Stat check: numbers you should stop quoting
Some famous statistics have no real source, are out of date, or are quoted wrong. Here is where each one came from and, when we have one, the checked number to use instead.
Myths by topic
Numbers with no real source
Each card names the claim, says what is wrong with it and where it came from.
Stanford's AI Index found that 88% of organizations use AI
The 88% is from McKinsey's global survey (Nov 2025, 1,993 respondents, many at large firms). The AI Index repeats it. It is a share of survey respondents, not of all businesses; Census puts US employer firms at 24.6%. Where it came from
Quote this instead: 24.6%78% of US businesses now use AI
78% comes from an NBER survey of executives in seven countries, and separately from the Atlanta Fed estimate that 78% of the labor force works at AI-adopting firms (a share of workers). The Census share of US firms is 24.6%. Where it came from
Quote this instead: 24.6%Companies that adopt AI are cutting jobs
Common in headlines about single companies. In the Census AI supplement, 95.7% of AI-using firms said AI did not change total employment; 2.0% said it decreased. Where it came from
Quote this instead: 95.7%Only about 5% of US businesses use AI
Census numbers from 2023-2024 (3.7% to 6%) used the old question about AI "in producing goods or services". Census widened the question in November 2025. Where it came from
Quote this instead: 24.6%AI skills earn a 56% wage premium
From PwC's 2025 AI Jobs Barometer. The 2026 edition gives 62% and restates last year as 57%. Where it came from
Quote this instead: 62%AI made more images in a year than photography made in 149 years
Spread by news sites from Everypixel's 2023 estimate. Everypixel corrected the comparison; its page now says photographers took "a few dozen years" to reach 15 billion. The 15 billion figure itself is an extrapolated 2023 estimate. Where it came from
AI made as many images in 1.5 years as photography made in its first 149 years
Everypixel's August 2023 blog estimate. Everypixel later published a correction saying it made a calculation error in the photography comparison; the page now says photographers took 'a few dozen years'. The 15 billion AI images figure is itself a 2023 vendor estimate. Where it came from
AI search and zero-click
60% of Google searches are zero-click
A rounding of SparkToro's 2024 US figure (58.5%, Datos panel, Sep 2022-May 2024), repeated with no year, region or panel. The current US figure is 68.01% (Jan-Apr 2026, a different panel). Where it came from
Quote this instead: 68.01%65% of US Google searches end without a click (64.82%)
64.82% is a worldwide figure for all of 2020 (Similarweb via SparkToro, published March 2021). The US figure in the same post was 68.3% for Q4 2020. SparkToro now flags the post as possibly out of date. Where it came from
Quote this instead: 68.01%SEO is dead
A recurring headline when Google changes results. Even SparkToro, whose data shows 68.01% zero-click searches, writes that SEO 'still matters as much or more than ever before', but earns less traffic. Branded, local and high-intent searches still send clicks. Where it came from
Quote this instead: 23193% of online experiences begin with a search engine
From a 2006 report. Ahrefs found 700+ sites linking to a page for this stat that no longer even mentioned it. Where it came from
Quote this instead: 42%AI Overviews now appear on almost every Google search
Early 2024 reports claimed AI Overviews on up to 84% of queries (noted by SparkToro). Pew's tracked browsing found an AI summary on 18% of Google searches in March 2025; SparkToro's 2026 post says 20%+. Where it came from
Quote this instead: 18%Marketing and advertising
It takes 7 touches (the 'rule of 7') before someone buys
Usually credited to 1930s film studios and movie posters. No study, memo or data has been traced; the number is repeated from blog to blog. Research on repeated exposure exists, but none sets a threshold of seven. Where it came from
Email marketing returns $42 for every $1 spent
The figure is from the UK Data & Marketing Association's Marketer Email Tracker 2019: just over 42 pounds per pound, estimated by UK marketers in a 2019 survey. It was never a US dollar figure and is now seven years old. Where it came from
Quote this instead: 35%Email returns $36 for every $1 spent
