- Matches the source 72
- Restated wrong 17
- No study behind it 4
The enterprise AI debate runs on a dozen numbers, and almost nobody quotes them correctly. Type the one you heard. See what the study measured, its sample, and how it gets restated wrong.
- 93studies, reports and enforcement actions
- 17circulate in a form the study does not support
- 4have no study behind them at all
- 50report the size of their sample
Every study in the index.
Newest first. Each one with its verdict, what it measured, its sample and the primary source.
- Matches the source
Adobe Analytics2026, method not recorded
What it measuredUS online holiday spending reached $257.8 billion between November 1 and December 31, 2025, up 6.8%. Traffic to US retail sites referred from generative AI sources rose 693% year over year.
SampleMore than 1 trillion visits to US retail sites, 100 million SKUs and 18 product categories, published January 7, 2026
The fine print
Vendor-published data from Adobe's own analytics customers, US only. The 693% is growth from a small base and says nothing about conversion or revenue from AI referrals. Use it as direction of travel, never as a share of traffic.
Holiday shopping season drove a record $257.8 billion online with consumers embracing generative AI tools
- Matches the source
American Transportation Research Institute2026, method not recorded
What it measuredThe average marginal cost of operating a truck reached $2.336 per mile in 2025, up 3.4% and the highest since the series began in 2016. Operating margins in the truckload and refrigerated sectors stayed below 1.0%, flatbed carriers ran at a 0.5% loss, and tank carriers averaged 4.0%.
SampleAnnual survey of motor carriers' financial and operational data, 2025 data year, published July 15, 2026
The fine print
US carrier costs in US dollars. Participating carriers self-select into the survey and the release does not publish the carrier count. Costs are per mile, not per load, and margins are sector averages, so they say nothing about any one fleet or brokerage.
An Analysis of the Operational Costs of Trucking: 2026 Update
Primary source, American Transportation Research InstituteLink to this study
- Matches the source
Association of Corporate Counsel and FTI Consulting2026, method not recorded
What it measured36% of legal departments are actively deploying GenAI. 35% cite chronic budget and resource constraints as their top barrier. 63% expect headcount to stay flat.
Sample1,049 chief legal officers across 43 countries, released January 29, 2026
The fine print
CLO respondents skew larger than 10 to 100 person companies. Global sample, self-reported. ACC's site blocks automated fetches, wording verified via the syndicated press release.
2026 ACC Chief Legal Officers Survey
Primary source, Association of Corporate Counsel and FTI ConsultingLink to this study
- Matches the source
BizBuySell2026, method not recorded
What it measuredSmall-business AI adoption reached 63-88% depending on survey, with productivity gains the #1 cited motivation at 78%.
SampleNo sample published.
The fine print
Adoption range spans multiple surveys with different definitions of 'using AI'; cite as a range, never a single number.
Small Business AI Adoption Survey
- Matches the source
BuildForce Canada2026, method not recorded
What it measuredCanada's construction industry must hire 306,200 workers by 2034 to cover 32,100 workers of growth plus projected retirements. About 271,900 new entrants under 30 are expected, leaving a shortage of as many as 34,300 workers. In residential construction alone, 135,000 workers, more than one-fifth of that sector's labour force, are projected to retire by 2035.
SampleBuildForce Canada national labour market forecast, built on Statistics Canada labour force and investment data with industry validation, published July 20, 2026
The fine print
A forecast, not a count. The 135,000 retirements figure is residential only and quoted from the Canadian Home Builders' Association in the release. The release itself mixes 2034 and 2035 horizons, so carry the horizon with each number.
Construction and Maintenance Looking Forward, 2026 to 2035 national summary
- Matches the source
Canadian Medical Association and Canadian Federation of Independent Business2026, method not recorded
What it measuredPhysicians in Canada spend about nine hours a week on administrative tasks, roughly one-fifth of a 49-hour week and 42.7 million hours a year nationally. Respondents judged 47% of that time, about 19.8 million hours, unnecessary for a physician to do, the equivalent of 9,093 full-time physicians. Family physicians report 9.9 hours a week. More than half plan to reduce their hours.
Sample1,924 practising physicians, residents and fellows, open-link voluntary survey fielded July 31 to August 21, 2025, published January 26, 2026
The fine print
Self-selected respondents, so burdened physicians may be over-represented, and the unnecessary share is the respondent's own judgment. The 9,093 figure is an arithmetic equivalence, not physicians who would be hired. The CMA site blocks automated fetches, verified from the report PDF in the CMA digital library.
Losing doctors to desk work: Canadian physicians lose 20 million hours each year to red tape
Primary source, Canadian Medical Association and Canadian Federation of Independent BusinessLink to this study
- Restated wrong
Cloudera / Harvard Business Review Analytic Services2026, survey
Usually quoted as
Only 7% of enterprises are AI-ready, according to Harvard.
What it measuredOnly 7% of enterprises say their data is completely ready for AI; 73% say preparing data for AI is challenging; top obstacles are siloed data (56%) and no clear data strategy (44%).
Sample231 members of the Harvard Business Review audience involved in their organization's AI data decisions, online survey fielded October 2025. The full report is free and ungated and publishes its methodology and participant profile, including organization size and industry.
The fine print
It measured self-reported readiness of data, not of the enterprise. Respondents rated their own organization's data for AI adoption on a five option scale and 7% picked the top option, completely ready, against 15% almost completely ready, 51% somewhat ready and 27% not very or not at all ready. The base is 231 members of the Harvard Business Review audience, surveyed online in October 2025, all involved in their organization's data decisions and all at organizations already practising, piloting or exploring AI, so the sample is screened toward AI adopters. HBR publishes the organization size, industry, region and seniority splits in a free report, though nothing at country level. Cloudera sponsored the report and sells a hybrid data platform, so a data readiness gap is the sponsor's own sales case, and the usual misquote drops both the word data and the word say.
Enterprise data readiness for AI
Primary source, Cloudera / Harvard Business Review Analytic ServicesLink to this study
- Restated wrong
Construction Owners2026, survey
Usually quoted as
85% of construction firms say AI has cut the time they spend on repetitive tasks.
What it measuredDodge Construction Network, SmartMarket Brief "AI for Contractors", released 5 December 2025 in partnership with CMiC. 235 US general and trade contractors surveyed in September and October 2025. 87 percent expect AI to transform construction. 19 percent have actually adapted a workflow.
Sample235 general and trade contractors in the United States, surveyed September and October 2025
The fine print
The research is not by Construction Owners Club, the site that carried it. constructionowners.com is a construction industry news and marketing site with no disclosed corporate entity and no named editorial staff, and its write-up does credit the original: it names Dodge Construction Network and CMiC, quotes Dodge's Steve Jones, and links to Construction Dive's earlier reporting. The study itself is Dodge Construction Network's SmartMarket Brief "AI for Contractors", released 5 December 2025 with CMiC, a construction financial and project management software vendor, as partner. It surveyed 235 US general and trade contractors in September and October 2025, so it measured contractors rather than owner organizations, and the year is 2025 rather than 2026. Nearly every headline figure is an expectation about the future: 87 percent expect transformation against 19 percent who have changed a workflow.
Construction AI adoption doubles in 2026
- Restated wrong
Deloitte2026, survey
Usually quoted as
Deloitte found that 80% of companies deploying AI agents have no governance in place.
What it measured21% of the companies that plan to deploy AI agents within two years say they have a mature agent governance model. That group is close to three-quarters of the 3,235 leaders surveyed, so the figure is about 15% of all respondents, not 1 in 5 of all organizations.
Sample3,235 business and IT leaders, director through C-suite, split equally between IT and line of business, all with direct involvement in their organization's AI work, across 24 countries in the Americas, Asia Pacific, Europe and the Middle East, and 6 industries. Fielded August and September 2025. Published 21 January 2026 at Davos.
The fine print
Deloitte publishes this figure two different ways and the gap matters. Its press releases say 21% of the companies planning to deploy AI agents within two years report a mature agent governance model, and that group is close to three-quarters of the 3,235 leaders surveyed. Its Deloitte Insights article drops the restriction and says 21% of all respondents, adding that roughly 80% of organisations surveyed lack mature governance capabilities. Cite whichever base you use and say which. Either way the finding is about governance that is not yet mature, measured on agent decision boundaries, real-time monitoring and audit trails, so "no governance in place" overstates it. Maturity is self-assessed by the respondent and never audited, and Deloitte sells AI governance consulting. Fieldwork ran in August and September 2025, not 2026.
