We asked five AI assistants for statistics and followed 1,285 of them back to where they came from. Fewer than half led to a study that says the same thing.Mack Grenfell28 October 2026 · 12 min read
\nWe asked ChatGPT, Perplexity, Gemini, Claude and Google’s AI Overviews 200 questions that call for a number, the kind people ask every day. Then we clicked the citations, and kept clicking. Here is one trail, from the answer to the end.
\nGoogle AI Overviews
\nWhat percentage of salespeople give up after one follow-up?
\n44% of salespeople give up after just one follow-up attempt.
\nA sales firm’s stats page 2020
\n44% of salespeople give up after one follow-up attempt (source).
\nA freelance writer’s sample article by 2017
\n80 Percent of Sales Require 5 Follow-up Calls after Meeting; 44 Percent Salespeople Give Up After One Follow-up
\nNo study behind it This page cites no research.
\nIn the answer
\nEvery assistant got the same question: “What percentage of salespeople give up after one follow-up?” All five assistants gave the same figure, and only ChatGPT said the evidence behind it was unclear.
\nOne click in
\nGoogle AI Overviews cited a sales firm’s stats page, from 2020. The number is right there. Its “source” link goes to a writing sample.
\nTwo clicks in
\nThat leads to a freelance writer’s sample article, online by 2017. The number is in its headline. It lists two slide decks, a blog list and two news stories. None is a study of follow-ups.
\nThe end of the trail
\nThat’s where the trail ends. We tried every other route to this number we could find, and each one stopped at a page that cites no research. This is what we mean by a zombie stat: a number that keeps getting repeated with no study behind it.
\nMost numbers aren’t like this one. The trouble is that you can’t tell which ones are from the answer. This one came with the same confident tone and the same kind of citation as any other, and all five assistants agreed on it. So we followed 1,285 statistics the same way.
\nWe took a random sample of the statistics each assistant cited, 1,285 in all, and walked every trail the same way: the cited page, then that page’s source, up to five pages deep. Where a page named a source without a link, we searched for it.
\n40% reached a study that says the same thing. Another 8% reached the study, but the number or its context had changed on the way. The rest never got there: 40% stopped at a source we couldn’t check, because it was locked, gone or silent on the number, and 11% led nowhere at all.
\nEach square is one in a hundred of the 1,285 statistics we traced, sorted by where its trail ended.
\nThe last 1%: trails whose pages wouldn’t load for us.
\nTrakkr Research, five AI assistants asked 200 questions that call for a number, October 2026. A random sample of each assistant’s cited statistics, weighted by how many each quoted, and corrected with 150 trails checked by hand. Sourced: 40%, 95% range 34% to 46%.
\nChatGPT did best, and quoted the fewest numbers: 1.3 per answer, against 6.8 for Claude. The other four landed between 38% and 46%, too close to rank.
\nThe same hundred squares for each assistant, from its own sample of cited statistics.
\nChatGPT
\n57% sourced
\n1.3 statistics per answer
\nPerplexity
\n46% sourced
\n2.2 statistics per answer
\nAI Overviews
\n39% sourced
\n4.8 statistics per answer
\nGemini
\n38% sourced
\n6.2 statistics per answer
\nClaude
\n38% sourced
\n6.8 statistics per answer
\nTrakkr Research, five AI assistants asked 200 questions that call for a number, October 2026. Traced per assistant: ChatGPT 108, Perplexity 277, Google AI Overviews 300, Gemini 300, Claude 300. The four after ChatGPT sit inside each other’s ranges, so they aren’t ranked.
\nTo check this wasn’t down to the questions we picked, we also traced 300 statistics from Google AI Overviews answers to real everyday searches. 40% were sourced, about the same share.
\nThe AI mostly wasn’t making numbers up. The page it cited had the figure 82% of the time. But only 21% of the pages cited for a number were the original research.
\nMore than half were company blogs. Add news stories and statistics roundups and you have most of the citations: pages that repeat figures from somewhere else, often without linking to it.