Litmus reports ROI in bands, not an average. In its 2025 State of Email data, 35% of marketing leaders said $10-$36 and 30% said $36-$50; 21% do not measure email ROI at all. $36 is the edge of a band. Where it came from
Quote this instead: 35%Our attention span is 8 seconds, shorter than a goldfish
A 2015 Microsoft Canada report credited 'Statistic Brain'; the bodies it named had no such research. Where it came from
57% of the buying process is done before talking to sales
A 2011 CEB and Google study of B2B buyers at 22 companies; an average that hides wide variation, also quoted as 65% and 67%. Where it came from
88% of people trust online reviews as much as personal recommendations
A 2014 BrightLocal survey headline. The article body reports that 88% had read reviews to judge a local business, which measures reading, not trust. The 'trust' figure in the 2013 survey headline was 79%. Where it came from
Web and software development
GitHub Copilot makes developers 55% more productive
A 2022 controlled test (arXiv 2302.06590): 95 Upwork freelancers recruited, one task (an HTTP server in JavaScript); the Copilot group finished 55.8% faster (95% CI 21% to 89%). It measured one task's speed, not overall productivity. Where it came from
Quote this instead: 55.8%AI makes experienced developers 19% slower
METR's early-2025 trial of 16 open-source developers. METR now labels it out of date; its February 2026 update points to a likely speed-up but says the new data is unreliable because developers refused to work without AI. Where it came from
Quote this instead: 19%Only 29% of developers trust AI tools
Repeated by roundup sites for the 2025 Stack Overflow survey. The survey page says 33% trust AI accuracy and 46% distrust it; in 2026 the question changed and 48% trust AI output when they can easily verify it. Where it came from
Quote this instead: 48%53% of mobile visitors leave a site that takes longer than 3 seconds to load
Google/DoubleClick 'The Need for Mobile Speed' (2016). The Think with Google page that hosted it now redirects elsewhere; use current Core Web Vitals data instead. Where it came from
Quote this instead: 48%SaaS
The average company uses 305 SaaS apps (Zylo 2026)
Roundup sites attribute 305 apps to Zylo's 2026 SaaS Management Index. Zylo's own 2026 release gives no such count (it says app counts were largely flat); its 2025 release gave 152 apps for smaller companies and 660 for 10,000+ employees. Where it came from
Quote this instead: $9,455SaaS spend per employee doubled in a year, from $4,830 to $9,455
Zylo's 2025 release gave an average of $4,830 per employee; its 2026 release gives a median of $9,455 on a larger spend base ($75B vs $40B). Average and median from different datasets cannot be compared as growth. Where it came from
Quote this instead: $9,455Companies waste 30% of their cloud spend
Flexera State of the Cloud figures are respondents' own estimates of waste, not measured waste; the 2026 report gives 29%. Where it came from
Quote this instead: 29%E-commerce
95% of all purchases will be made online by 2040
Credited to Nasdaq (dated 2017 or 2019 depending on the site). No Nasdaq report or method has been found; roundup sites repeat each other. The official US online share is 17.1% (Q2 2026). Where it came from
Quote this instead: 17.1%US holiday sales topped $1 trillion for the first time in 2025
NRF's October 2025 forecast said 'just over $1 trillion' and its January 2026 Retail Monitor read was +4.1%. NRF's October 2026 release puts 2025 holiday sales at $972 billion, up 3.1%. Where it came from
Quote this instead: $972BMost shopping in the US now happens online
Usually a misread of mobile share of online orders (56.4% of online holiday transactions were on smartphones, Adobe). Online sales are 17.1% of all US retail sales (Census, Q2 2026). Where it came from
Quote this instead: 17.1%Cyber security
60% of small businesses close within six months of a cyber attack
Traced by Nextgov (May 2017) to an uncited line in a 2011 Business Insider article, later repeated in an NCSA infographic that NCSA took down; NCSA and Symantec both said it was not their research. Where it came from
Quote this instead: 42%AI-powered business email compromise caused $2.77 billion in losses in 2024
Roundup sites relabel the FBI's total BEC losses for 2024 ($2,770,151,146 across 21,442 complaints, all BEC) as 'AI-powered'. IC3's AI-related BEC losses for 2025 were about $30 million. Where it came from