State of AI in the Enterprise 2026
- Matches the source
Higher Education Policy Institute and Kortext2026, method not recorded
What it measured95% of UK undergraduates use generative AI in at least one way and 94% have used it to help with assessed work, up from 92% and 88% in the 2025 edition. 12% say they include AI-generated text directly in assessed work. 65% report their assessments have changed substantially in response to AI.
Sample1,054 full-time UK undergraduates, fielded by Savanta in December 2025, published March 12, 2026
The fine print
UK undergraduates only, online panel. 'Help with assessed work' includes explaining concepts and summarising articles, so it is not a misconduct rate. The 12% is the only figure about submitting AI text. Co-published with a courseware vendor.
Student Generative AI Survey 2026
Primary source, Higher Education Policy Institute and KortextLink to this study
- Matches the source
IDC2026, method not recorded
What it measured88% of agentic AI pilots never reach production.
SampleNo sample published.
The fine print
Reported via press coverage of IDC research rather than the primary report. Use it as the reason we gate on evals and scope, not as a precise industry benchmark.
Agentic AI pilot-to-production rates
- Matches the source
JAMA2026, method not recorded
What it measuredClinicians using ambient AI scribes saved 16 minutes of documentation time and spent 13 fewer minutes in the medical record per 8 hours of patient care, and added roughly one extra patient visit every two weeks.
Sample1,800 clinicians across 5 academic medical centres, 2023 to 2025
The fine print
The denominator is per 8 hours of patient care, NOT per encounter, and the study characterizes the savings as modest against vendor claims. It found no significant effect on after-hours EHR time. Never restate this as a per-visit number.
Ambient AI scribes and clinician documentation burden
- No study behind it
JPMorgan Chase2026, method not recorded
Usually quoted as
JPMorgan's AI deployment saves $2 billion a year and delivers 30% to 40% productivity gains across 230,000 employees.
What it measuredNo stat was supplied and no study exists to correct against. If a JPMorgan figure is cited at all, attribute it as a company statement. Dimon said in an October 2025 Bloomberg interview that roughly $2 billion of AI expense returned roughly $2 billion of benefit.
Sample1,469 leaders of US midsize companies (annual revenue $20 million to $500 million), online survey November 4 to 25, 2025, margin of error +/- 2.5 points, published January 7, 2026
The fine print
JPMorgan has published no study behind these numbers. The figures are executive statements and trade coverage, not research with a sample or a method. In an October 2025 Bloomberg interview Jamie Dimon said that for about $2 billion of AI expense the bank saw about $2 billion of benefit, which is a wash rather than a net saving. The 230,000 employees on its internal LLM Suite, the 450 production use cases and the 30% to 40% efficiency figures come from company statements with no published measurement method and no external audit. Attribute all of it to JPMorgan as a company claim, never as measured evidence.
2026 US Business Leaders Outlook
- Matches the source
JPMorganChase Institute2026, method not recorded
What it measured17.7% of US small businesses were paying for AI tools by the end of 2025, and the newest adopter cohort reached 10% adoption within 6 months.
SampleNo sample published.
The fine print
Built from anonymized Chase small business payment records, so it counts paid AI tools only and free-tier use is invisible. US firms, not Canada. Confirm the exact report title and publication date before citing.
JPMorganChase Institute research on small business adoption of AI tools
- Restated wrong
Kyndryl2026, survey
Usually quoted as
Kyndryl's 2025 survey of 3,700 executives found only 29% say their AI is ready to manage future risks.
What it measuredThe 2025 Kyndryl Readiness Report reports 54% of leaders seeing a positive return on their AI investments, up 12 points year over year, against 36% who say their AI is completely ready to manage future risks and 62% who say they are still in the experimentation phase.
Sample3,700 senior leaders and decision makers across 21 countries, fielded by Edelman DXI for Kyndryl between June 20 and August 15, 2025, by online survey and telephone interview. 50% C-suite, 50% senior directors and business unit leaders. Half represented companies with $1 billion or more in revenue. Paired with aggregated telemetry from 1,200 companies on Kyndryl Bridge, dated August 2025.
The fine print
All of these are executive perceptions collected by Edelman DXI for Kyndryl, not measured returns, and Kyndryl sells the IT modernization work that a readiness gap implies. The 2025 report surveyed 3,700 senior leaders in 21 countries between June 20 and August 15, 2025. It gives 54% seeing a positive return on AI investments, up 12 points year over year, 36% saying their AI is completely ready to manage future risks, up 7 points, and 62% saying they are still in the experimentation phase. The 29% often quoted as the AI readiness number is the 2024 wave of the same annual report, republished by Kyndryl as an AI Readiness Report. A separate 29%, workforce readiness, comes from Kyndryl's People Readiness Report, a different survey of 1,100 leaders in 8 markets.
2026 People Readiness Report: Beyond AI Adoption
- Matches the source
National Automobile Dealers Association2026, method not recorded
What it measuredThe average franchised US dealership wrote 16,252 repair orders in 2025 and booked $9.69 million in service and parts sales, $494 per customer repair order, at an average customer labour rate of $186 an hour. Across 16,990 dealerships that was 276 million repair orders and $164.6 billion, 13.3% of total dealership sales dollars.
SampleNADA's annual financial profile, compiled from dealer financial statements and industry data for full-year 2025
The fine print
US averages in US dollars, skewed by large stores, and per-store figures are means, not medians. NADA Data does not publish a service absorption rate, so do not attribute one to it. Canadian dealer economics come from CADA, not NADA.
NADA Data 2025: Annual Financial Profile of America's Franchised New-Car Dealerships
Primary source, National Automobile Dealers AssociationLink to this study
- Matches the source
North American Electric Reliability Corporation2026, method not recorded
What it measuredAggregated summer peak demand across North America is forecast to rise by over 224 GW in the next 10 years, 69% higher than the 132 GW projected a year earlier, and winter peak demand by 245 GW. New data centers for AI and the digital economy account for most of the increase. 13 of 23 assessment areas face resource adequacy challenges over the 10 years.
SampleLoad-serving entity and system planner forecasts submitted to NERC in mid-2025, covering the contiguous US, Canada and northern Baja California
The fine print
A forecast built from planner submissions, not a measurement, and NERC itself notes that large loads add volatility as project timelines shift. Published January 2026 under the 2025 title. The widely quoted 24% figure is a press calculation and is not in the report.
2025 Long-Term Reliability Assessment
Primary source, North American Electric Reliability CorporationLink to this study
- Matches the source
Recon Analytics2026, method not recorded
What it measuredOnly 35.8% of workers with paid Copilot access actually adopt it, vs 83.1% for ChatGPT.
SampleUS worker survey, July 2025 to January 2026
The fine print
The companion '64% of Copilot licenses unused' figure comes from vendor adoption analyses, not an independent study; treat it as directional and cite the 35.8% instead.
AI Choice 2026: Why Licenses Don't Equal Adoption
- Matches the source
Salesforce2026, method not recorded
What it measuredSellers spend 40% of their time actually selling. 48% of reps say they lack the bandwidth to do adequate cold outreach. 87% of sales organizations already use some form of AI.
Sample4,050 sales professionals, surveyed late 2025
The fine print
Vendor-published by a company selling sales AI. Global sample across company sizes, self-reported time allocation, not Canada-specific and not limited to mid-size companies. Confirm exact field dates on the report page before citing them.
State of Sales 2026
- Matches the source
SHRM2026, method not recorded
What it measured39% of organizations have brought AI into their HR function, with recruiting the leading use at 27%. 56% of HR functions that adopted AI never formally measure whether it worked.
SampleNo sample published.
The fine print
SHRM member survey of HR professionals, US-weighted, self-reported. Confirm respondent count and field dates on SHRM's research page before citing.
The State of AI in HR 2026
- Matches the source
Statistics Canada2026, method not recorded
What it measured19.2% of Canadian businesses used AI to produce goods or deliver services in the 12 months to Q2 2026, up from 12.2% in Q2 2025 and 6.1% in Q2 2024. Among businesses with 100 or more employees the figure is 27.8%. The finance and insurance industry leads at 40.4%. Data analytics is the most common use, named by 36.6% of AI-using firms.
