We had five finished case studies sitting ready for a manufacturing client's outreach campaign. Real photos, real projects, written and formatted and one approval away from going out. Then the client's own manager read them and said something that stopped the whole thing: some of those jobs were not theirs.
He was right. Every client case study in that pack came from the wrong half of a shared file archive. We had to throw all five away and start again with three.
What actually went wrong
This client shares one big file archive with a second company. Two brands, one storage space, folders sitting side by side. The only thing separating them is a short prefix at the front of each folder name.
I read that prefix as a department name. It was a company name.
So the photography I pulled — restaurants, rooftops, hotels, a museum — was genuinely finished work, genuinely well shot, and genuinely someone else's. Not a stock photo, not a fake. A different company's completed jobs, filed one shelf over.
Worse: months earlier I had used the same misreading to lift a caution on two of those projects. An earlier version of me had flagged them as unproven. I overruled that flag on bad information. The original caution was correct all along.
Why a shared archive is a trap for your best proof
If you have ever worked with a partner, a contractor, a previous owner or a sister company, your files probably live closer together than your businesses do. And the material most likely to get mixed up is exactly the material you want for sales:
- Finished-job photography, because it is the only thing a prospect really looks at
- Reference lists, which get copied between teams and lose their origin
- Anything with a client's name as the folder title, since the name tells you who the customer was, not who did the work
The failure is quiet. Nothing errors. A confident, attractive case study assembled from the wrong folder looks exactly like a correct one.
How do you check that a case study is really yours?
Three questions, in this order. They take minutes and they would have saved us a week.
1. Who owns the folder, not the job?
Trace the picture back to the top-level folder and ask a human what that folder is. Do not guess a prefix. I guessed, and a guess is what cost us the pack.
2. Is the proof in your own job file?
Every one of this client's real jobs has a slot for completed photos. On every job we checked, that slot was empty. Active bids had drawings and site surveys; finished installs had nothing. That emptiness was the tell, and I read past it because a full folder was sitting nearby.
3. Did you do the whole job, or part of it?
The rule the client set afterwards is the cleanest version I have heard: use only projects you designed, produced and installed in full. If a different producer touched any part of it, it is not your case study.
Three honest case studies beat five borrowed ones
We asked for five. The archive yields three. So we sent three — ranked strongest first, with a plain note at the top naming every project we had dropped and why.
Then we listed five more genuine jobs that have no photograph at all, so the client can decide whether to go find one. And we noticed that two of the three good projects were photographed on the same day, which means one phone call to that photographer could roughly double what they are able to show.
That last part only became visible because we stopped defending the pack and started counting what we could actually prove.
A case study is a claim about who did the work. If you cannot say whose folder the photo came from, you do not have a case study yet — you have a nice picture.