How to Write a Case Study for a Law Firm (Without Breaching Client Confidentiality)
Most law firms sit on a goldmine of proof they never use. Every settled matter is evidence of what a prospective client wants to know before they call: can this firm actually do the thing? Yet the case studies page, if it exists, is usually three vague paragraphs about "achieving a favourable outcome" that could describe any matter, ever. We get it 💖 — confidentiality isn't optional, and legal advertising rules are strict. But "we can't share details" and "we can't write a case study" aren't the same thing. Here's how to do the second one properly.
What most law firms get wrong
Two failure modes, seen across every practice area. First, firms go so vague protecting confidentiality that the case study proves nothing — "a client came to us with a difficult matter and we achieved a great result" could be about literally anyone. Safe, and useless, because prospective clients can't see themselves in it. Second, less often but more dangerously, firms include just enough real detail — a suburb, a settlement figure, a distinctive fact pattern — that anyone who knew the matter could work out who it was. Neither serves the firm. The goal is a story specific enough to feel real, generic enough that the client themselves wouldn't recognise it.
The copy-paste case study template
Steal this structure — it works across family law, commercial, criminal defence and most practice areas we've applied it to.
1. Situation (2–3 sentences)
The type of matter and the client's core problem, using a role rather than a name ("a small business owner"). State roughly what was at stake, no exact figures.
2. Approach (3–4 sentences)
The strategy, legal mechanism and judgement calls. This is where expertise actually shows — the hardest part to fake and the part clients are really evaluating.
3. Outcome (2–3 sentences)
The result in ranges or general terms ("settled well ahead of the court date"), never exact figures or dates, and never a guarantee of the same result for someone else.
4. De-identification checklist:
- Exact dates ("in 2023," not a specific week)
- Exact dollar figures (use ranges or relative terms)
- Suburb, workplace, school or other location detail
- Ages, family size, distinctive structure
- Anything the client called "unusual" about their matter
- Detail combinations that are identifying together, even if fine alone
- Written consent on file, even when you think it's de-identified enough not to need it
Three worked examples
How the de-identification actually works
Separate the facts that make a matter legally interesting from the facts that make it identifiable, and publish only the first. A prospective client doesn't need the $1.4 million settlement in Mermaid Beach — they need to know you've handled settlements involving hidden business income, and can explain how. Change identifying combinations, not just single details; three innocuous facts together can identify someone as surely as one obvious one. Then ask: would the client recognise themselves? If yes, it's not de-identified enough, consent or no consent.
Mistakes to avoid
- Publishing without written consent because it "feels" de-identified enough
- Using real or rounded settlement figures in a small practice community
- Writing the outcome as a guarantee or implied promise for future clients
- Going so vague the case study could describe any matter — it needs to feel specific to be believable
- Framing a client's worst day as a marketing win without acknowledging the human cost
- Reusing the same template details so every case study sounds identical
Frequently asked questions
Do I always need written client consent, even for a de-identified case study?
Yes, as good practice and often a professional obligation under your state law society's rules. Even a well de-identified matter can occasionally be recognised by someone with inside knowledge, so consent is the client agreeing their story can be told, not a formality.
Can I use exact settlement figures if I round them?
Generally, no. Rounding to the nearest hundred thousand can still tell a small circle of people — former colleagues, opposing counsel, the client's own family — roughly which matter it was, especially with a suburb or timeframe attached. Use ranges instead.
What if a case study isn't punchy without a few identifying details?
Sometimes that's true, and the right call is to let that matter go rather than compromise confidentiality for a better sentence. The strategy and judgement calls are usually the interesting part anyway, not the numbers.
Is it ever okay to combine two real matters into one case study?
Yes — a composite built from two or three similar matters can protect confidentiality better than a single de-identified one, provided you don't present it as one client's experience. Be transparent internally, and check your local rules on how composites should be described.
Keep reading 🤍
I help Gold Coast and Brisbane businesses grow with branding, websites and marketing that actually works.
Work with me ✦