Sample9,251 business responses from a 21,105 sample, fielded April to May 2026 (Canadian Survey on Business Conditions)
The fine print
Government survey measuring business-level AI adoption across all functions, never a single department. The 40.4% is the finance-and-insurance industry classification, not the finance function, and the 36.6% is a share of AI-using businesses, not of all businesses.
Analysis on artificial intelligence use by businesses in Canada, second quarter of 2026
- Matches the source
Statistics Canada2026, method not recorded
What it measuredCanadian farm cash receipts rose 4.7% to $102.2 billion in 2025, but farm operating expenses rose faster, up 5.1% to $83.0 billion, with fertilizer up 6.9% to $9.7 billion, so realized net income slipped 0.3% to $8.3 billion.
SampleNational farm income accounts, released May 27, 2026
The fine print
Sector aggregates, not a per-farm figure, and realized net income excludes inventory change. A national average hides wide spreads by commodity and province. Say expenses grew faster than receipts, never that farms lost money.
Farm income, 2025
- Matches the source
Statistics Canada2026, method not recorded
What it measuredAccommodation and food services had the highest job vacancy rate of any sector in Canada in June 2026, at 4.3%, against a national rate of 2.8% and 509,100 vacancies nationally.
SampleJob Vacancy and Wage Survey and Survey of Employment, Payrolls and Hours, reference month June 2026, released August 27, 2026. Sector rate from Table 14-10-0406-01, seasonally adjusted
The fine print
A vacancy rate, not a turnover rate. It measures unfilled positions as a share of labour demand in one month and moves seasonally. Use it for the hardest sector to staff, never as a churn figure.
Payroll employment, earnings and hours, and job vacancies, June 2026
- Matches the source
The CMO Survey (Deloitte, Duke Fuqua School of Business, American Marketing Association)2026, method not recorded
What it measuredGenerative AI was used in 22.4% of marketing activities in 2026, up from 15.1% in 2025 and 7.0% in 2024. Companies project AI will account for 55.9% of marketing activities within three years. Marketing headcount grew 2.5% over the past year, down from 5.4% in 2025. 41.5% of companies already work on showing up in AI-generated answers.
Sample308 marketing leaders at US for-profit companies (14.6% response rate, 97% VP level or above), in the field January 7 to 29, 2026
The fine print
US marketing leaders, self-reported, skews larger companies. The 55.9% three-year projection covers AI and machine learning overall, not generative AI alone.
The CMO Survey 2026: Highlights and Insights Report (35th edition)
Primary source, The CMO Survey (Deloitte, Duke Fuqua School of Business, American Marketing Association)Link to this study
- Matches the source
Thomson Reuters Institute2026, method not recorded
What it measured41% of law firms and 47% of corporate legal departments say their organizations already use generative AI. Across all respondents organizational use rose to 40%, up from 22% in the 2025 edition.
Sample1,514 legal, tax, accounting, risk and government professionals in 27 countries, online survey October and November 2025, published February 2026
The fine print
Not limited to in-house teams or mid-size companies. Self-reported.
2026 AI in Professional Services Report
- Matches the source
Wolters Kluwer2026, method not recorded
What it measuredOver 90% of legal professionals report using at least one AI tool, while firm-level adoption sits far below individual use.
SampleNo sample published.
The fine print
Individual tool use is not firm deployment. Cite this as the gap between personal use and governed firm systems, never as firm-level adoption.
Future Ready Lawyer Survey 2026
- Matches the source
Zendesk2026, method not recorded
What it measured74% of consumers now expect 24/7 service because of AI.
Sample6,182 consumers in 22 countries
The fine print
Vendor-published research from a helpdesk company with an AI product to sell. Measures consumer expectations, not business results. Confirm field dates on the report page before citing.
Zendesk CX Trends 2026
- Matches the source
AICPA and CIMA2025, method not recorded
What it measuredUS accounting bachelor's and master's completions fell 6.6% to 55,152 in the 2023-24 academic year, and new CPA exam candidates dropped from 42,626 in 2023 to 28,082 in 2024. Accounting enrollment then grew 12% year over year for two consecutive semesters in 2024-25, and 75% of firms that hired in 2024 expect to hire the same number or more.
SampleIPEDS degree completions, National Student Clearinghouse enrollment data and an AICPA survey of public accounting firms, published October 27, 2025
The fine print
US pipeline data. Canadian CPA supply runs through CPA Canada and the provincial bodies, which publish no comparable series. The 2024 candidate drop partly reflects the CPA Evolution exam transition, so it is not a pure demand signal. Firm hiring figures are self-reported.
2025 Trends: A Report on Accounting Education, the CPA Exam and Public Accounting Firms' Hiring of Recent Graduates
- Matches the source
AIxEnergy2025, method not recorded
What it measuredEnergy companies report roughly 170% average ROI on deployed AI despite below-average adoption rates.
SampleNo sample published.
The fine print
Industry-analysis estimate; treat as directional.
What US utilities are actually doing with AI
- Restated wrong
American Bar Association2025, survey
Usually quoted as
Nearly half of law firms are now using AI.
What it measuredAmerican Bar Association, 2024 Legal Technology Survey Report, released March 2025: 30% of responding attorneys reported using AI tools, up from 11% in 2023. By firm size, 46% at firms with 100 or more attorneys, 30% at firms with 10 to 49 attorneys, and 18% at solo practices.
Sample512 attorneys in private practice in the United States, fielded 2024. Distributed to a sample of ABA members with voluntary response, so the risk is nonresponse bias rather than self-selection. No response rate is published, and per-band counts are not published either.
The fine print
There is no 2026 ABA Legal Technology Survey Report. The most recent edition with AI figures broken out by firm size is the 2024 report, released 3 March 2025, based on 512 United States attorneys in private practice. Response was voluntary and the ABA publishes no response rate, so the respondents are not a random cross-section of the profession. The report does not define "artificial intelligence", a gap Robert Ambrogi flagged in his review, so it is not clear whether respondents counted generative tools only or any AI-based feature. That ambiguity also reaches the jump from 11% in 2023 to 30% in 2024. Per-band respondent counts are not published, so the solo and largest-firm figures rest on small cells.
2024 Legal Technology Survey Report
- Matches the source
American Medical Association2025, method not recorded
What it measuredThe Permanente Medical Group's ambient AI scribes saved physicians an estimated 15,791 hours of documentation time.
SampleNo sample published.
The fine print
A 2026 multi-center study (STAT) found more modest average savings, roughly 16 minutes per 8 hours of patient care. Cite the Permanente figure as one health system's estimate, not a universal result.
AI scribes save 15,000 hours and restore the human side of medicine
Primary source, American Medical AssociationLink to this study
- Matches the source
Association for Financial Professionals2025, method not recorded
What it measured23% of FP&A practitioners use AI on a daily, weekly or monthly basis. 40% are in a testing phase and plan to implement AI in the next year.
Sample362 FP&A practitioner responses
The fine print
Global and self-selected from AFP's membership, not Canada-specific and not limited to mid-size companies. Survey field dates not stated on the public page.
2025 AFP FP&A Benchmarking Survey Report: Technology and Data
Primary source, Association for Financial ProfessionalsLink to this study
- Matches the source
Association of Corporate Counsel and Everlaw2025, method not recorded
What it measuredNearly 60% of in-house counsel reported no noticeable savings from their outside counsel's use of AI. Of the 40% who did see a benefit, only 13% pointed to fewer billable hours.
SampleNo sample published.
The fine print
Measures what clients perceived, not what firms saved. The honest reading is that efficiency gains are not reaching clients, which is a pricing question before it is a technology one.
In-house counsel on outside counsel AI use
Primary source, Association of Corporate Counsel and EverlawLink to this study
- Matches the source
Association of Corporate Counsel and Everlaw2025, method not recorded
What it measured52% of in-house legal professionals actively use GenAI, up from 23% in 2024.