Quote this instead: $893MHackers attack every 39 seconds
A 2007 University of Maryland study of brute-force attempts on four deliberately weak Linux test computers (2,244 attempts a day). It does not measure attacks on businesses or people. Where it came from
Quote this instead: 1,008,597The average data breach costs $4.45 million (or $4.44 million)
IBM Cost of a Data Breach 2023 ($4.45M) and 2025 ($4.44M) figures still quoted as current; the 2026 report gives $4.99M globally and $11.5M in the US. Where it came from
Quote this instead: $4.99MSocial media and internet
Most Americans use TikTok
True only for young adults (65% of 18 to 29 year olds). Among all US adults it is 37% in 2026, behind YouTube (84%), Facebook (70%) and Instagram (51%). Where it came from
Quote this instead: 37%Teens have left YouTube for TikTok
TikTok is the platform teens are most often on 'almost constantly' (21%), but YouTube is still used by 92% of teens and visited daily by 76%, against 68% and 61% for TikTok. Where it came from
Quote this instead: 92%Young people don't use Facebook any more
Among US adults 18 to 29, 69% use Facebook (Pew 2026). It is teens 13 to 17 where Facebook fell, from 71% in 2014-15 to 31% in 2025. Where it came from
Quote this instead: 70%Small business
Half of small businesses fail in their first year
A mix-up of the five-year figure. BLS data show about half of new establishments close within five years, not one. In the first year about 78% survive. Where it came from
Quote this instead: 77.9%Only about 8% of small businesses use AI
The SBA Advocacy FAQ (Feb 2026) quotes 8.2% for firms under five employees, from the older Census BTOS question over Sept 2024 to Aug 2025. Census changed the question in Nov 2025; firms with 1 to 4 staff are now at 24.3%. Where it came from
Quote this instead: 24.3%Nearly half of small businesses do not have a website
A 2016 Clutch survey found nearly 50% of small businesses had no website. Clutch's own 2025 survey puts website ownership at 83% (vendor survey of 406 owners). Where it came from
Quote this instead: 83%Healthcare
86% of doctors use telehealth
The 2021 NCHS figure (86.5%). The same survey found 80.0% in 2024, published June 2026. Where it came from
Quote this instead: 80.0%81% of doctors use AI
AMA's 2026 headline is 'awareness or use'. It adds 9% of physicians who were unsure which AI tools their practice offers to the 72% who named at least one AI use. Where it came from
Quote this instead: 72%Real estate
The typical first-time home buyer is in their late 20s
True in the 1980s, says NAR. The median first-time buyer was 38 in the 2024 Profile and 40 in the 2025 Profile. Where it came from
Quote this instead: 40Since buyers search online, most now buy homes without an agent
A common claim after online listing sites and the 2024 commission rule changes. NAR's 2025 Profile: 52% found their home online, yet 88% bought through an agent. Where it came from
Quote this instead: 88%Law firm marketing
93% of law firms use AI.
Clio's March 2025 press release for its Legal Trends for Mid-Sized Law Firms report. The headline says mid-sized firms; the body says 'mid-sized and solo'. Roundup sites drop the size band. Where it came from
Quote this instead: 71%79% of lawyers use AI every day.
Clio's 2024 Legal Trends Report. Clio's own page says 79% of surveyed legal professionals use AI 'in some capacity' and only 25% widely or universally; 'daily' is not in the finding. Where it came from
Quote this instead: 71%AI will shrink the number of lawyer jobs.
Common claim in AI and legal-tech commentary. BLS projects lawyer jobs to grow 5% from 2025 to 2035 and says efficiency gains are not expected to reduce overall demand. Where it came from
Quote this instead: 5%Dental marketing
Dentistry is already about 35% consolidated into DSOs.
An unnamed advisory firm's claim, published without saying which consolidation measure it used; DentistryIQ checked it against ADA data. ADA HPI counts 16.1% of dentists as DSO-affiliated in 2024. Where it came from
Quote this instead: 16.1%Most dentists no longer own their practice.
Repeated in DSO and consolidation coverage. ADA HPI: 72.5% of US dentists were practice owners in 2023 (down from 84.7% in 2005). Where it came from
Quote this instead: 72.5%Home services marketing
More than 70% of home service pros use AI.
