Editorial & AI Policy
How the claims on this site are sourced, why our case studies are anonymous, and where AI is and is not used in our work.
Why every case study is anonymous
Most of our clients compete in categories where naming their agency tells a competitor exactly which levers are being pulled. Several have it written into the contract. Rather than publish a thin roster of the few who said yes, we anonymise all of it and publish everything else in more detail than a named case study usually contains: the vertical, the revenue band, the platform, the actual constraint, the work, the timeframe, and the numbers.
If you want references, ask on the call. We will connect you with a current client in a comparable category who has agreed to take those calls.
Where the numbers come from
Every metric published on this site is read from the client’s own Shopify Analytics, GA4 or Google Search Console — not from an ad platform’s self-reported attribution, and not from a dashboard we control. Comparisons are like-for-like periods against a stated baseline, and the timeframe appears on the study.
We do not round up, we do not cherry-pick the best week, and we have deliberately left in engagements where the headline number was modest. An agency page where every result is spectacular is a page that has been edited.
Reviews
Reviews are published with the client’s permission and with their name and company withheld at their request. Role, category, approximate size and city are accurate. We do not write them, we do not edit them for tone, and we have not removed the four-star ones.
We publish no AggregateRating or Review structured data for these. Google treats reviews hosted by the reviewed party as ineligible for star treatment, and marking them up regardless is a structured-data manual-action risk. Ratings that carry weight belong on third-party profiles, and that is where we ask clients to leave them.
Where we use AI
In our work: for research synthesis, first-draft product copy, alt-text generation at catalogue scale, code review, and analysing large volumes of session recordings or support tickets. Every output is reviewed by the person whose name is on the deliverable. We do not ship AI-generated product descriptions, blog posts, or metadata to a client’s store without a human editing pass.
On this site: AI assisted the drafting and research behind these pages. The service definitions, pricing, process and claims were set by us and are accurate. Location pages are written per market rather than templated, and the market descriptions were checked for accuracy.
What we will not do: mass-generate thin location or programmatic pages to farm search traffic, fabricate a review or a case study, or invent a statistic and dress it up with a citation. If a number appears on this site without a source, it is our own aggregate and is labelled as such.
Corrections
If you find something on this site that is wrong, tell us at [email protected] and we will correct it and say what changed.