Sample657 in-house legal professionals in 30 countries, released October 14, 2025
The fine print
The year-over-year comparison covers US respondents only, while the 52% is the 30-country sample. Scope any doubling language to US respondents. Co-published with a legal-tech vendor, verified via everlaw.com/press/release/acc-report-2025. A separate 2025 ACC and Everlaw study (id accEverlaw2025) measures perceptions of outside counsel AI use. Do not conflate the two.
Generative AI's Growing Strategic Value for Corporate Law Departments
Primary source, Association of Corporate Counsel and EverlawLink to this study
- Restated wrong
BCG2025, vendor analysis
Usually quoted as
BCG research shows 70% of AI value comes from people and process, not technology.
What it measuredBCG's 10-20-70 rule advises companies to put 10 percent of their AI effort into algorithms, 20 percent into data and technology, and 70 percent into people and process change. It is guidance from BCG's client case work, not a measurement of where value came from.
SampleNo sample is published for the 10-20-70 rule itself. The report it appears in, BCG AI Radar 2025, surveyed 1,803 C-level executives across 19 markets and 12 industries, fieldwork September to December 2024, at companies above 500 million dollars revenue in the US, Europe, Japan and Australia and above 100 million dollars elsewhere.
The fine print
The 10, 20 and 70 are a BCG rule of thumb from its own consulting case work, which BCG describes as hundreds of client engagements. No sample frame, no measurement method and no underlying data are published, so nobody outside BCG can check it. BCG states it two ways. Its case-work language says about 10 percent of AI value comes from the algorithms, 20 percent from the technology, and 70 percent from rethinking people and process. The 2025 AI Radar states it as an effort split that top performers follow, which is a claim about what leaders do rather than a measurement of where value came from. The rule is not new to that report. BCG's Sylvain Duranton was giving the same three numbers in a TED talk published in January 2020. BCG sells AI transformation work, so read the 70 percent as guidance from an interested party.
The 10-20-70 rule of AI transformation
- Matches the source
Beeontrade2025, method not recorded
What it measuredAI-driven freight operations report up to 20% lower logistics spend.
SampleNo sample published.
The fine print
Industry-blog estimate; treat as directional, phrase as 'up to'.
Leveraging AI to transform logistics in 2025 and beyond
- Matches the source
C.H. Robinson2025, method not recorded
What it measuredC.H. Robinson runs 30+ AI agents managing more than 3 million shipment tasks, including quoting, booking, appointment scheduling and tracking, trained on 37 million annual shipments.
SampleNo sample published.
The fine print
Company-reported figures via trade press.
AI in freight brokerage (via Tank Transport)
- Matches the source
Canadian Federation of Independent Business2025, method not recorded
What it measuredCanadian small businesses that use generative AI report saving 1.08 hours a day.
Sample1,683 SME owner responses
The fine print
Member survey, Canada only, self-reported time savings from adopters only. Confirm the exact report title and field dates on CFIB's research page before citing.
CFIB member survey on AI use in small business, 2025
Primary source, Canadian Federation of Independent BusinessLink to this study
- Matches the source
Capgemini2025, method not recorded
What it measured82% of insurance executives plan to integrate AI agents within three years.
SampleNo sample published.
The fine print
Capgemini finding as cited by a third-party industry blog.
AI agents in insurance (as cited by Digiqt)
- Matches the source
CarLotAI2025, method not recorded
What it measuredDealerships typically miss around 60% of inbound calls, and 85% of missed callers do not leave a voicemail; they call the next dealer.
SampleNo sample published.
The fine print
Industry-guide figures; directional, not an audited study.
AI BDC platforms guide
- Matches the source
CenterPoint Energy (via T&D World)2025, method not recorded
What it measuredCenterPoint Energy delivered a 50% year-over-year reduction in vegetation-related outage minutes in 2025 using a LiDAR-based AI digital twin.
SampleNo sample published.
The fine print
Utility-reported outcome via trade press.
LiDAR and AI reshape vegetation risk management
Primary source, CenterPoint Energy (via T&D World)Link to this study
- Matches the source
Cohere Health2025, method not recorded
What it measuredCohere Health's AI auto-approves up to 90% of prior-authorization requests for covered health-plan members.
SampleNo sample published.
The fine print
Vendor-reported; the company sells the platform it measures.
National survey on AI prior authorization
- Matches the source
Cox Automotive2025, method not recorded
What it measuredDealerships handle 12% fewer service visits than they did in 2018, with their share of all service visits down to 29% from 33%. Among owners of vehicles two years old or newer, 54% returned to their selling dealer for service in 2025, down from 72% in 2023. 45% of vehicle owners report dissatisfaction with a dealership service experience, most often over unexpected costs and poor communication.
Sample1,974 US vehicle owners aged 18 to 75 who had a vehicle serviced in the prior 12 months, fielded April to May 2025, published November 11, 2025
The fine print
US consumer survey by a company that sells dealer software and marketplaces. The 29% share is of visits, not revenue, and dealer service revenue rose over the same period. Not a Canadian measurement.
2025 Service Industry Study
- Restated wrong
CPA.com2025, vendor analysis
Usually quoted as
A 2025 CPA.com study found AI saves accounting firms 30 to 70 percent of their time.
What it measuredCPA.com and AICPA, 2025 AI in Accounting Report, June 2025, produced with the consultancy be radical. It is a 19 page trend narrative, not a survey, and it publishes no sample. Its two most quotable numbers are attributed inside the report to vendors: time savings of 30 to 70 percent on common workflow tasks, and over 80 percent automation of individual return preparation at unnamed firms.
SampleNo sample published.
The fine print
The 2025 AI in Accounting Report is not a survey. CPA.com states that it synthesizes the AICPA and CPA.com AI Symposium, ecosystem research, vendor and practitioner interviews, and firsthand observations, and it publishes no respondent count, no sampling frame and no field dates. Its figures are vendor claims by its own wording: vendors report time savings of 30 to 70 percent, and over 80 percent automation of return preparation is credited to unnamed "some firms". CPA.com is the AICPA's commercial technology subsidiary, and the report's own footnote reads "Some solution providers listed within this report are CPA.com Preferred Partner Solutions".
2025 AI in Accounting Report
- Matches the source
Deloitte2025, method not recorded
What it measured93% of AI investment goes to technology and 7% to people.
SampleNo sample published.
The fine print
Deloitte's own figure ('The investment imbalance: 93% on tech / 7% on people', Tech Trends 2026, Figure 1, released 2025-12-10). The report cites 'Deloitte research on AI investment allocation patterns, 2025' and publishes no method or sample, so use it as direction, never as a precise benchmark. Re-sourced 2026-10-15 from General Assembly's paraphrase.
Tech Trends 2026
- Matches the source
DORA (Google Cloud)2025, method not recorded
What it measured90% of technology professionals use AI at work, yet 30% have little or no trust in AI-generated code.
SampleNo sample published.
The fine print
Google Cloud research program, global survey of technology professionals, self-reported and vendor-published. Confirm respondent count and field dates in the report before citing.
DORA State of AI-assisted Software Development 2025
- Matches the source
EliseAI2025, method not recorded
What it measured85% of AI-using property operators saw measurable lead-to-lease conversion improvement; 78% report losing business to AI-enabled competitors.
SampleNo sample published.
The fine print
Vendor-published industry survey; treat as directional.
Multifamily AI operations survey
- Restated wrong
EY2025, survey
Usually quoted as
EY found that companies lose 40% of their AI gains because only 12% of employees get proper training.
What it measuredEY's 2025 Work Reimagined Survey found 12% of employees received 81 or more hours of AI training in the past 12 months. EY separately says AI can unlock up to 40% more productivity when talent foundations are strong. That is an upside ceiling from a segment comparison, not a measured 40% of gains forfeited.
Sample15,000 employees and 1,500 employers across 29 countries and 19 sectors, organizations with 1,000+ global employees, fieldwork August 2025
The fine print
EY surveyed 15,000 employees and 1,500 employers across 29 countries in August 2025. The 12% is measured and real, but it has a specific threshold behind it: 12% of employees reported 81 or more hours of AI training in the past 12 months. The 40% is not a measured loss. EY's own body text says AI can unlock up to 40% more productivity when it lands on strong talent foundations, which is a ceiling drawn from comparing survey segments, and EY sells the workforce consulting that closes the gap it describes.