Contractor Magazine headline (3 Jun 2025) on Housecall Pro's 2025 survey: over 70% had tried AI tools, about 40% used AI actively. Housecall Pro's 2026 survey finds 48% actively use AI; ServiceTitan's April 2026 survey finds about 25% of residential contractors do. Where it came from
Quote this instead: 48%The US home services market is worth $600 billion.
Agency and roundup pages give conflicting market sizes ($226B, about $600B, $650-750B) without a primary source or a shared definition. Where it came from
AI agents
79% of companies have adopted AI agents
PwC asked 300 US senior executives in May 2025 whether AI agents are "being adopted" in their companies, which counts agent features in everyday apps. Gartner's 2026 CIO survey puts deployed agents at 17%. Where it came from
Quote this instead: 62%40% of enterprise apps already have AI agents
A Gartner prediction (Aug 2025) that 40% of enterprise apps will include task-specific agents by the end of 2026, up from less than 5% in 2025. It is a forecast, not a measurement. Where it came from
Over 40% of AI agent projects have failed
Gartner predicted (Jun 2025) that over 40% of agentic AI projects will be canceled by the end of 2027. It is a prediction, not a count of failures. Where it came from
Reconciled
Same question, different answers
The different numbers you see quoted are often all real: they come from different panels, years and questions. Here they are side by side.
What share of US businesses use AI?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 24.6% | 5-18 Oct 2026 | US Census Bureau, BTOS | Every US employer firm counts once, and most are small. Asks about AI in a business function in the last two weeks. |
| 32% of jobs | Nov 2025 - Jan 2026 | Census working paper CES-WP-26-25 | Same Census survey weighted by employment: 18% of firms, but they hold 32% of jobs. |
| 46% | Sep - Nov 2025 | Federal Reserve Banks, Small Business Credit Survey | Small employer firms (1-499 staff), convenience sample; counts use by the business or any employee. |
| 78% | Summer 2026 | NBER w34836 | Nearly 9,000 senior executives in seven countries; bigger firms weigh more. |
| 78% of workers | Nov 2025 | Atlanta Fed Survey of Business Uncertainty (via Fed note) | Share of the labor force at AI-adopting firms, not a share of firms. |
| 88% | Jun - Jul 2025 | McKinsey global survey | Share of 1,993 respondents worldwide; 38% work at $1B+ firms. |
| 20.2% | 2025 | OECD | Firms in OECD countries with data, mostly 10+ staff; not US only. |
| 19.95% | 2025 | Eurostat | EU enterprises with 10+ staff; not US. |
How many US workers use generative AI at work?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| One in four | Oct 2025 | Federal Reserve, SHED | Used it at work in the prior month. |
| About 41% | Nov 2025 | Real-Time Population Survey (via Fed note) | Share of the workforce using GenAI for work; a broader question. |
| 38% | Feb 2026 | Pew Research Center | Employed adults who use chatbots for tasks at work, at any time. |
| 20.8% of firms | Nov 2025 - Feb 2026 | Census BTOS AI supplement | Share of firms whose employees used generative AI in the last six months; counts firms, not workers. |
What share of companies use AI agents?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 62% | Jul - Aug 2026 | KPMG AI Quarterly Pulse | 314 US leaders at $1B+ firms; counts building, deploying or developing. |
| 79% | May 2025 | PwC AI agent survey | 300 US senior executives; "being adopted" includes agent features in apps. |
| 23% scaling, 62% at least experimenting | Jun - Jul 2025 | McKinsey | Global respondents; scaling means expanding in at least one function. |
| 40% vs 22% | May - Jun 2026 | McKinsey | Scaling agents: respondents at $1B+ organizations vs smaller ones. |
| 17% | 2026 survey | Gartner CIO and Technology Executive Survey | Organizations that have deployed AI agents; a stricter test. |
What share of Google searches end with no click?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 49% | 2019 | SparkToro / Jumpshot panel | First SparkToro study; a now-defunct panel, US. Not comparable with later panels. |
| 64.82% | 2020 | SparkToro / Similarweb | Worldwide, not US, desktop and mobile; often misquoted as US. Same post gives US Q4 2020 as 68.3%. |
