Case Studies: What the Work Actually Produced
Three engagements written up properly: the situation as it was found, what the diagnosis turned out to be, the order the work was done in, and what changed as a result. Including the parts that did not work.
The brands are not named. Most of this sits under confidentiality, and I would rather keep a client's numbers private than trade them for the next engagement. Everything that makes a case study useful, the reasoning and the sequence, is here in full.
Four Engagements, Four Different Problems
Plenty of traffic, no pipeline. The content was ranking for questions the product did not answer, and the conversion path assumed a reader who had already decided. The fix was intent reclassification and rebuilding the route from article to trial.
A Shopify catalogue where collection architecture, duplicate paths and template performance were capping everything. Category-first rebuild, then a content engine feeding the collections that could convert.
A brand ranking for clinical queries that shared vocabulary with its products but no purchase intent. The fix meant deliberately losing most of the impressions, and building landing pages for a query type with almost no competition.
A brand entirely dependent on paid social, with rising acquisition costs and attribution that had stopped being trustworthy. Technical cleanup, category architecture, then content, in that sequence.
What a Case Study Here Contains
Written to be useful to someone with a similar problem, rather than to be impressive.
The situation as found
What was actually happening when I arrived, including the things everyone believed that turned out not to be true. This is usually the most useful section and the one most case studies skip.
The real diagnosis
Almost never what the brief said. The named constraint that was capping everything else, and the evidence for it.
The sequence
What was done first, second and third, and why that order. Sequencing is the part that transfers between businesses even when the specifics do not.
What did not work
Every engagement has hypotheses that failed. Leaving them out makes the story cleaner and the document useless.
The measured outcome
Reported as measured. Where a number is a range, the measurement was a range.
What I would do differently
With hindsight, which parts were wasted effort. The most honest section and the one I would read first.
Why There Are No Logos on This Page
A wall of client logos is the standard proof device in consulting, and I have chosen not to use one. Partly that is contractual, since most of the substantial work I have done sits under agreements that do not permit it. Partly it is a preference.
A logo proves somebody paid an invoice. It says nothing about what was done, whether it worked, or whether the person showing you the logo did the work rather than sat in the pitch. The things that would actually help you judge me are the reasoning, the sequence and the honest account of what failed, and none of those require a brand name attached.
For a serious engagement I will arrange a reference conversation with a client whose situation resembles yours. That is a better signal than a logo grid and it costs the client something, which is why I ask rather than publish.
Questions About This Work
Why are the brands not named?
Because most of the work sits under confidentiality, and because the ones that do not are still commercially sensitive. Naming a client to win the next one is a trade I would rather not make. What is published here is the situation, the diagnosis, the sequence of work and the measured outcome, which is the part that is actually transferable.
Are the numbers real?
Yes, and they are reported as they were measured rather than rounded upward. Where a figure is a range, it is a range because the measurement was. Where something did not work, it is in the write-up, because a case study with no failed hypothesis in it is marketing rather than a record.
Can I speak to a reference?
For a serious engagement, yes. I will ask the relevant client first rather than hand out contacts, and I will match you with someone whose situation resembles yours rather than whoever had the best result.
Why so few case studies for twenty years of work?
Because most of what I have done was in-house or under agreements that do not permit a write-up. These are the engagements where I have both permission and enough measured detail to be useful. I would rather publish three that stand up than twenty that read like brochures.
What does a typical engagement look like?
Almost always a diagnosis first, then a defined sequence of work, then either your team executing with me directing or a longer consulting arrangement. The specifics are on the services page, and the honest version is that the smallest option that solves the problem is usually the right one.
Do you only work with e-commerce and SaaS?
No. Those happen to be the studies I can publish. The sectors I work in are broader and include healthcare, insurance, automotive, marketplaces, careers platforms and local service businesses. See industries.
More context: how I work, the full list of services, the sectors I work in, and the writing, which is the most honest sample of how I think about these problems.
Have a problem that resembles one of these?
The first conversation is thirty minutes and costs nothing. Bring your numbers and I will tell you what I think is actually happening, including if the answer is that you do not need to hire anyone.
Book a Free Strategy CallGet an SEO Audit