AI productivity and talent strategy survey
- Matches the source
FP&A Trends Group2025, method not recorded
What it measured46% of FP&A time is spent on data collection and validation rather than analysis. Only 2% of organizations consider their FP&A teams optimized. 29% need more than 10 days to produce a forecast.
SampleNo sample published.
The fine print
Industry research group whose annual survey is vendor-sponsored. Sample size and field dates are not disclosed on the article page. Self-reported time allocation from FP&A practitioners globally.
2025 FP&A Benchmarks: Where We Are, Where Leaders Are Going
- Matches the source
Gallup (via Engageli)2025, method not recorded
What it measuredTeachers who use AI weekly save an average of 5.9 hours per week, about six weeks per school year.
SampleNo sample published.
The fine print
Gallup finding as cited by an education-platform roundup.
AI in education statistics
- Matches the source
Gartner2025, method not recorded
What it measuredMore than 40% of agentic AI projects are predicted to be canceled by end of 2027 due to escalating costs, unclear business value, or inadequate risk controls.
SampleNo sample published.
The fine print
A prediction, not an observed outcome.
Over 40% of Agentic AI Projects Will Be Canceled by End of 2027
- Restated wrong
Gartner2025, prediction
Usually quoted as
Gartner found that over 40% of agentic AI projects fail.
What it measuredOnly an estimated ~130 of the thousands of vendors claiming agentic AI are genuine; the rest is rebranded RPA, chatbots, and assistants.
SampleNo sample is published for either headline claim. The only sample in the release is a January 2025 Gartner poll of 3,412 webinar attendees, which measured investment posture: 19% significant investment, 42% conservative investment, 8% no investment, 31% waiting or unsure. Webinar attendees are self-selected. No geography, industry or job title breakdown is published.
The fine print
Neither number is a measurement. The 40% is a forecast about 2027, and Gartner published no model and no sample for it in public, because the report behind it is sold to clients. The 130 vendor figure is an analyst estimate with no published denominator and no published method, so "thousands" is Gartner's word rather than a count anyone can check. The only sample in the release is a January 2025 poll of 3,412 self-selected webinar attendees, and that poll asked how much organizations had invested in agentic AI, not whether anything was canceled.
Agentic AI vendor analysis (agent washing)
- No study behind it
Gartner2025, prediction
Usually quoted as
By 2028, 15% of all work will be done autonomously by AI agents.
What it measuredNo Gartner publication titled "AI as a coworker" exists. If a Gartner agent prediction is wanted, use: Gartner predicts that by 2028, at least 15% of day-to-day work decisions will be made autonomously through agentic AI, up from 0% in 2024, and that 33% of enterprise software applications will include agentic AI by 2028, up from less than 1% in 2024. Both come from the Top 10 Strategic Technology Trends for 2025, announced October 2024.
Sample2,986 employees, July 2025
The fine print
Gartner has published no document titled "AI as a coworker". The nearest real Gartner publications on AI in the workplace are a 24 April 2025 press release predicting that one third of finance staff will occupy "shared jobs" with AI by 2029, and an 11 November 2025 release on four scenarios for human and AI collaboration at work. The most quoted Gartner agent number, that at least 15% of day-to-day work decisions will be made autonomously through agentic AI by 2028, up from 0% in 2024, comes from the Top 10 Strategic Technology Trends for 2025, announced 21 October 2024. It is a forecast rather than a measurement, and Gartner publishes no sample or confidence interval for it. Gartner separately forecasts that more than 40% of agentic AI projects will be cancelled by the end of 2027, so quote both halves or neither.
HR survey on employee AI use
- Matches the source
Gartner2025, method not recorded
What it measuredGartner predicts AI will resolve 80% of common customer service issues without a human by 2029.
SampleNo sample published.
The fine print
A prediction, not a measurement, and copy must label it as one. Gartner's respondent base skews large enterprise. Gartner.com blocks automated fetches, verify wording via the press release or syndicated coverage.
Gartner Predicts Agentic AI Will Autonomously Resolve 80 Percent of Common Customer Service Issues Without Human Intervention by 2029
- Matches the source
Gartner2025, method not recorded
What it measured59% of finance leaders report AI use in their finance function in 2025, up from 58% in 2024 and 37% in 2023. Top uses among adopters: knowledge management 49%, AP automation 37%, error and anomaly detection 34%.
Sample183 CFOs and senior finance leaders, surveyed May to June 2025
The fine print
Gartner's respondent base skews large enterprise, not 10 to 100 person companies. Gartner.com returns 403 to automated fetch. Figures verified via CPA Practice Advisor's reprint (cpapracticeadvisor.com, 2025-11-18) and corroborated by CFO Dive and CFO.com coverage.
Gartner Survey Shows Finance AI Adoption Remains Steady in 2025
- Matches the source
GE (via UITG industry roundup)2025, method not recorded
What it measuredGE reports a 20% reduction in unplanned downtime from AI predictive maintenance; industry benchmarks for the category run 15-30% downtime reduction and 10-25% maintenance cost savings.
SampleNo sample published.
The fine print
Operator-reported outcomes compiled by an industry roundup, not an audited study.
AI predictive maintenance outcomes in manufacturing
Primary source, GE (via UITG industry roundup)Link to this study
- Matches the source
HeroHunt2025, method not recorded
What it measuredIn 2025, 41% of talent-acquisition teams piloted AI interview scheduling and 23% standardized it; 99% of hiring managers report using AI somewhere in hiring.
SampleNo sample published.
The fine print
Vendor roundup of industry surveys; treat the 99% figure as 'somewhere in the process', not full automation.
AI adoption in recruiting: 2025 year in review
- Matches the source
Hostie2025, method not recorded
What it measuredRestaurants deploying voice AI report 87% fewer missed calls and $3,000-$18,000 per month in recovered revenue per location.
SampleNo sample published.
The fine print
Vendor-reported; the company sells the voice agents it measures.
Voice AI restaurant reservations: 2025 adoption, accuracy, revenue impact
- Matches the source
HSMAI Asia2025, method not recorded
What it measured86%+ of hoteliers now depend on AI for demand forecasting; advanced revenue-management systems are associated with 10-15% RevPAR lift.
SampleNo sample published.
The fine print
Industry-association reporting; RevPAR lift is a typical range, not a guarantee.
AI revolution in hotel revenue management
- Restated wrong
Iowa State / University of Arkansas2025, field trial
Usually quoted as
Iowa State found targeted spray technology cuts herbicide use by 76%.
What it measuredTwo separate programmes, not one joint study. Iowa State Extension, 2024: a 415 acre unreplicated demonstration on 5 soybean fields with a John Deere See and Spray Ultimate. Product savings by field of 43.9, 71.2, 87.2, 87.6 and 90.6 percent, averaging 76 percent, worth $15.70 per acre. University of Arkansas, reported March 2025 from a trial begun in 2022: 43 to 59 percent reduction in postemergence herbicide at the high sensitivity setting, worth $30.49 per acre, breaking even at 819 acres.
SampleIowa State: 5 conventionally managed soybean fields using conventional or vertical tillage, 415 acres, Boone and Story counties, Iowa, 2024 season, unreplicated field-scale demonstration. Arkansas: 3 year soybean trial from 2022, Northeast Research and Extension Center, Keiser, Arkansas. Peer reviewed Arkansas paper: 4 site years, 2021 to 2022, Keiser AR and Greenville MS.
The fine print
These are two separate programmes, not one study. Iowa State ran a field scale demonstration in 2024, not 2025: five conventionally managed soybean fields, 415 acres, Boone and Story counties, using a John Deere See and Spray Ultimate. Product savings ran from 43.9 percent to 90.6 percent by field and averaged 76 percent, the fields were not replicated, and Iowa State states that savings are a function of both the technology and the underlying weed pressure, so a single average travels badly. Arkansas is a separate plot trial reporting 43 to 59 percent postemergence reduction at the high sensitivity setting, and its peer reviewed paper in Weed Technology was funded by Blue River Technology and Deere and Company, which also supplied the machine, with three Blue River employees as co-authors who disclose that employment.
See & Spray field research
Primary source, Iowa State / University of ArkansasLink to this study
- Matches the source
John Deere2025, method not recorded
What it measuredIn 2025, See & Spray ran across 5 million acres and saved more than 31 million gallons of herbicide mix, roughly a 50% average reduction.