| 58.5% | Sep 2022-May 2024 | SparkToro / Datos panel | US; few iPhones in the panel; weighted 63% mobile. |
| 60.45% | 2024 (restated) | SparkToro, 2026 post | The 2026 post restates the 2024 US figure as 60.45%; it does not explain the change from 58.5%. |
| 68.01% | Jan-Apr 2026 | SparkToro / Similarweb panel | US, browser searches on desktop and mobile; Google app excluded; SparkToro calls the comparison with earlier years 'a bit of apples and oranges'. |
| 62.1% to 69.5% | Jan-Apr 2026 | SparkToro / Similarweb, six countries | Germany lowest, UK highest; same panel and period, so these can be compared with each other. |
How much do AI Overviews cut clicks?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 15% to 8% | Mar 2025 | Pew Research Center, 900 US adults, 68,879 searches | Share of search visits with a click on a normal result, without vs with an AI summary. Real tracked browsing; Google only; one month. |
| 1% | Mar 2025 | Pew Research Center | Visits where a link inside the AI summary was clicked. |
| 16% to 26% | Mar 2025 | Pew Research Center | Share of results pages after which the browsing session ended, without vs with a summary. |
| -34.5% | Mar 2024 vs Mar 2025 | Ahrefs, 300,000 keywords (vendor) | Click rate of the number one result, desktop, vs a forecast without AI Overviews. |
| -58% | Dec 2023 vs Dec 2025 | Ahrefs update (vendor) | Same method, two-year gap and new baseline, so not a trend from -34.5%. Positions 2-10 also lose (-50.8% to -19.4%). |
| nearly 60% | 2026 | SparkToro (citing Ahrefs) | Restates the Ahrefs figure; not a separate measurement. |
How much did US holiday sales grow in 2025?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| +4.1% | Nov-Dec 2025, read 12 Jan 2026 | NRF, CNBC/NRF Retail Monitor | Early read from card transaction data, before Census December data |
| +3.1% ($972B) | Nov-Dec 2025, stated 15 Oct 2026 | NRF | Based on Census data; core retail without cars, gas and restaurants |
| +6.8% ($257.8B) | Nov-Dec 2025 | Adobe Analytics (vendor) | Online spending only, from retailer site data |
How many Americans use buy now, pay later?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 21.2% | 2022 | CFPB | Consumers with a credit record, from six lenders' loan data matched to credit files |
| 16% | Year to Oct 2025 | Federal Reserve Board, SHED | Survey of US adults, self-reported use in the past year |
| $20B | Nov-Dec 2025 | Adobe Analytics (vendor) | Dollar value of online holiday spending paid with BNPL, not a share of people |
How big is US digital video advertising?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| $78B | 2025 | IAB/PwC Internet Advertising Revenue Report | Revenue reported by ad sellers; includes CTV, social video, online video and short-form video |
| over $80B | 2026 projection | IAB with Guideline and Advertiser Perceptions | Agency billing data plus a buyer survey; growth of 11% implies a 2025 base below $78B, so the two methods differ |
Does AI make developers faster? Why studies disagree
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 55.8% faster | 2022 test, published 2023 | GitHub Copilot experiment (arXiv 2302.06590) | 95 Upwork freelancers, one new JavaScript task, early Copilot. |
| 19% slower | Feb to Jun 2025 | METR randomized trial | 16 experienced developers on their own large repositories, early-2025 tools; METR now calls it out of date. |
| -18% for returning developers | Aug 2025 to early 2026 | METR update, Feb 2026 | Late-2025 tools; METR reads it as some evidence of speed-up but says selection effects make it weak evidence. |
| 73% daily use | Jun to Aug 2026 | Stack Overflow Developer Survey 2026 | Self-reported use, not measured speed. |
Do developers trust AI? 2025 vs 2026 Stack Overflow numbers
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 33% trust, 46% distrust | 2025 | Stack Overflow Developer Survey 2025 | Five-point trust scale on AI accuracy. |
| 48% trust when easy to verify; 6.6% for important decisions | 2026 | Stack Overflow Developer Survey 2026 | New answer options about when people trust AI; not comparable as a trend. |
How much did Americans lose to online fraud and cybercrime in 2025?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| $20.877B | 2025 | FBI IC3 | All complaints to IC3, including businesses (BEC, ransomware). |
