Why our case studies page is called Lab
We are a young studio with no client logos to show. Instead of inventing some, we built four working systems and published them. Here is the reasoning, and a checklist for auditing anyone's case studies.
Every new agency faces the same temptation on day one. You need a case studies page, because buyers expect one. You do not have case studies, because you are new. The industry's quiet solution is to invent them: blur a logo, round up a metric, describe a project that was really a tutorial. Nobody checks. Everybody knows.
We came close. Early drafts of this site had a case studies page with impressive percentages on it. Then we asked ourselves the question we would ask any vendor: who can verify this? The answer was nobody, because the clients did not exist. So we deleted the page and did something slower instead.
What we did instead
We built the systems for real. A multi-property hotel booking platform with channel sync. An offline-first restaurant POS. A patient portal designed around audit logging. An AI layer over a CRM with human review gates. No client asked for them, and we say so plainly: they are internal builds, labeled MVP, Prototype, or Research. The page is called Lab because that is what it is.
The builds took months we could have spent cold-emailing. But they gave us something a logo wall cannot: verifiable evidence of how we think. Anyone can read what we built, why, and what it taught us. A prospect can push on any detail and we can answer, because we actually did the work.
How to audit anyone's case studies
If you are hiring a software partner, the case studies page is where the fiction concentrates. Some questions that separate real work from marketing:
- Can you speak to the client? A real engagement has a reference. Confidentiality is sometimes real, but if every single client is confidential, that is an answer too.
- Who exactly did the work? Agencies inherit case studies from people who left years ago. Ask whether the team on your project built the thing on the page.
- What was the failure? Real projects have a part that went wrong. A vendor who cannot name one is describing a project that did not happen or a lesson they did not learn.
- Do the numbers have denominators? A 300% improvement of an unstated baseline is decoration, not evidence.
- Can they go deep on demand? Ask one specific technical question about the case study. Watch what happens.
What honesty costs
We will lose some deals to agencies with better-looking walls of logos. Buyers who select on logo walls are, in a sense, not our buyers. The ones we want are the ones who read a lab page and recognize the difference between demonstrated ability and claimed history. So far, those are also the projects worth doing.
The standard we are holding ourselves to
When we have real client work we can publish, it will appear with names, numbers with denominators, and the parts that went wrong. Until then, the lab is the evidence. We would rather show you working software than tell you a story.