Don’t guess. Measure.

Jacob Whitmore

Independent research into the future of search.

Search is changing. I measure what happens next. Original research, public datasets and predictions scored in public.

Explore the research The Jacob Whitmore Index

Jacob Whitmore, portrait in round glasses and a dark jacket
Jacob Whitmore
  • 12 yearsin SEO, across agency, freelance and in-house roles
  • Open dataa CSV with every study, free to reuse with a link
  • On the recorddated predictions, scored right or wrong, misses left up
  • MonthlyUK AI-crawler snapshots, running since October 2026

01 Latest research

Original research. Data attached.

Each study states when the data was collected, the sample, the method and its limits. The data file is published with it.

Latest study · 18 Nov 2026
40 of 40everyday UK questions where Google showed its own answer above the first ordinary link.

Google answered 40 of 40 UK questions before the first link

Question
What does Google put above the first ordinary link on everyday UK questions?
Collected
18 November 2026, google.co.uk, desktop
Sample
40 questions in four groups of ten: quick facts, how-to, health and money
Finding
An AI Overview on 38, Google’s own answer box on 3. On 22 of 40, no ordinary link was visible without scrolling.
Limits
A pilot. One location, one browser, one morning. It shows what Google displayed, not what searchers clicked.
Next
Same 40 questions, around 18 December 2026

Read the study Get the data

02 Findings

What the evidence says so far.

Each figure is one dated reading with the data behind it. Free to quote with the date and a link.

Nine of the UK’s ten biggest newsbrands block at least one AI crawler in robots.txt.
Read 2 Nov 2026 · Study and data
None of the UK’s ten biggest supermarkets blocks any of the same 15 AI crawlers.
Read 10 Nov 2026 · Study and data
None of 32 widely repeated AI search statistics passes all five source checks.
Read 29 Oct 2026 · Study and data

03 The Jacob Whitmore Index

The numbers, in one table.

A fixed set of readings on how UK websites and AI search treat each other. Each has the date it was measured and the date it gets measured again.

MeasureLatest readingMeasuredNext reading
JWI-1 · AI crawler checks blocked by the UK’s ten biggest newsbrands99 of 1502 Nov 20262 Dec 2026
JWI-3 · The same checks blocked by the UK’s ten biggest supermarkets0 of 15010 Nov 202610 Feb 2027
JWI-4 · The same checks blocked by the UK’s ten most-visited retail sites27 of 15012 Nov 202612 Feb 2027
JWI-8 · Most-repeated AI search statistics that pass all five source checks0 of 3229 Oct 202629 Jan 2027
JWI-9 · AI visibility trackers that publish evidence their numbers are accurate0 of 1330 Oct 202630 Apr 2027
JWI-11 · Everyday UK questions Google answered above the first ordinary link40 of 4018 Nov 202618 Dec 2026

Six of the twelve current readings, as published on 19 November 2026. A reading is never edited. A new study adds a new one.

See the full Index The studies behind it

04 Public Prediction Ledger

Called early. Scored in public.

Each prediction is written down with a date and a test before the result is in. On the check date it gets marked. Wrong calls stay up, in red, unedited.

Right
2
Wrong
1
Still open
30
Couldn’t score
0

33 on the record as of 19 Nov 2026. Open calls count neither way. Board as CSV ↓

P19 RightLatest scored · written down 14 Oct 2026 · checked 30 Nov 2026

After the follow-up 'When is that figure from? Give the month or period and the official source.', at least 95% of AI answers will state a period for the figure.

Scored 18 Nov 2026: hit. 80 of 80 arm D second answers stated a period (75 of 80 if a year is required; the registered test does not require one). Results: https://jacobwhitmore.com/date-follow-up-results/

Due next

  1. P17On the first Russell Group robots.txt snapshot, a smaller share of the 24 universities will block GPTBot than the UK's ten biggest newsbrands did (7 of 10).Written 14 Oct 2026 · checked 2 Dec 2026
  2. P18On the first Russell Group snapshot, universities will block training crawlers more often than retrieval crawlers, by at least 10 percentage points.Written 14 Oct 2026 · checked 2 Dec 2026
  3. P38On the 18 December 2026 rerun of the same 40 UK questions, Google will show an answer (an AI Overview or its own answer box) above the first ordinary link on at least 38 of the 40.Written 19 Nov 2026 · checked 18 Dec 2026

See every prediction How a prediction is written

05 Current investigations

What I’m measuring next.

Work that is planned or under way, with the date each result is due. Nothing here has a result yet.

InvestigationWhere it standsDue
UK news sites and AI crawlers, reading threeMonthly series. Two readings published.2 Dec 2026
Russell Group universities and AI crawlersFirst snapshot. Predictions P17 and P18 get checked against it.2 Dec 2026
Undocumented AI crawler rulesPlan published. First reading runs with the December news snapshot.2 Dec 2026
Google answers above the first link, reading twoPilot published 18 Nov 2026. Same 40 questions again.Around 18 Dec 2026
The industry’s 2026 SEO predictions, scoredPlan published. Pages collected 1 Dec 2026, scored in January.By 31 Jan 2027

Every registered plan

06 Selected analysis

Worth reading first.

Picked for what they settle, not how new they are.

Search the library Ideas & Perspectives All analysis

  1. Does publishing original research get you cited by AI?

    No controlled study shows original research raises AI citations. Why it should help AI use your work, why credit is less certain, and the test this site will score on 2 Jan 2027.

  2. Will AI agents do your shopping? Three of the UK’s biggest retail sites have said no in writing

    Checked 12 November 2026. Amazon, eBay and Vinted tell AI shopping agents to stay out of checkout. Seven of the UK's ten biggest retail sites say nothing. What the files and one court ruling show.

  3. How do ChatGPT, Claude, Gemini, Perplexity and Copilot choose their sources? What the companies actually say

    Checked 8 Nov 2026: what OpenAI, Anthropic, Google, Perplexity and Microsoft say about how their AI answers find and pick sources. Only one explains how it chooses.

  4. How many times should you run an AI prompt before you trust the number?

    To pin an AI brand-mention rate to within ±10 points you need about 97 runs of one prompt. To show a change worked, about 388 per version. The maths and a free calculator.

  5. Do backlinks get you cited in AI answers? What 16 studies actually measured

    Checked 30 October 2026. 16 studies on links, authority and rankings vs AI citations. All correlational, none with raw data, one naming a country. Brand mentions beat links.

  6. Gartner said search volume would drop 25% by 2026. Can it be scored?

    Gartner's 2024 call that search volume would fall 25% by 2026, checked in October 2026. It can't be scored as written. On the closest public data, it looks like a miss.

07 Press and media

For journalists.

Bios, a press photo, quotable findings and citation guidance are in the media centre. Mistakes get corrected in public under the editorial policy.

Jacob Whitmore walking along a dry-stone wall on a hillside path

08 About Jacob

Hello, I’m Jacob.

I’ve worked in SEO for around 12 years. Now I study what happens when search changes.

Who gets discovered, which businesses AI tools recommend, and what the evidence says about where search is heading. I test it with small, repeatable studies and publish the data behind each one.

This site is where it all goes: the findings, the method, and the calls I get wrong.

Read the full biography

09 Follow the research

Follow the research.

contact@jacobwhitmore.com

New results go up on LinkedIn first. The RSS feed carries every study and analysis piece. For press, research questions or corrections, email is fastest.

Follow on LinkedIn RSS feedContact