Notes
What the weekly measurement turns up, and the engineering decisions behind it.
Index
All notes
| 3 Sept 2026 | Twelve weeks of measuring Google, and the four numbers that came out
We ran 4,185 queries through Google once a week for 12 weeks. Here is what held steady and what did not. |
| 27 Aug 2026 | Google cites what it does not rank
38.8% of AI Overview citations point at domains absent from the ten results below them. |
| 20 Aug 2026 | The same query in ten countries shares about half its results
Run an identical query from nine other countries and roughly half the domains change. |
| 13 Aug 2026 | Absent, not null: a response contract worth copying
The difference between a missing key and a null value is the difference between a clean integration and a 3am page. |
| 6 Aug 2026 | How to build a query corpus worth measuring
A measurement is only as good as the list it runs on. Here is how ours is built, and why it is generated rather than curated. |
| 30 Jul 2026 | Nine SERP APIs, one response body: what actually differs
Field names, block coverage, vertical count and metering policy — in roughly that order of how much trouble they cause. |
| 23 Jul 2026 | Pagination without an upper bound, and why page 40 matters
Most search APIs quietly stop somewhere around page ten. The work that needs page forty is the work worth doing. |
| 16 Jul 2026 | Why we publish our own data instead of describing it
Every search API claims to be accurate and comprehensive. None of them shows you the results. So we show ours. |
Filter