BLOG MATRIX / CASE STUDIES
Our first published case study is CGL Typing, a client site we built and manage, with its Search Console data unedited. This page also sets out how to read anybody’s case studies, ours included.
PUBLISHED CASE STUDIES
More will appear here as clients agree to it, each with the scope, the timeframe and the source of every number.
CLIENT PROJECT · WEB DEVELOPMENT AND SEO
A free SSC CGL typing test we designed, built in hand-written code and manage for the client, SEO included. No ads, only SEO.
The engagements that stalled, the ones where the client never implemented the recommendations, and the ones that simply did not work are not written up. That is not dishonesty. It is what a portfolio is. But it does mean a page of successes tells you the ceiling of somebody’s work and never the average.
So the useful question is not whether a provider has a good example, because almost everyone does. It is what share of their engagements resemble it, and what happened in the ones that did not. Somebody who can describe a project that went badly and what they took from it is giving you more information than somebody with nine polished stories.
A leading D2C brand in the home category cannot be checked. Neither can a dashboard screenshot with the property name cropped out, or a chart whose axis begins at a number nobody will state. If the client is not named and reachable, you are being asked to take the figures on trust, which is the one thing evidence is supposed to remove.
There are legitimate reasons for anonymity. Confidentiality agreements are real and some clients simply decline. But the honest response to that constraint is to say so and offer a reference call, not to publish an unverifiable story and present it as proof. When you meet an anonymous case study, read it as a description of method rather than a demonstration of result, and judge the method.
A rising line is the most common image in SEO marketing and among the least informative. Sessions can rise because a site started ranking for things nobody buys, because one article caught a news cycle, or because the comparison year was unusually poor. None of that reaches the bank.
What is worth looking for is whether the exhibit connects to money, or to something that leads to money: enquiries, qualified enquiries, bookings, orders. And whether absolute numbers appear at all. A four hundred per cent increase is a very different claim if the starting point was nine visits a month rather than nine thousand. Where a case study shows percentages and withholds the base, that choice was made deliberately. Which measures are the right ones depends on the business, and the table further down sets out how that differs.
Attribution is the hardest part of an honest case study, because SEO never runs in isolation. Across the same twelve months a business might have raised funding, hired a salesperson, run paid campaigns, been written about, launched a product, or simply watched a competitor go quiet. Search itself changed several times.
None of that makes the work worthless. It makes single-cause claims unreliable. A case study worth reading names what else was in play and says which part of the movement it can reasonably claim. One that draws a clean line from we did SEO to revenue doubled is telling you about the confidence of its author rather than about the business.
Ourselves included. We have one published case study so far, CGL Typing, and one example is a sample, not a track record. What we would suggest instead is the thing that actually predicts whether an engagement goes well: ask any provider to look at your site and tell you what they think is wrong with it, and why, before you sign anything.
A specific diagnosis of your website is harder to fake than a case study and considerably more relevant, because it is about you. This page competes with that one. This section generates URLs nobody should see. This service has demand you are not addressing. It also shows you how somebody thinks, which is the thing you are buying. If you want to see how we approach it, the technical, content and consulting pages set out the method in more detail than a success story would.
WHAT A CASE STUDY SHOULD TELL YOU
Measured against everything above, this is what we would have to put in our own. Everything we publish here has to carry these three things, or it is not worth reading.
The business goal, website situation, market and baseline before the engagement.
The priorities, actions and implementation responsibilities, with a clear timeframe.
Relevant changes, the source of the data and the limitations behind any result.
MEASUREMENT THAT FITS THE BUSINESS
The right measures depend on how your business serves customers and records enquiries or sales.
| Business context | Measures to consider | Context to keep |
|---|---|---|
| Local service business | Relevant local visibility and qualified enquiries | Service areas, seasonality and lead quality |
| Online store | Organic product/category visits and attributed sales | Stock, promotions and attribution limits |
| B2B or professional services | Relevant demand and qualified business enquiries | Longer buying cycles and sales follow-up |
QUESTIONS, ANSWERED
Yes, as client work matures and clients agree to it, described without dressing it up. The first is CGL Typing. A page of anonymous, unverifiable results would not tell you anything useful anyway.
Something checkable. A named client you can contact, read-only access to a live report, or a small piece of paid work with a defined output. All three tell you more than a portfolio you cannot verify.
Rarely on its own. It is the easiest exhibit to select favourably, and something else was usually happening at the same time — a season, a campaign, a site change. Ask what else changed in that window.
A USEFUL NEXT STEP
We can discuss the outcomes that would matter to your business and how to track them.
