The Case Study You Cannot Publish

Two different problems

The common one is a customer who would rather not be identified, which is solved by removing the identifying details and keeping the substance.

The harder one is a contractual restriction: an agreement saying you may not disclose the engagement, name the client, or describe the work.

That second constraint is not solved by anonymising, because the restriction covers the work itself rather than the name attached to it.

Businesses in that position frequently conclude they can publish nothing, which is usually wrong and costs them their strongest evidence.

Read what you actually agreed

Since the assumed restriction is often wider than the written one.

Many agreements prohibit naming the client and disclosing specific information, and say nothing about describing the type of work.

Some prohibit disclosing that a relationship exists at all, which is genuinely restrictive and is less common than people assume.

Some are time-limited, so work from four years ago may no longer be covered.

Reading the clause takes ten minutes and frequently reveals more room than the general sense of caution suggested.

Where it is genuinely unclear, that is a question for somebody qualified rather than a reason to assume the strictest reading.

What you can usually still publish

• The method, without the engagement.
• Aggregate figures across many jobs.
• The type of client and the type of problem.
• What you learned, stated generally.
• A composite, clearly labelled as one.

The second is the most underused. Saying that across roughly forty engagements the typical result was a certain range discloses nothing about any one of them and is more persuasive than a single case anyway.

Write about the method instead

The substitution that works and is frequently better than the case study would have been.

A page explaining how you approach a kind of problem, what you check first, what usually turns out to be the cause, and what you do about it, discloses no engagement.

It also demonstrates competence more directly than a case study, which asks the reader to infer skill from an outcome.

Draw the examples from across your work rather than from one job, so no single client is identifiable from the detail.

Where a specific detail is distinctive enough to identify somebody, change it or leave it out, which is the same discipline as anonymising with a higher threshold.

Aggregates disclose nothing

And are worth building deliberately.

Across a number of jobs, publish the range of outcomes, the typical duration, the proportion where a particular problem was found, and the common causes.

None of that identifies a client, and it is checkable in a way a single anecdote is not.

It requires you to have looked at your own records, which is the actual work, and the resulting figures are useful internally as well.

Twenty jobs is enough for a range to mean something for a small business.

Say how many the figures come from, since a range from four jobs and a range from forty are different claims.

A worked example

A business working almost entirely under confidentiality agreements had no published work at all.

They reread three of the agreements and found that two prohibited naming the client and disclosing specifics, and said nothing about describing categories of work.

They published a page describing their approach to a common problem, with figures drawn from across roughly thirty engagements and no reference to any of them.

They also asked two long-standing clients directly whether a short, specific write-up would be acceptable, and one agreed with a review of the wording.

That single named example, alongside the method page and the aggregate figures, was enough.

Their assessment was that they had assumed a total restriction for years without reading the clause.

Ask properly rather than generally

Because a specific request is granted far more often than a vague one.

Do not ask whether you can use them as a case study, which sounds open-ended and invites a no.

Write the piece first, send it, and ask whether they are comfortable with it as written, offering to remove anything.

Ask at the point they are most pleased, which is shortly after the work concluded successfully.

Offer the option of approving it without being named, which many will take and which still gives you a real engagement to describe.

Get the answer in writing and keep it, since a verbal agreement two years old is difficult to rely on when somebody new is in the role.

The composite, labelled honestly

A legitimate option with one condition attached.

A description drawn from several similar engagements, presented as a typical case, discloses nothing and is often more useful than any single one.

The condition is that you say it is a composite, plainly, in a line.

An invented case presented as a real one is a fabrication regardless of how ordinary the practice is, and a reader who works it out has learned something about everything else you publish.

A composite honestly labelled costs nothing in credibility and frequently reads as more considered, since it is describing the pattern rather than an outlier.

The counter-case

Some restrictions really are total.

Certain sectors, security work, and some public sector engagements genuinely prohibit acknowledging that a relationship exists, and the correct response is to publish nothing about it.

In those cases the site has to be built on other evidence: method, credentials, named people, and the aggregate figures if even those are permitted.

There is also a risk of getting this wrong in a way that costs a client relationship, which is far more expensive than the page was worth.

Read the clause, ask about the specific piece you have written, and where you are unsure, do not publish.

What to do

1. Read the actual clause, not your memory of it.
2. Check whether it expires.
3. Write about the method instead.
4. Publish aggregate figures across many jobs.
5. Say how many the figures come from.
6. Write the piece, then ask about that piece.
7. Label composites as composites.

Step one is the step most often skipped, since businesses operate for years on a general sense of what they may not say rather than on what they actually agreed.

The simpler version of this problem is covered in writing up a job without naming the customer.

Frequently asked questions

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.