
Step 2. Pin down when, where and how often
You cannot measure a problem that happens “sometimes”. Write down what triggers it, where it happens, how often, and what people do at that moment today. The trigger tells you when a solution has to show up. The workaround is your real competitor, even if it is only a sticky note.
Describing the job the customer hires a product for helps if the context keeps sliding toward your idea.
Output: one context line that includes a frequency.
Step 3. Find the cause you can act on
Ask why the problem happens, and stop when the answer is something you can test and change.
Four methods:
- 5 Whys: ask why on each answer.
- Journey mapping: lay out each step the customer takes and mark where it breaks.
- Funnel analysis: count the people who drop out at each stage.
- Affinity mapping: group interview notes by theme, then count the notes under each theme.
PMI warns against presupposing a cause or placing blame4. Keep your causes as labeled hypotheses under the statement. The sentence itself describes the condition.
Output: two or three cause hypotheses, ordered by how much evidence supports each.
Step 4. Set the target: baseline, number, date
Write the baseline first, then the target, then the date. Check the direction. A target to “reduce the share of employees who feel valued” describes a worse company, though it reads like a goal.
Spell out percentages. In our example, the no-show rate falls from 10% to 5% (both assumptions): a drop of 5 percentage points, or 50% in relative terms. In our example, a target written as “a 5% drop” (an assumed figure) could mean either one.
Test the metric before you build. A fake door test is one way to get a baseline cheaply.
Output: one metric with a baseline, a target, a date and a pass line.
Step 5. Write the sentence and cut the solution out
Fill the formula with your outputs from steps 1 to 4. Then delete every word that names a fix: “app”, “platform”, “hire”, “automate”, “dashboard”. Run the three-solution test from rule 2. If only one solution fits, the solution is still hidden in your wording. A “How might we” question can come after this step, when you move to ideas.
When the sentence passes, turn it into a concept.
Output: one to three sentences, with evidence lines underneath.
Test your draft before you share it
| Question | Fix if the answer is no |
|---|---|
| Is the statement free of product, feature, hire and tool words? | Delete them and move them to your solution list |
| Could three different solutions each answer it? | Rewrite the problem in the customer’s terms |
| Does Who name a role, a market and a trigger? | Add the missing one |
| Is every number sourced or labeled “assumption”? | Add the source or the label |
| Does the target move toward “better”? | Flip the direction |
| Are share changes written as points or as a relative change? | Write “from 10% to 5%, 5 points” |
| Does it describe a condition without blaming a person or team? | Replace the name with the condition4 |
| Does it cover one problem? | Split it; PMI advises against tackling several at once4 |
Worked example: from a vague pitch to a finished statement
This is an example by our method, the one our Guides use to validate ideas. Every figure below is an assumption.
The pitch: “Groomers hate no-shows. We’ll build an app with deposits.” It names two solutions (an app, deposits), skips the customer and the cost, and sets no finish line.
Step 1. Suppose a solo groomer’s calendar shows 30 appointments a week and 3 no-shows (both assumptions). In our example, these assumed figures give a no-show rate of 10%.
Step 2. Regulars book their next visit at checkout, weeks ahead. Today the groomer texts reminders by hand the night before, and regulars push back on deposits (assumption, to be confirmed in interviews).
Step 3. The chain:
- Why did the client not come? They forgot.
- Why did they forget? The booking was made 3 or more weeks out (assumption).
- Why so far out? Recurring slots are booked at checkout.
- Why does that lead to forgetting? Nothing reminds the client closer to the date except a manual text that does not always go out.
The chain stops at four: “no reminder closer to the date” is a cause she can test and change.
Step 4. In our example, an average ticket of $70 and 3 no-shows a week (both assumptions) mean the no-shows cost $210 a week. In our example, assuming 50 working weeks, that comes to $10,500 a year ($210 × 50). In our example, the target (an assumption) is to cut no-shows from 10% to 5% of booked appointments, a drop of 5 percentage points, within 90 days.
Step 5. Cut “app” and “deposits”. Three fixes now fit: automated reminders, a deposit policy and a waitlist that refills empty slots.
