Fundamentals
How to Write a Problem Statement: Template and Before-and-After Examples
Part of our guide to business analyst roles and responsibilities.
Symptoms, problems and solutions: know the difference
When people first describe a problem at work, they rarely describe the problem itself. They describe what they feel (a symptom) or what they want to buy (a solution). Your job as a business analyst is to find the real problem between the two. If you are new to the role, What does a business analyst do? explains where this fits in the wider job.
- Symptom: something people notice or feel. 'Customers are angry.' 'The report is always late.' It is real, but it does not tell you why.
- Problem: the condition causing the symptom, stated so you can measure it. 'Refund requests wait several days for manual approval.'
- Solution: a way to fix the problem. 'Build an automatic refund feature.' Useful later, but it does not belong in the problem statement.
A quick test: if the sentence contains a tool, a system, a feature or a verb like build, buy or automate, it is a solution. If it only describes a feeling, it is a symptom. A problem sits in the middle. It is a specific, observable gap between how things are and how they should be.
Symptom, problem and solution side by side
| Aspect | Symptom | Problem | Solution |
|---|---|---|---|
| What it is | What people notice or feel | The condition causing the pain, with a measure | A way to remove the problem |
| Typical wording | Customers are unhappy with refunds | Refunds wait several days for manual finance approval | Add automatic refunds for small orders |
| Question it answers | What hurts? | What is going wrong, for whom, and how much? | What should we do about it? |
| Who usually says it | Customers and frontline staff | The business analyst, after analysis | Sponsors, developers and vendors |
| Where it belongs | Elicitation notes | Problem statement and business case | Options analysis and requirements |
A simple problem statement template
You do not need a fancy format. A good problem statement answers five questions in one short paragraph, usually three to five sentences.
- Who is affected? Name the role, team or customer group.
- What is going wrong? Describe the gap in plain words.
- Where and when does it happen? Give the process step, channel or time.
- How much? Add a measure from real records, such as a count, a time or a rate.
- So what? Say the impact on customers, money, risk or staff, and what success would look like.
TEXT
[Who] experience(s) [what goes wrong]
when or during [where or when it happens].
In [period], [measure from real records].
This causes [impact on customers, cost, risk or staff].
We will know it is solved when [measurable target].Another common format reads: 'The problem of [X] affects [Y]. The impact of which is [Z]. A successful solution would [W].' Both work. Pick one and use it the same way across your documents, so readers always know where to look for each part.
Some teams leave out the success line, because the target often moves into the business case or into SMART objectives. Keep it in your first drafts anyway. It forces you to think about how the result will be measured.
How to write a problem statement in six steps
- Collect the complaints word for word. Write down what stakeholders say, in their own language. Do not fix or judge yet. Interviews, observation and support tickets are good sources. See requirements elicitation techniques for more options.
- Ask why until you reach a cause you can observe. Use the five whys gently. Stop when the answer is something you can see happening in a process, not an opinion.
- Name who is affected. 'Users' is too broad. Say 'new customers on the mobile app' or 'night-shift warehouse supervisors'.
- Measure the gap. Find one or two numbers from real records, for example how many, how long or how often. Say where each number came from. If no records exist, say so and plan how to get them. Never guess a number to make the statement look stronger.
- State the impact. Connect the problem to something the sponsor cares about: lost revenue, cost, customer complaints, compliance risk or staff time.
- Remove hidden solutions, then test it. Delete any tool, system or feature. Read the statement to two or three stakeholders and ask: 'Is this the problem? Is anything missing or wrong?'
Diagram in words
- 1. Collect complaints: Write down what stakeholders say, in their own words, before you judge anything.
- 2. Find the cause: Ask why until the answer is something you can observe in a process.
- 3. Name who is affected: Pick a specific role, team or customer group instead of everyone.
- 4. Measure the gap: Add a count, time or rate from real records, and note where it came from.
- 5. State the impact: Link the gap to customers, cost, risk or staff time.
- 6. Remove solutions and test: Delete tools and features, then check the wording with stakeholders.
