[ Case study ]
The site's medically reviewed articles answered questions better than the aggregators that outranked them, yet answer engines paraphrased the aggregators. Reviewer credentials — the site's strongest trust asset — were invisible to machines.
CLIENT a patient-education health site — FOCUS Surface the review chain as data
Representative examplesEvery case study in this library is an illustrative composite of the kind of engagement we deliver — written to show our method and standards, not to name clients.
Trust is the entire product of this patient-education health site: every article is written by a named author, reviewed by a credentialed medical reviewer, and dated. Its answers are better than the aggregators that outrank it — yet answer engines paraphrased the aggregators, because the review chain, the site's strongest trust asset, existed only as visible byline text machines could not weigh. The editorial calendar runs on a fixed review cycle; the review board guards wording strictly; and nothing may ever imply a claim was made without review.
The site's medically reviewed articles answered questions better than the aggregators that outranked them, yet answer engines paraphrased the aggregators. Reviewer credentials — the site's strongest trust asset — were invisible to machines.
We proposed surfacing the review chain as data: author, medical reviewer, credentials, review date, and sources became structured facts on every article, always visible to humans as well. Question-first sections with concise, extractable answers would be added under the existing review process, so machines quote accurately while humans still read the full context — and no claim ever bypasses the board. Measurement would use the niche's real patient questions: a weekly panel of actual queries run against the answer surfaces, logging whether and how the site is cited.
Just as important is what we ruled out, and why:
Author, medical reviewer, review date, and sources became structured facts on every article — the trust signals machines can weigh.
Question-first sections with concise, extractable answers were added under the existing review process, so machines quote accurately and humans read the full context.
A weekly panel of the site's actual patient questions ran against the answer surfaces, logging attribution.
Delivered by the growth pod — content strategist + technical SEO specialist over 10 weeks, with working increments reviewed with the client every week.
Obstacle
Three articles went through medical review twice because schema batches landed mid-cycle and changed pages already approved.
Handled: Schema work was batched per review cycle from then on; the three double-reviewed articles became the calendar template the editor still uses.
Obstacle
Extractable short answers tended to strip the clinical hedges — 'speak to a clinician if' disappeared under compression — and the board flagged the first batch.
Handled: Answer blocks now carry the hedge and a source line by rule; the board approved the extraction pattern as standing policy rather than per-article negotiation.
Obstacle
Two reviewers shared near-identical names with unrelated practitioners online, and early attribution attached the wrong person's credentials.
Handled: Dedicated reviewer profile pages with credentials and disambiguating sameAs links were shipped before the next schema batch went live.
The headline: weekly panel questions where the site is the cited source, weeks 1 to 10 — 2 → 31, read from Weekly answer-panel tracking. A second check: organic entrances to the restructured articles at +17%.
The site's name now appears as the cited source for questions it has answered correctly for years, and reviewers see their credentials travel with their words instead of sitting under a byline nobody weighs. The editorial team treats the review chain as an asset rather than overhead, and the board approves new article formats quickly because the extraction pattern has never once moved a claim.
The result was read from Weekly answer-panel tracking against the pre-engagement baseline over the stated window, with a guardrail check on organic entrances to the restructured articles. Where platform-reported numbers and business outcomes differ, this record says which layer it is quoting.
What we would do differently
We would have batched the schema work per review cycle from the start — three articles went through review twice because of our sequencing.
[ Related service ]
[ Related builds ]
14 31Qualified inquiries per month attributed to service pages, over the 8 weeks after rollout versus the 8 before
3 43Referring domains from non-sponsored coverage, over 12 weeks
[ Next step ]
Next case study