SampleNo sample published.
The fine print
Vendor-reported acreage and savings; see the university research entry for independent measurement.
See & Spray herbicide savings
- Matches the source
Klarna2025, method not recorded
What it measuredKlarna says its AI assistant performs the work of 853 full-time agents, handling roughly two-thirds of customer-service chats.
SampleNo sample published.
The fine print
Company-reported figure via trade press.
AI assistant performance (via CX Dive)
- Restated wrong
KPMG / Melbourne Business School2025, survey
Usually quoted as
57% of workers hide their AI use from their bosses.
What it measured57% of employees who use AI at work say they have used it in non-transparent ways, meaning presenting AI-generated content as their own or avoiding revealing that they used AI. The figure counts anyone who did either behavior, including rarely.
Sample48,340 adults across 47 countries, fielded November 2024 to mid January 2025 by online panel (Dynata), nationally representative on age, gender and regional distribution, country samples of 1,001 to 1,098. 32,352 respondents were working full or part time and answered the workplace questions.
The fine print
It measured two self-reported behaviors rolled into one number. The report says 57% of employees admit using AI in non-transparent ways, meaning presenting AI-generated content as their own or avoiding revealing that they used AI. Treat the number as soft. The report prints no base line for it, and the two underlying chart rows sit at 61% and 55% for doing it rarely or more often, so 57% is a composite the report never explains. A nearby sentence scopes a related finding to employees who use AI, while the headline sentence, the executive summary and the press release all say employees, so the base is genuinely ambiguous. KPMG funded the work and sells AI advisory, and its press release restates the finding as employees who hide their use of AI and present AI-generated work as their own, which turns an either/or into a both. The research itself was designed and run independently by University of Melbourne academics, and the authors flag social desirability bias, meaning the true rate is probably higher.
Trust, attitudes and use of AI (global study)
Primary source, KPMG / Melbourne Business SchoolLink to this study
- Matches the source
LayerX2025, method not recorded
What it measuredAbout 18% of enterprise employees paste data into GenAI tools, and over half of those paste events include corporate information; 72-82% of GenAI access happens through unmanaged personal accounts.
SampleNo sample published.
The fine print
LayerX's separate 77% figure is ChatGPT's share of LLM traffic, not the share of employees leaking data. Use the 18% / over-half-of-pastes framing.
Enterprise AI & SaaS Data Security Report 2025
- Matches the source
McKinsey2025, method not recorded
What it measured88% of organizations report regular AI use in at least one business function, up from 78% a year earlier. About 6% qualify as AI high performers, the group attributing 5% or more of EBIT to AI.
SampleNo sample published.
The fine print
The 6% is McKinsey's 'AI high performer' segment, which requires both significant reported value and an EBIT impact of 5 percent or more attributed to AI (McKinsey's own wording, rechecked 2026-10-15). Do not phrase it as '6% see any return': 39% report enterprise-level EBIT impact, most of it below 5%. The URL is pinned to the 2025 edition: /the-state-of-ai now opens the 2026 survey.
The state of AI in 2025: Agents, innovation, and transformation
- Restated wrong
McKinsey2025, survey
Usually quoted as
McKinsey found employees are 3x more ready for AI than their leaders believe.
What it measured48% of employees rank training as the most important factor for AI adoption, yet nearly half receive minimal or no training.
Sample3,613 employees (managers and independent contributors) and 238 C-level executives, fieldwork October to November 2024, 81 percent United States with the remainder in Australia, India, New Zealand, Singapore and the United Kingdom
The fine print
McKinsey surveyed 3,613 employees and 238 C-level executives in October and November 2024. It did not score readiness on any scale. It measured what employees said they already do against what leaders guessed employees do, and the two groups answered different questions. Leaders estimated 4 percent of employees use gen AI for at least 30 percent of daily work, employees self-reported 13 percent, and that 3x gap is the headline. McKinsey's conclusion that employees are ready is an inference from that gap plus what employees said they want, such as the 48 percent asking for formal training. The reported findings cover US workplaces only, so quoting this as a global figure is wrong.
Superagency in the Workplace
- Matches the source
MIT NANDA2025, method not recorded
What it measured95% of enterprise GenAI pilots show no measurable P&L impact; externally purchased tools succeed roughly three times as often as internal builds.
Sample150 interviews, 350-employee survey, 300 deployments analyzed
The fine print
Measures no measurable P&L impact, not total project failure. Phrase as 'show no measurable return', never 'fail completely'.
The GenAI Divide: State of AI in Business 2025
- Matches the source
National Retail Federation and Happy Returns2025, method not recorded
What it measuredRetailers expected 15.8% of annual sales, $849.9 billion, to be returned in 2025, down from 16.9% and $890 billion in 2024. The online return rate was expected at 19.3%. 82% of shoppers call free returns a major consideration when buying, up from 76%.
Sample358 ecommerce professionals at merchants over $500 million in revenue and 2,006 consumers who returned an online purchase in the prior 12 months, fielded summer 2025, released October 15, 2025
The fine print
US retail, and the retailer sample is large merchants, so the rates are not an SMB benchmark. The dollar figure is a projection from retailer estimates, not a count. Co-published with a returns-logistics vendor owned by UPS.
2025 Retail Returns Landscape
Primary source, National Retail Federation and Happy ReturnsLink to this study
- Matches the source
Numa2025, method not recorded
What it measuredNuma serves 1,200+ dealerships; one multi-location GM group recovered over $853,000 in missed service revenue, handling roughly 16,500 calls per store per year.
SampleNo sample published.
The fine print
Vendor-reported case figures.
AI for dealership service departments
- Matches the source
OECD2025, method not recorded
What it measuredFull-time lower secondary teachers report working 41 hours a week on average across OECD countries, of which 22.7 hours are teaching, 7.4 hours preparing lessons, 4.6 hours marking and 3 hours administrative work. 36% of teachers report having used AI in their work, and 75% of those who have not say they lack the knowledge and skills to teach with it.
SampleAbout 280,000 lower secondary teachers and school leaders across 55 education systems, fielded 2024, results published October 7, 2025
The fine print
Self-reported hours for full-time lower secondary teachers, OECD average. Canada did not take part as a system, so there is no Canadian line, and the survey says nothing about training providers or higher education. The OECD site blocks automated fetches, figures verified from the OECD's own country notes.
Results from TALIS 2024 (Teaching and Learning International Survey)
- Matches the source
OpenAI2025, method not recorded
What it measuredDaily AI use reaches only about 10% of the US workforce; power users get roughly 6x the productivity of typical employees on the same tools.
SampleNo sample published.
The fine print
Vendor-published report; the company sells the tools it measures.
The State of Enterprise AI 2025
- Restated wrong
PayPal2025, survey
Usually quoted as
PayPal found that 99% of large enterprises are already adopting agentic commerce.
What it measuredNo stat was supplied, so nothing can be confirmed or corrected. A defensible figure from the primary report: 52% of large enterprises have integrated AI into regular operations, and roughly 1 in 5 merchants have 80% or more of their product catalog available as structured, machine-readable data.
Sample498 US merchants (166 small under $3M revenue, 143 mid-market $3M to $20M, 189 large enterprise over $20M), plus 7 qualitative interviews, fielded February 23 to March 3, 2026
The fine print
PayPal's Agentic Commerce Pulse is an online survey of 498 US merchants, fielded between February 23 and March 3, 2026 by Stripe Partners, an independent London research consultancy that is not Stripe the payments company. Seven interviews ran alongside it. It measures merchant awareness, readiness, trust and perceived risk. It does not measure sales, conversion or revenue. The 99% figure for large enterprises is familiarity with a term PayPal defined for respondents inside the survey, not adoption. Cite the primary PDF and not PayPal's article page, which prints the mid-market sample as 243 and makes the segments sum to 598 against a stated total of 498. PayPal commissioned the research while selling agentic commerce services.
Beyond Efficiency: Small Businesses Look to AI for Competitive Edge
- Matches the source
Phenom2025, method not recorded
What it measuredMastercard cut interview scheduling time by more than 85% and scheduled 88% of interviews within 24 hours of request.
SampleNo sample published.
The fine print
Vendor case study published by the platform provider.