Expect two or three drafts. That is normal. The first draft is often a symptom with a number attached. A later draft, after stakeholders have pushed back, is usually the one people agree on.
Three before-and-after problem statement examples
Here are three vague complaints turned into clear problem statements. All three companies are fictional, and the numbers are part of the story, not facts about any real business.
Example 1: Kestrel Credit Union, a fictional lender
Before: 'Our loan system is too slow. We need a new one.' This mixes a symptom (slow) with a solution (a new system). In a short first meeting in April, the business analyst learned that the system itself was not the main issue. The time went into re-typing paper forms, and the operations team already had two sources of real data from March: a timed sample and a callback log.
Finding the problem behind 'we need a new system'
Kestrel Credit Union, a fictional lender, in April: the first meeting with the project sponsor about slow loan decisions.
- Helena Project sponsor
Our loan system is too slow. I want a new one by the end of the year.
- Nadia Business analyst
Before we look at systems, can you walk me through where the time goes, from receiving the form to the decision?
- Helena Project sponsor
Loan officers type everything from the paper form into the system. Sunil's operations team timed 20 applications in March, and the re-typing averaged 40 minutes each.
- You The BA learner
How many applications came in during March, and how many had problems?
- Helena Project sponsor
In March, 200 across our 4 branches. The callback log shows 30 of them needed a call to the customer because of typing errors.
- Nadia Business analyst
So the slow part may be the re-typing, not the system itself. Let's write the problem around that, say where each number came from, and keep a new system as one possible option.
The people and the company in this scene are fictional.
After: 'Loan officers in Kestrel's 4 branches re-type every paper loan application into the loan system. In a timed sample of 20 applications in March, re-typing averaged 40 minutes per application. The March callback log shows that 30 of 200 applications (15%) had a typing error that needed a customer callback. Callbacks delay loan decisions and frustrate applicants. We will know this is solved when no more than 10 of every 200 applications (5%) need a callback, measured monthly.'
Notice that the statement says the 40 minutes comes from a sample of 20, not from all 200 applications. That honesty matters. If someone later times more applications and gets a different average, the team can update the figure without losing trust in the whole statement.
Anatomy of the Kestrel problem statement
Kestrel Credit Union problem statement
Loan officers in Kestrel's 4 branches re-type every paper loan application into the loan system.1
In a timed sample of 20 applications in March, re-typing averaged 40 minutes per application.2
The March callback log shows that 30 of 200 applications (15%) had a typing error3
that needed a customer callback.
Callbacks delay loan decisions and frustrate applicants.4
Success: no more than 10 of every 200 applications (5%) need a callback, measured monthly.5
- Who, where and the cause. A specific role in a specific place, doing something you can watch happen. No new system is mentioned.
- Measure with its source. The time comes from a timed sample of 20 applications, and the statement says so. It is not an opinion or a guess.
- Measure from real records. A count, a period and a source: the callback log. The percentage comes from the count: 30 of 200 is 15%.
- Impact. Says why the sponsor should care: slower decisions and unhappy applicants.
- Success target. Gives the team a finish line it can check every month using the same callback log. 10 of 200 is 5%.
Example 2: Marigold Meals, a fictional food delivery company
Before: 'Customers hate our app.' This is a pure symptom. When the analyst read a sample of support tickets, most of the anger was about refunds, not the app design.
After: 'Customers who report a missing or wrong item wait an average of 5 days for a refund, because finance approves every refund by hand. In April, support received 500 refund tickets, and 120 of those customers (24%) sent a second complaint while waiting. This adds support work and puts repeat orders at risk. Success: at least 90% of refund tickets are completed within 1 day.'
Example 3: Larkspur Logistics, a fictional warehouse company
Before: 'We need a dashboard for the warehouse team.' This is a hidden solution dressed up as a request. A dashboard might help, but so might a simple alert or a change to the picking order. The problem has to come first.
After: 'Shift supervisors at Larkspur's main warehouse only find out which orders will miss the 6 pm same-day dispatch cut-off during the last hour of the shift, when it is too late to re-plan. In the last 4 weeks, 80 of 2,000 same-day orders (4%) missed dispatch. Each miss means a late delivery and a customer contact. Success: no more than 20 of every 2,000 same-day orders (1%) miss dispatch.'
