Future of search research by Jacob Whitmore
Where search is heading. Called early, scored in public.
How people find things is changing. Answer engines decide whether to credit a source, two people asking the same question get different answers, and publishers decide what machines can read. I measure it on stated dates, publish the raw data, and call where it goes next. Every call has a check date, and misses stay up, in red.
The scoreboard
As of 10 October 2026
11 calls: 11 open, 0 hit, 0 miss, 0 void.
At least 7 of the UK’s 10 biggest newsbrands will still block AI training crawlers more often than the bots that fetch a page when someone asks a question.
What I’m measuring
Who still gets the click?
When an answer engine uses someone’s work, it can link, name the source, or do neither.
Study registered. Run due December 2026.
What do publishers let machines read?
The robots.txt files of the UK’s ten biggest newsbrands, against 15 AI crawlers, every month.
Latest: 2 October 2026. Next: 2 November 2026.
Does everyone get the same answer?
The same question, asked by a plumber and by a finance director, gets different brands.
Do answers say when their facts were true?
Prices and market figures quoted by answer engines, checked for a date.
Latest research
- 9 Oct 2026
Does ChatGPT recommend different brands to different customers?
Three of the four finance-director answers recommended none of the products the no-persona answers named.
Download the CSV - 9 Oct 2026
How much is the purple pound worth?
Both AI engines that answered gave £274bn, and neither said it was a 2020 figure.
Download the CSV - 2 Oct 2026
Which UK news sites block GPTBot and other AI crawlers?
Nine of the UK’s ten biggest newsbrands block at least one AI crawler, and they block training bots far more often than the bots that fetch pages for answers.
Download the CSV
How it works
- MeasureI collect the data myself, on one stated date. No surveys, no borrowed numbers.
- PublishThe answer in the first line, one plain table, the full CSV and the method.
- Call itA dated prediction, with what counts as a hit and a miss written down before the result.
- Score itOn the check date I rerun the test. Misses stay up, in red. Nothing gets reworded.
Some calls can’t be scored. Here’s one I made in May that I can’t mark, and what it should have said.
Who’s measuring
I’ve worked in SEO for around 12 years, across in-house, agency and freelance roles. This site is where I test my own calls about where search is heading. More about me
Follow the data
A new crawler snapshot lands on the 2nd of every month, and scores land on each call’s check date. I post both on LinkedIn first.