SALT HARBOR REAL ESTATE · SNAPSHOT SEPTEMBER 5, 2026
A real blog.
Measured results.
One client’s published articles, measured through Google Search Console and verified AI requests. Here is what the numbers show—and what they don’t.
Results from one client; outcomes vary. Shared with permission. No ranking, traffic or AI citation is guaranteed.
GOOGLE SEARCH
24 of 75 articles
Had at least one reported search query with an average position in Google’s top 10.
May 7–September 3, 2026
How we count rankingsAI ARTICLE ACCESS
148 verified fetches
137 ChatGPT-User requests and 11 Claude-User requests to individual articles.
July 9–September 5, 2026
Fetches are not confirmed citations.
What a fetch establishes01 / SEARCH
A query average,
not a fixed position.
We count an article once if at least one page-query pair has an average position from 1 to 10 during the reporting window. All 75 published articles are in the denominator, including those without reported rankings. The result is 24 of 75, or 32%.
For the Amelia Island Plantation article, the query “amelia island plantation resort” averaged position 2.0. That is a historical query average, not a current placement or a promise about every search.
Reporting window and other search measures
The first Salt Harbor article was published June 28, 2026. September 3 is 67 elapsed days later. Search Console was queried for May 7–September 3; returned daily records begin June 28, contain no pre-publication impressions, and end September 2.
The daily records total 12,449 impressions. Impressions count appearances in search results, not unique people or website visits. No minimum impression threshold was applied to the ranking measures; some queries are infrequent.
Sixteen of the 75 articles had at least one page-query average from 1 to 3. Those articles are included in the 24 top-10 articles. The named-query results contained 348 phrases with an average position of 10 or better, including 236 at 3 or better. Article counts and query counts are different measures.
02 / AI ACCESS
Opened by an assistant.
Citation unconfirmed.
Each counted request reached an individual article and matched the provider’s published IP ranges. Index pages, utility paths and spoofed requests are excluded. These are fetch events, not 148 unique people, conversations or leads.
A fetch establishes that the assistant accessed an article. It does not establish that the final answer cited it. We do not present these requests as screenshots or proof of an AI recommendation.
Weekly counts and cutoff
Seven-day periods starting July 9, July 16, July 23, July 30, August 6, August 13, August 20 and August 27 recorded 28, 22, 14, 11, 18, 22, 15 and 13 verified fetches. The partial period starting September 3 recorded five through September 5 at 01:08 UTC. Together they total 148. This shows recurring activity, not a claim that activity increased each week.
Inspect the source.
Read the measurement snapshot or visit the client’s published blog.
Articles are published on the client’s domain, under their name, after approval. Published content remains theirs after cancellation. A growing library does not guarantee a ranking or third-party authority score.