\nEvery page an assistant cited first for a number, by kind. Each column’s width is its share of those citations. Green is how often its trails reached a study that says the same thing.
\n62%
\nResearch firm 13% of citations
\nResearch firm or survey: 13% of first citations, 62% of them sourced.
\n61%
\nGovernment 7% of citations
\nGovernment or official body: 7% of first citations, 61% of them sourced.
\n37%
\nCompany blog 54% of citations
\nCompany blog: 54% of first citations, 37% of them sourced.
\n37%
\nNews 10% of citations
\nNews or magazine: 10% of first citations, 37% of them sourced.
\n35%
\nStats roundup 8% of citations
\nStats roundup: 8% of first citations, 35% of them sourced.
\n41%
\nOther 8% of citations
\nOther: 8% of first citations, 41% of them sourced.
\nTrakkr Research, five AI assistants asked 200 questions that call for a number, October 2026. 1,268 traced statistics with a readable first page. Page types judged from the page itself. Outcomes corrected with 150 trails checked by hand.
\nThe first page made a big difference. When a trail started at a research firm or a government page, it reached a matching study 61% of the time. From a company blog, 37%; from a statistics roundup, 35%. Even research firms fell short 38% of the time, often because the full report sat behind a form.
\nA citation tells you where the AI found a number, not where the number came from.
\nReaching the study didn’t always mean the number held. Of the statistics that got there, 17% had changed on the way. The digits moved only about one time in four. Usually it was the context: what was measured, the year, the source, or who was asked.
\nWhat changed, across every change we logged
\nThe number changed
\n26%
\nWhat was measured changed
\n20%
\nThe year changed
\n17%
\nThe source changed
\n16%
\nWho was asked changed
\n14%
\nThe place changed
\n4%
\nIn one case, a 2004 study of 15 women judging health websites turned into a claim about first impressions of every website. Below are the clearest trails we followed, each checked by hand.
\nEach trail runs from the original study to an answer an assistant gave in October 2026. Step along it: the number stays, and the words around it change.
\nAskedWhat percentage of first impressions of a website are design-related?
\n\n\nThe study · Sillence and colleagues, CHI, 2004
\nOf the comments 15 women made about rejecting or mistrusting health websites, 94% were about design.
\n
A 2004 study watched 15 British women judge health websites, and 94% of their comments about rejecting a site were about design. Twenty years of retelling made it “94% of first impressions are design-related”, and Google AI Overviews states it as fact.
\nStep 1 of 4
\nEvery step checked by hand on the open web, with archived copies kept. We name the original studies; the pages that repeated them are described by type, not named.
\nAsk five assistants the same question and you’ll often get five different numbers. On 28 of the 71 questions where all five gave a percentage, we got four or five different figures.
\nAsked “What percentage of marketers use marketing automation tools?”, word for word, on the same day.
\n49–51%AI Overviews
\n62%ChatGPT
\n70–79%Gemini
\n81%Perplexity
\n96%Claude
\nChatGPT There isn’t one universal figure, it varies by survey and what counts as “marketing automation.” One commonly cited survey found 62% of U.S. digital marketers used automation software, but that data is from 2020, so it may not reflect current adoption.
\nPerplexityApproximately 81% of marketing organizations use marketing automation.
\nGemini Across major industry benchmarks, between 70% and 79% of marketers and companies use marketing automation tools in some capacity
\nClaude96% of marketers have used a marketing automation platform or plan to use one
\nAI Overviews Roughly 49% to 51% of companies and marketing teams currently use marketing automation tools, according to industry benchmarks from Moosend and related research studies.
\nTrakkr Research, five AI assistants asked 200 questions that call for a number, October 2026. Answers collected on 25 October 2026. A range is drawn from its low to its high.
\nAcross those questions, the typical gap between the highest figure and the lowest was 16 points.
\nEvery question where all five gave a percentage, one line each, from the lowest figure to the highest. Closest at the top.