Companies using AI recruiting platforms (Mastercard case)
- Restated wrong
S&P Global Market Intelligence2025, survey
Usually quoted as
42% of companies abandoned their AI initiatives in 2025, up from 17% in 2024.
What it measured42% of respondents said most of their proof of concept AI projects are abandoned before reaching production, up from 17% in the prior wave. The data was collected October 21 to November 25, 2024, not in 2025. The 17% figure was collected November 14 to December 14, 2023. S&P's exact wording is: "The proportion of companies that abandon most of their AI initiatives has increased from 17% to 42%, with the average organization scrapping 46% of its proof-of-concept projects prior to production."
Sample1,006 midlevel and senior IT and line-of-business professionals across North America and Europe, online, fielded October 21 to November 25, 2024. Margin of error plus or minus 3 points at the 95% confidence level. The 17% comparison comes from the prior wave: 1,001 respondents, fielded November 14 to December 14, 2023. The abandonment question was asked only of respondents whose organization had AI or ML in production or proof of concept, and that base size is not published.
The fine print
This measures proof of concept projects, not AI programs. Respondents estimated what share of their organization's AI projects sitting in proof of concept get abandoned before production, and 42% put that share above half. The fieldwork ran October 21 to November 25, 2024, so the 2025 in the survey name is a label and not the collection year. It gets quoted as companies walking away from AI altogether, which is not what the question asked.
Voice of the Enterprise: AI & Machine Learning 2025
Primary source, S&P Global Market IntelligenceLink to this study
- Matches the source
Salesforce2025, method not recorded
What it measuredAI and agents influenced $262B, 20% of all global online holiday sales in 2025; retailers operating their own AI agents grew 59% faster (6.2% vs 3.9% YoY).
SampleNo sample published.
The fine print
Vendor-published commerce data drawn from Salesforce's own platform traffic.
2025 holiday shopping data
- Matches the source
Salesforce2025, method not recorded
What it measured79% of service leaders say investing in AI agents is essential to stay competitive.
Sample6,500 service professionals in 40 countries
The fine print
Vendor-published by a company selling AI agents. Self-reported, global sample, not limited to mid-size companies. Confirm field dates on the report page before citing.
State of Service 2025
- Matches the source
SHRM2025, method not recorded
What it measured45% of US workers already use AI in their jobs, and 77% of those say it helps them get more done in less time.
SampleNo sample published.
The fine print
Worker-level survey, US only, self-reported productivity, not specific to the HR function. Confirm the exact report title, respondent count and field dates before citing.
SHRM 2025 research on AI use by US workers
- Matches the source
SHRM2025, method not recorded
What it measuredSHRM's 2025 benchmarking found an average cost per hire of $5,475 for non-executive roles and $35,879 for executive roles.
Sample2,371 SHRM members responding, surveyed January 9 to March 3, 2025, published October 15, 2025
The fine print
US employer averages in US dollars, skewed by large organisations, and the figures are means rather than medians. They measure in-house hiring cost, not staffing-agency fees or margins. SHRM's member base is HR professionals, so it is an employer view, not a candidate view.
SHRM 2025 Benchmarking Reports (Recruiting)
- No study behind it
Spendflo2025, method not recorded
Usually quoted as
Companies manage over 300 SaaS tools and only about half get used, according to Spendflo's 2025 SaaS and AI spend benchmarks.
What it measuredThere is no verifiable Spendflo stat to correct. If the index needs an AI spend benchmark, use SpendHound's AI Spend Report 2026: 46% of finance and procurement leaders exceeded their AI budget in 2025, against 37% who exceeded budget on traditional finance and accounting software. Sample is 172 survey respondents plus spend data from 1,300+ companies.
SampleNo sample published.
The fine print
This entry carries no verifiable statistic. Spendflo's State of SaaS Procurement 2025 is a gated ebook that discloses no sample size, no field dates and no geography, and describes its inputs only as finance leader responses plus Spendflo's own contract data. Spendflo's State of SaaS Buying Survey 2026 likewise names no sample. Neither discloses enough method to cite as a benchmark. The widely repeated figure of roughly 300 SaaS applications per company with about half of licences unused comes from Zylo's SaaS Management Index, not from Spendflo. For an AI spend benchmark with a disclosed sample, use SpendHound's AI Spend Report 2026, which reports 172 survey respondents plus spend data from more than 1,300 companies, and note that SpendHound sells spend management software.
State of SaaS Buying and Procurement 2025
- Matches the source
SQ Magazine2025, method not recorded
What it measuredTeacher AI usage roughly doubled year over year to 53-61% in the 2024-25 school year.
SampleNo sample published.
The fine print
Aggregated survey range; cite as a range.
AI in education statistics
- Matches the source
Stack Overflow2025, method not recorded
What it measured84% of developers use or plan to use AI tools, up from 76% a year earlier. 66% say their biggest frustration with AI tools is answers that are almost right. 45% say debugging AI-generated code takes more time.
SampleNo sample published.
The fine print
Self-selected online survey of the Stack Overflow community, tens of thousands of developer responses worldwide, fielded mid-2025. Skews hands-on developers, not IT decision makers, and is not company-level adoption data.
Stack Overflow Developer Survey 2025
- Restated wrong
Stanford GSB2025, field trial
Usually quoted as
AI cuts monthly financial close time by 7.5 days.
What it measuredChoi and Xie, "Human + AI in Accounting: Early Evidence from the Field", Stanford GSB Working Paper 4261 and MIT Sloan Working Paper 7280-25, May 2025, published in the Journal of Accounting Research in 2026: AI use corresponds to a 7.5-day reduction in monthly close time and a 12% increase in general ledger granularity, measured across 79 small and mid-sized firms.
Sample277 accountants surveyed, plus hundreds of thousands of transaction-level records from 79 small and mid-sized firms covering January 2023 to March 2025, United States, plus a framed field experiment run November 2024 to March 2025
The fine print
The 7.5 days is an association, not a causal estimate. It comes from field data on 79 small and mid-sized firms supplied by one AI accounting software vendor, so the sample is that vendor's own customer base and adoption was self-selected rather than assigned. The authors write that AI use "corresponds to" the reduction, and their own framed experiment found accountants sometimes accepted AI classifications that were wrong. Nothing in the study covers large enterprise closes, which is exactly where the number usually gets applied.
AI in accounting study
- No study behind it
Statista2025, method not recorded
Usually quoted as
According to Statista, most organizations report an AI skills shortage.
What it measuredNo corrected phrasing is possible. No study is named and no sample is published, so there is nothing to correct.
SampleNo sample published.
The fine print
There is no study behind this entry. The link points at the Statista homepage rather than at a statistic, and the stat string carries no number, no sample and no year of fieldwork. Statista both runs its own surveys and republishes other people's, and the source detail usually sits behind a subscription, so a Statista link on its own does not tell a reader who fielded the survey, when, or how many people answered. Name the group that ran it, or drop the entry.
Barriers to AI adoption survey
- Matches the source
The CMO Survey (Deloitte, Duke Fuqua School of Business, American Marketing Association)2025, method not recorded
What it measuredGenerative AI was used in 15.1% of marketing activities in 2025, up from 7.0% a year earlier. Marketing leaders expect AI to run 44.2% of marketing activities within three years. Marketing headcount grew 5.4% in 2024.
SampleNo sample published.
The fine print
Survey of marketing leaders at US for-profit companies, a few hundred respondents per wave. US-only, self-reported, skews larger companies. Respondent count and field dates are published per wave on the results page, confirm before citing.
The CMO Survey, 2025 edition
Primary source, The CMO Survey (Deloitte, Duke Fuqua School of Business, American Marketing Association)Link to this study
- Matches the source
University of Kansas / Togal.AI2025, method not recorded
What it measuredA 2025 University of Kansas study found Togal.AI 76% faster than leading takeoff tools; the vendor reports a full architectural takeoff in roughly 12 minutes.
SampleNo sample published.
The fine print
The 76% figure is the university study; the 12-minute and 97-98% accuracy claims are vendor-reported.
AI takeoff speed study
Primary source, University of Kansas / Togal.AILink to this study
- Matches the source
US Energy Information Administration2025, method not recorded
What it measuredUS electricity customers experienced an average of 11 hours of interruptions in 2024, nearly twice the annual average of the decade before. Major events accounted for 80% of those hours, about nine hours, while interruptions not tied to major events routinely average about two hours a year.
