A webmaster running rank-tracking for third-party publications found that three months of data logged as Yandex positions were in fact pulled from Google, after a paid API provider's two endpoints turned out to differ by a single character in the URL.
The script ran from July 7 to October 13, 2026, and collected 42,477 measurement rows, all labelled 'Yandex positions' in reports. On October 13 it emerged that every row had come from Google instead.
One character, two endpoints
The API provider offers separate endpoints for Yandex and Google search results, both returning an identical JSON structure and the same fields. The script parsed whatever came back without checking which address it had actually queried. A single character in the URL was the only difference between the two, and it went unnoticed because the data looked plausible: positions moved within normal ranges and showed no obvious anomalies.
The fix added a validation step: log the actual endpoint URL on every request, check the search engine identifier returned in the response against what was expected, and throw an error on a mismatch instead of writing to the database.
Different search engine, different rankings
Once the same keywords were pulled correctly from Yandex, positions diverged from the Google data by a wide margin on some queries. Three months of optimisation decisions had been based on the wrong search engine, even though some of the tracked publications draw their actual traffic from Yandex.
For anyone tracking rankings through a third-party API that serves multiple search engines in the same response format, explicit source validation is not optional. A response that looks correct is not proof it came from the engine you asked for.
