The case study: STAR, and before and after

Why:You’ll learn to present a working case as proof: situation, task, action, result and an honest “before and after”. Afterwards you have one case study of your service from which the client can’t be identified, and you know what a case study must never reveal.

30 minutes

Read by speech synthesis (Qwen3-TTS); the voice is not a person’s.

A case study is a working case told as proof, in the STAR form: situation, task, action, result. A “before and after” compares one and the same quantity.

Proof in the form of a story

We have said it already: a story persuades more strongly than a list because it ties events together. A case study is such a story about a working case, one that serves as proof: “here is what we did for someone like you”. Good case studies often persuade better than any assurances, because they show the process and the result.

But a case study is also a responsibility. You are publishing the story of a real person, and readers believe you exactly as far as you are precise in it. So two things are needed: an honest form and the safety of the people you write about.

The STAR form

A simple framework for a case study is called STAR, from the first letters of four words: Situation, Task, Action, Result.

  1. Situation. What the person came with. Who they are (in general terms) and what position they were in.
  2. Task. What had to be done and what stood in the way: a deadline, a law, a constraint.
  3. Action. What you did. Three or four steps, each with a clear reason.
  4. Result. How it ended. With a number, if there is one.

The framework wasn’t invented for sites: it is used for telling stories in job interviews, because it forces an answer to “what was there, what did I do, how did it end”. For a case study on a site it is worth adding a term and a “before and after”.

Check yourself: what are the four parts of STAR?

Situation, task, action, result.

“Before and after”

The shortest proof is a pair: “Before:” and “After:”. But the pair has a condition: you must compare one and the same quantity. “Before: half a year of uncertainty. After: 14 days” is comparable. “Before: hard. After: good” says nothing.

Three rules for an honest “before and after”.

  • Name only what you measured. A time, a number of visits, a sum.
  • If there is no number, say so: “the result wasn’t measured with a number; the matter was closed the first time”.
  • Don’t promise the same result to another person. Write “in this case”, not “this is what you’ll get”. In the US, the FTC’s Endorsement Guides (16 CFR 255.2(b)) expect that when an ad shows a result that isn’t what most people can expect, it either has proof the result is typical or clearly discloses the result people can generally expect; a “results not typical” label alone no longer does the job. In the UK, the CAP Code asks that claims can be backed by evidence. For many professions a guarantee of a result is not allowed in advertising at all, and the ethics lesson comes back to this.

Play with the diagram below: the slider moves the figures from “before” to “after”. In the lower part, paste in the text of a case study and see what the program thinks is risky.

Before and after, and what a case study must not carry
Before and after
Time to settle the matter: 180 days
Trips to offices: 6

The numbers are invented. In a real case study name only what you measured, and if you didn’t measure, say so.

What must not stay
  • phone: +1 000 000 0000 → remove
  • phone: 000-000-000 → remove
  • email: john@example.com → remove
  • parcel number: parcel APN 000-000-000 → replace with “a lot in the suburbs”
  • case number: case No. 00-CV-0000. → remove, keep “in court”
  • address: 0 Shore Road → replace with “in town N”

If a set of details (a profession, a rare place, a date, an unusual situation) lets someone recognize the person, the case study still has to be generalized, or consent obtained.

Check yourself: what can you write if the result has no number?

Say honestly “no number” and describe the outcome in words. An invented number is worse.

What a case study must not carry

A case study is the story of a real person, so the first rule is do no harm to them. Under the EU and UK GDPR, personal data is any information relating to an identified or identifiable person, directly or indirectly (Article 4(1)), and data counts as anonymous only if the person can no longer be identified by any means reasonably likely to be used (Recital 26). In the US there is no single law, but state privacy laws such as the California Consumer Privacy Act use a similar idea, and health information has its own rule: HIPAA lists the identifiers that must go before health data is treated as de-identified (45 CFR 164.514). A practical rule follows: publish only a case study from which the person can’t be recognized, or one you have the person’s written consent to publish.

What to remove.

  • Direct identifiers: first and last name, phone, email, address.
  • Numbers in public records: a court case number, a parcel number, a document number. With them a person can be found in a minute.
  • Indirect clues: a rare profession, a small village, an exact date, an unusual situation. A set of such details can identify a person as easily as a name does.
  • Photographs and documents, if anything on them is visible.

Replace the details with a generalization: “a landowner in the suburbs”, “in spring”, “in a trial court”.

Some professions have stricter rules. In the US, a lawyer’s duty of confidentiality is broader than the courtroom privilege: under ABA Model Rule 1.6 and its state versions, a lawyer may not reveal information relating to a client’s representation without informed consent, even if parts of it are public record. Doctors and other health professionals face HIPAA. If you are a lawyer, a notary, a doctor or an expert witness, show the text to a specialist in your own profession before publishing a case study. This lesson gives landmarks, not legal advice; for your own case, ask your state bar or professional board, or, for data protection, the ICO in the UK or your state attorney general in the US.

Check yourself: why isn’t it enough to remove the client’s name?

A person can also be recognized from a case number, an address and a combination of rare details. Everything that identifies them, directly or indirectly, has to go.

Practice

Fill in one entry of a “Cases” collection, as is done in a headless CMS: an entry has fields, and the page shows them.

  1. Title and service. Take the service from your brief.
  2. Situation and task. In your own words, without details that could identify the person.
  3. Actions. At least three steps.
  4. Result and term. With a number, or an honest “no number”.
  5. Before and after. The same quantity.
  6. A check. Reread it as a stranger: could the person be found from the text?

Practice

Preparing your page…

Test

Question 1 of 6

Put the parts of a case study in STAR order

Put them in order with the “up” and “down” buttons.

Cards

These cards go into your review once you finish the lesson.

  • What is STAR?

    Answer

    Situation, Task, Action, Result. Four questions to ask of a working case.

  • How do you make a “before and after” honest?

    Answer

    Compare one and the same quantity (a time, a number of visits), name only what was measured, and if there is no number, say so.

  • What do you remove from a case study?

    Answer

    The name, phone, email, address, case numbers and parcel numbers, and also rare combinations of details by which the person could be recognized.

  • Can you write that the result is always like this?

    Answer

    No. One case is not a guarantee. Write “in this case”, not “this is what you’ll get”.

Sources

Next

You now know: the case study.