Solo dog groomers in Austin who book recurring clients weeks ahead lose about 3 of 30 weekly appointments to no-shows (assumption). Today they text reminders by hand the night before; in our example, with assumed figures, the empty slots cost about $210 a week. In our example, it is solved when the no-show rate falls from 10% to 5% of booked appointments within 90 days; the target and deadline are assumptions.
| Part | Where it sits in the statement |
|---|---|
| Who | Solo dog groomers in Austin who book recurring clients weeks ahead |
| Problem | Lose appointments to no-shows |
| Context | About 3 of 30 a week, on slots booked weeks ahead |
| What they do today | Text reminders by hand the night before |
| Impact | About $210 a week |
| Solved when | 10% to 5% of booked appointments within 90 days |
Problem statement examples by type
Each example is an example by our method. Every figure is an assumption.
Internal operations: refunds at an online shop
A 12-person online home-goods shop takes 9 business days from the moment a returned parcel arrives to the moment the refund posts (assumption). Support answers about 40 “where is my refund” emails a week (assumption) by checking the warehouse sheet against the payment dashboard by hand.
Customers of our online shop wait 9 business days after their return arrives for a refund (assumption). Today support answers about 40 refund-status emails a week by hand (assumption). It is solved when the median time from parcel arrival to refund falls from 9 to 3 business days by June 30, 2027 (assumption).
Who is the returning customer, the trigger is parcel arrival, and the cost is support time. The statement names no tool.
Project: a paper inventory count
This one follows PMI’s pattern of moving from the current state to a future state by a target date4. A restaurant group with 3 locations counts inventory weekly on paper, and a manager retypes the counts into a spreadsheet (assumption). Each count takes about 5 staff hours per location, retyping included (assumption), or 15 staff hours a week across the group by arithmetic.
Current state: each of our 3 locations spends about 5 staff hours a week counting inventory on paper and retyping it (assumption). Future state: counts are entered once and take about 2 staff hours per location (assumption). Target date: March 31, 2027 (assumption).
The future state describes the result and leaves the method open.
Research: why food-truck owners leave delivery platforms
A research statement needs no figures. Suppose your literature review finds no study of why food-truck owners stop using a delivery platform soon after they join.
Little is documented about why food-truck owners stop using delivery platforms within their first month. Without that, neither platforms nor owners can tell which onboarding problems to fix. This study asks: what makes food-truck owners leave a delivery platform within the first month after joining?
The one-month window is a scope you choose and defend in the method section.
FAQ
How long should a problem statement be?
One to three sentences, with evidence lines underneath. For research work, MIT’s D.I.S. course has you summarize each part in 20–50 words and combine them into one statement of 100–150 words3. Market size goes in a separate document: see how to calculate market size.
Should a problem statement include the solution?
No. PMI advises against suggesting solutions prematurely4. Run the three-solution test: if only one fix fits your wording, rewrite it.
What are the 5 Ws of a problem statement?
Who, What, When, Where and Why. They leave out the current workaround and a measurable finish line, so add both.
What is the difference between a problem statement and a goal or hypothesis?
A goal is the finish line alone; a problem statement adds the customer and the cost. A hypothesis is a testable guess about a cause or a fix, and it sits under the statement as evidence to check.
How do you write a problem statement for a research paper?
State the gap in current knowledge, the consequence of leaving it open and the question your study answers. Purdue OWL frames it as the move that shows the reader why your document is important2.
Does a federal grant application need a problem statement?
Follow the sections of the notice of funding opportunity. Federal rules in 2 CFR Part 200, Appendix I set the full text a notice must contain, including a program description with the purpose of the funding and what it is expected to achieve6. Executive Order 14332, published in the Federal Register on August 12, 2025, set oversight of federal grantmaking7. The Department of Energy issued its implementation for funding announcements on February 19, 20268. Read the current notice for your agency and fit the six parts to its need or significance section.
You have the problem. Check whether people will pay to solve it: Check your idea for free.
Sources
- en.wikipedia.org, 10/23/2026
- owl.purdue.edu, 10/23/2026
- ocw.mit.edu, 10/23/2026
- pmi.org, 10/23/2026
- d-eship.com, 10/23/2026
- ecfr.gov, 10/23/2026
- govinfo.gov, 08/12/2025
- energy.gov, 02/19/2026
Check your idea for free
Describe your idea and the check collects demand, competitors, complaints and prices, with sources.