SampleUtility-reported SAIDI data compiled in the Electric Power Annual 2024, released December 1, 2025
The fine print
A US national average skewed by three hurricanes. The steady-state figure a utility can act on is the two hours a year of non-major-event interruptions. Canadian utilities report to provincial regulators, not to EIA, so this is a comparator, not a Canadian number.
Hurricanes in 2024 led to the most hours without power in the United States in 10 years
Primary source, US Energy Information AdministrationLink to this study
- Matches the source
Voxel512025, method not recorded
What it measuredComputer-vision quality inspection reaches 98-99% defect-detection accuracy with 50-70% inspection labor savings as industry benchmarks.
SampleNo sample published.
The fine print
Benchmark ranges across deployments, compiled by a vision-tooling vendor; treat as directional.
Visual AI in manufacturing: 2025 landscape
- Matches the source
American Transportation Research Institute2024, method not recorded
What it measuredDrivers reported being detained at a shipper or receiver in 39.3% of all stops in 2023, and in 9.9% of stops the detention ran longer than two hours. Detention cost for-hire trucking an estimated $3.6 billion in direct expenses and $11.5 billion in lost productivity, and carriers collected detention fees on fewer than 50% of the invoices they sent.
SampleDriver and motor carrier survey responses plus ATRI truck GPS data, 2023 data year, published September 10, 2024
The fine print
US for-hire trucking. The 39.3% is self-reported by drivers. The 9.9% over two hours is the figure for significant detention, never the 39.3%. Respondent counts are in the full report, not the release.
The Impacts of Truck Driver Detention
Primary source, American Transportation Research InstituteLink to this study
- Matches the source
Canadian Agricultural Human Resource Council2024, method not recorded
What it measuredMore than 28,200 agriculture jobs went unfilled during the 2022 peak season, and employers attributed a 3.7% decline in sales, about $3.5 billion, to labour shortages. Vacancies are projected to surpass 100,000 by 2030 as 85,300 workers, 30% of the workforce, retire. Even with more foreign workers, 22,200 positions are projected to remain vacant.
Sample1,048 employer responses to the 2023 CAHRC Employer Survey plus worker and stakeholder surveys, modelled by the Conference Board of Canada, published February 15, 2024
The fine print
The lost-sales figure applies an employer-reported 3.7% to total sector sales, so it is an estimate, not a count, and the 2030 numbers are a forecast. The employer survey was voluntary and over-represents some provinces. Full report verified from the CAHRC PDF.
Sowing Seeds of Change: Agriculture Labour Market Forecast 2023-2030
Primary source, Canadian Agricultural Human Resource CouncilLink to this study
- Matches the source
Gartner2024, method not recorded
What it measured30% of generative AI projects predicted abandoned after proof-of-concept by end of 2025.
SampleNo sample published.
The fine print
A prediction, not an observed outcome.
30% of Generative AI Projects Will Be Abandoned After Proof of Concept by End of 2025
- Restated wrong
RAND Corporation2024, interviews
Usually quoted as
RAND found that more than 80% of AI projects fail, twice the rate of non-AI IT projects.
What it measuredRAND cites this figure rather than measuring it, and the hedge matters. The report says: "By some estimates, more than 80 percent of AI projects fail. This is twice the already-high rate of failure in corporate information technology (IT) projects that do not involve AI." The first sentence is footnoted to a 2022 Fortune article, the second to a 2023 Harvard Business Review article. RAND's own finding is five root causes of AI project failure drawn from 65 interviews, four of which are organizational rather than technical.
Sample65 semistructured interviews, August to December 2023. 50 industry practitioners drawn from 379 people contacted, representing more than 50 organizations, plus 15 academics. All had at least 5 years building AI or ML models. Industry participants were recruited through LinkedIn Recruiter and paid a $100 honorarium for 45 minutes. Academics came from a convenience sample at conferences and the research team's own contacts. No geography published.
The fine print
RAND measured root causes, not a failure rate. The 80% figure is not RAND's own. It appears twice, once in the summary and once in the introduction, hedged both times as "by some estimates," and the footnote points to a July 2022 Fortune article that names no survey and puts the range at 83% to 92%. That article credits the range to unnamed surveys of business leaders, inside a profile of an AI vendor's CEO, so there is no study anyone can check behind the number the internet now credits to RAND. RAND's own work is 65 interviews that produced five root causes of failure and no failure rate at all. RAND calls it an exploratory analysis and warns its findings may skew toward blaming leadership, because most interviewees were engineers rather than executives.
The Root Causes of Failure for Artificial Intelligence Projects
- Matches the source
The Manufacturing Institute and Deloitte2024, method not recorded
What it measuredUS manufacturing could need as many as 3.8 million new employees between 2024 and 2033, and about 1.9 million of those roles could go unfilled if the skills and applicant gaps are not closed. 65% of manufacturers named attracting and retaining talent their primary business challenge.
SampleOnline survey of more than 200 US manufacturers plus executive interviews and labour supply and demand modelling, published April 2024
The fine print
US figures and a projection, not a count. The survey sample is small, 200-plus firms, and the 3.8 million combines growth with retirements, so do not quote it as retirements alone. No Canadian equivalent at this scale exists, CME publishes periodic labour surveys.
Taking charge: Manufacturers support growth with active workforce strategies
Primary source, The Manufacturing Institute and DeloitteLink to this study
- Restated wrong
US Federal Trade Commission2024, enforcement action
Usually quoted as
In 2024 the FTC fined five AI companies for overstating what their AI could do.
What it measuredThe FTC launched a coordinated crackdown on deceptive AI claims with five simultaneous enforcement actions, followed by roughly a dozen AI-washing cases in 2025.
Sample5 enforcement actions against 5 operations, announced 25 September 2024, United States only. No survey, no respondents, no sample. One of the five, FBA Machine, was filed in June 2024 and folded into the September announcement.
The fine print
This is a law enforcement sweep, not research. It measures nothing about how often AI claims are exaggerated, and it has no sample. The FTC characterised three of the five targets as business opportunity schemes that marketed AI-powered tools, so only DoNotPay and Rytr sold an actual AI product. No civil penalty was levied in the sweep. DoNotPay agreed to pay 193,000 dollars in consumer redress, which is the only money that changed hands. The FTC then reopened and set aside the Rytr order on 22 December 2025, so anyone citing Rytr as live precedent is citing a vacated order.
Operation AI Comply
Primary source, US Federal Trade CommissionLink to this study
Method.
How each study was checked, what a verdict means, and the interest we have in the result.
Every study Advizr cites anywhere on this site is recorded in a typed registry with the source address, the finding phrased to what was measured, the sample where one is published, and an honesty note. A reference to a study that is not in the registry fails the build, so the site cannot cite something this index does not list.
Each entry was researched against its primary document, then handed to a second reviewer told to refute the first and to reject when unsure. On the most recent pass that reviewer rejected 16 of 21 first drafts, almost always for over-claiming about a named firm's method. The corrected version is what appears here.
The verdict is about the number as it circulates. Restated wrong means the review recorded a common restatement the study does not support, or the finding as it had been written did not match the source. No study behind it means the citation leads to a page, a vendor post or a forecast with no study under it. The rest match their source.
The unit is one published study, report or enforcement action. Entries are chosen because they come up in the enterprise AI debate often enough to be worth correcting, not by a sampling frame, so this is a curated index, not a representative sample of the literature. A survey and a prediction can produce similar looking percentages, so the method is printed on every entry.
Advizr sells AI systems. That is a commercial interest in this subject, and it is why the notes are written to be usable against Advizr as readily as for it. Several entries undercut claims the company would find convenient. Corrections are welcome and versioned rather than made in silence.
The data, and how to cite it.
The data is CC BY 4.0, 93 rows and 14 columns. The prose is not.
Booth, James. The AI Evidence Index, version 1.0. Advizr, 2026. 93 studies, 2024 to 2026. CC BY 4.0. advizr.ca/research/ai-evidence-index
Run your whole company on AI.
Tell us how your business runs. If we're not the right fit, we'll say so on the call.
Your custom strategy and systems, delivered in under a month. Your named people using them every week by day 30. Or we keep working for free until both are true.
The AI operating layer for enterprises.