| $15.9B | 2025 | FTC Consumer Sentinel | Consumer fraud reports to the FTC; some victims report to both, so do not add them. |
What does a data breach cost?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| $11.5M | Mar 2025 to Feb 2026 | IBM / Ponemon, US average | Breaches of 2,590 to 115,380 records at 602 organizations; includes detection, lost business, response. |
| $4.99M | Mar 2025 to Feb 2026 | IBM / Ponemon, global average | Same study, all 16 countries and regions. |
| $20,699 | 2025 | FBI IC3, average loss per complaint | Money lost by complainants, mostly individuals; not breach response costs. |
SaaS spend per employee: $4,830 or $9,455?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| $4,830 | 2024 data, published Jan 2025 | Zylo SaaS Management Index 2025 | Average; $40B spend under management. |
| $9,455 | 2025 data, published Jan 2026 | Zylo SaaS Management Index 2026 | Median; $75B spend under management. Different measure and base. |
How many small businesses are there in the US?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 36.2M | 2022 data (FAQ Feb 2026) | SBA Office of Advocacy | All firms under 500 employees, including one-person firms with no payroll |
| 6.4M | 2022 | SBA Advocacy from Census Statistics of US Businesses | Only small firms with paid employees (6,395,635) |
| 29.8M | 2022 | SBA Advocacy from Census Nonemployer Statistics | Only firms with no employees (29,811,495), mostly sole proprietors |
| 4.34M | 2025 state profiles | SBA Office of Advocacy | California alone, the state with the most |
What share of small businesses use AI?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 8.2% | Sept 2024 to Aug 2025 | SBA Advocacy FAQ from Census BTOS | Firms under five employees; older BTOS question (AI used to produce goods or services) |
| 24.3% | Wave 5-18 Oct 2026 | Census BTOS | Firms with 1 to 4 employees; current question (AI in any business function, last two weeks) |
| 17.9% | 17 Nov 2025 to 8 Feb 2026 | Census BTOS AI supplement | All employer firms, pooled winter waves; not comparable with the biweekly series |
| 46% | Sept to Nov 2025 | Federal Reserve Small Business Credit Survey | Employer firms under 500 staff; counts any use by the business or its employees; convenience sample of 6,525 |
How many US doctors use AI?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 72% | Jan to Feb 2026 | AMA physician survey | Named at least one AI use case in their practice |
| 81% | Jan to Feb 2026 | AMA physician survey (headline) | Adds 9% who were unsure which AI tools their practice offers |
| 38% | 2023 | AMA physician survey | Earlier wave without the 'unsure' option; AMA warns waves are not strictly comparable |
| 71% | 2024 | ASTP/ONC from AHA IT Supplement | Hospitals, not doctors: predictive AI built into the EHR |
What share of US home buyers are first-time buyers?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 21% | Jul 2024 to Jun 2025 | NAR Profile of Home Buyers and Sellers | Annual mail survey of 6,103 primary-residence buyers |
| 24% | Jul 2023 to Jun 2024 | NAR Profile of Home Buyers and Sellers | Previous year of the same survey |
| 31% | Sept 2026 | NAR REALTORS Confidence Index | Monthly survey of agents about closed sales; different respondents and period |
How many lawyers use AI: 30%, 71% or 93%?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 71% / 75% | 2026 | Clio, solo and small firm report | Solo practitioners / small firms using AI to complete legal work; vendor survey. |
| 93% | 2025 (2024 survey) | Clio, mid-sized firm report | Mid-sized firms only; any AI use, with over half using it widely. |
| 30.2% | Oct-Dec 2024 | ABA Legal Technology Survey (Check first) | Asks whether the attorney's office uses AI-based tools; 512 respondents. Not yet verified on americanbar.org. |
What share of home service contractors use AI: 25% or 48%?
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 25% | 2026 (released Apr 2026) | ServiceTitan / Thrive Analytics | 1,000 residential contractor owners and managers; 'currently using AI'. |
| 48% | May-Jun 2026 | Housecall Pro | 248 home service pros, mostly owners of small firms on its platform; 'actively use AI in their business'. |
Every checked number links to its original source. Free to cite with a link to the stat page.