\nClosest 20 points or more apart Furthest apart
\nTrakkr Research, five AI assistants asked 200 questions that call for a number, October 2026. A range counts at its middle. 2 questions with figures above 100% are left off the scale.
\nWhen all five agreed, the number wasn’t any more likely to check out. They gave the same figure on 10 questions, and of the 5 we traced, 2 were sourced. The statistics quoted by three or more assistants did no better: 11 of 29 were sourced.
\nFour were quoted by all five assistants, and we couldn’t find a study behind any of them.
\nFour statistics every assistant gave us, each with a trail that ends at a page citing no research.
\n44%
\nEvery trail ends at a page that cites no research, one of them a freelance writer’s sample article. Its full trail opens chapter 1.
\nQuoted by all five
\n90%
\nStated with no source. No study with this figure was found.
\nQuoted by all five
\n75%
\nRepeated with no source, often credited to an old marketing roundup that no longer carries it. No study was found.
\nQuoted by all five
\n5 to 25 times
\nA 2014 Harvard Business Review article gives the range “depending on which study you believe”, and names none.
\nQuoted by all five
\nTrakkr Research, five AI assistants asked 200 questions that call for a number, October 2026. Statistics all five assistants quoted whose every trail ended at a page citing no research. Each trail was checked by hand on the open web.
\nEach was quoted by at least three of the five assistants, most often first, and its trail checked by hand.
\n95%
\nGenerative AI pilots that fail to deliver a measurable financial return
\nMIT NANDA’s 2025 report says 95% of organizations got no return from generative AI. That isn’t the same as 95% of pilots failing.
\n5 of 5 Changed Study from 2025
\n44%
\nSalespeople who give up after one follow-up
\nEvery trail ends at a page that cites no research, one of them a freelance writer’s sample article. Its full trail opens chapter 1.
\n5 of 5 Nowhere
\n4% to 10%
\nUnhappy customers who actually complain
\nGoes back to TARP’s 1979 research for the White House consumer office. Its author says rates that low apply to small purchases; in 2016, 22% to 41% complained about their most serious problem.
\n5 of 5 Changed Study from 1979
\n2 hours 21 minutes
\nTime people spend on social media each day
\nDataReportal’s Digital 2025 report, from GWI survey data: the typical internet user spends 2 hours 21 minutes a day on social media.
\n5 of 5 Sourced Study from 2025
\n46%
\nGoogle searches with local intent
\nTraces to a 2018 social post about a slide at a search conference. Google never published it, as far as we could find.
\n5 of 5 Can’t check
\n90%
\nSocial media users who follow at least one brand
\nStated with no source. No study with this figure was found.
\n5 of 5 Nowhere
\n94%
\nFirst impressions of a website that are about its design
\nBegan as a 2004 study of how 15 women judged health websites. Its full trail is in chapter 4.
\n5 of 5 Changed Study from 2004
\n75%
\nSearchers who never scroll past the first page of results
\nRepeated with no source, often credited to an old marketing roundup that no longer carries it. No study was found.
\n5 of 5 Nowhere
\nTrakkr Research, five AI assistants asked 200 questions that call for a number, October 2026. Statistics grouped by figure and meaning. Original year shown where the trail reached the study.
\nTreat it as a lead rather than a source. Here’s what we’d do, whether you use these numbers or publish them.
\nIf the cited page is a list of statistics or a company blog, use its own link to go one page further. Keep going until you reach the study itself, or until you can’t.
\nMost of the changes we found were in the context, not the digits: one country’s poll, a ten-year-old study, fifteen people.
\nFive assistants giving the same number isn’t five sources. The numbers they repeated most were no better sourced than the rest.
\nThe sample, the date and a link to the source, on the same page as the figure, so the next page, and the next answer, can trace it back to you.
\nReports behind a form were a common dead end: even trails that started at a research firm fell short 38% of the time. Put the headline numbers and the method on an open page.
\nIt is often a roundup or a blog that copied you, sometimes with an older figure. That page decides what people hear.
\n