Notice what changed in all three: a named group, a measured gap with a known source, a clear impact and a target, with no tool or feature in sight.
Common mistakes, including hidden solutions
Most weak problem statements fail in the same few ways. Check your draft against each one.
- Hiding a solution. 'The problem is that we do not have a CRM.' The lack of a tool is almost never the real problem. Ask what goes wrong because the tool is missing, and write that instead.
- Stopping at the symptom. 'Staff morale is low' may be true, but it gives the team nothing to analyse. Keep asking why.
- No measure. Without a number, nobody can tell if the change worked. A rough count from real records beats a precise guess.
- Presenting opinion as data. 'Managers think it takes ages' is useful as a lead, not as evidence. Time a small sample or check a log, and say which one you used.
- Too broad. 'Our customer experience is poor' covers everything and therefore nothing. Narrow it to one group and one process.
- Blaming people. 'The finance team is slow' puts people on the defensive. Describe the process instead: 'every refund needs manual approval'.
- Writing an essay. If it runs past a short paragraph, background has crept in. Move history and context to the business case.
- Inventing numbers. If you do not have the figures yet, write 'to be measured' and plan how to get them.
Strong versus weak problem statements
Strong habit
- Name one specific group of people, such as night-shift supervisors or new app customers.
- Describe the process gap you can observe, not the feeling it causes.
- Add a count, time or rate from real records, with the period and the source.
- Say what success looks like so the team knows when to stop.
Weak habit
- Write 'we need a new system' or 'we lack a dashboard'. That is a solution in disguise.
- Blame a team. Describe the step in the process instead.
- Guess a number, or pass off an opinion as a measurement.
- Try to cover the whole company in one statement.
Problem statement checklist
Run through this list before you share your draft. If any answer is no, revise it.
- It names a specific group of people who are affected.
- It describes what goes wrong, not how anyone feels about it.
- It contains no tool, system, feature or vendor name.
- It includes at least one measure from real records, with a period and a source.
- It states the impact on customers, cost, risk or staff.
- It says what success would look like.
- It fits in one short paragraph.
- At least two stakeholders have read it and agree with it.
Once agreed, the problem statement usually becomes the first section of your business case. This business case template and worked example shows what comes next. A before-and-after problem statement is also a strong first piece for a beginner portfolio, because it shows how you think, not just what tools you know.
Frequently asked questions
How long should a problem statement be?
Usually one short paragraph of three to five sentences. If it is longer, background or solution ideas have probably crept in. Move the extra detail to the business case or your analysis notes, and keep the statement short enough to read aloud in a meeting.
What if I do not have any data for the measure?
Say so honestly. Write 'to be measured' and add an action to collect the figures, for example by sampling support tickets for two weeks or timing one process step. A statement with an honest gap is better than one with a guessed number that someone later disproves.
Can I use a stakeholder's estimate as the measure?
Use it as a starting point, not as the final figure. Label it clearly as an estimate, then check it against a log or a small timed sample. Once you have real data, replace the estimate and note the source in the statement.
Is a problem statement the same as a business need?
They are close but not the same. A business need describes what the organisation must achieve, while a problem statement focuses on what is going wrong today and how much it hurts. Many teams write the problem statement first and then derive the business need and objectives from it.
Who should approve the problem statement?
Usually the project sponsor, after the people who live with the problem have checked that it is accurate. Frontline staff confirm the facts, and the sponsor confirms it matters enough to act on. Record the agreement so the team can refer back to it later.
Can a problem statement change during a project?
Yes. New evidence can show that the real cause is different from what you first thought. When that happens, update the statement, note why it changed and tell stakeholders, because a changed problem can also change scope and priorities.
This article was written by an AI system and published automatically after automated accuracy, safety and quality checks. All companies, people and numbers in the examples are fictional. BA Mentorship does not issue professional certifications and cannot guarantee any job or interview outcome. Spot an error? Email our support team and we will check and correct it.