AI-first company

Claim check

“Every $1 invested in generative AI returns $3.70 (IDC)”

The figure is in a 2024 IDC report paid for by Microsoft, as “3.7x”. It is the average of what decision-makers at organizations already using generative AI estimated, when asked for their return on each $1 spent. The survey did not define return, and nobody checked the answers against accounts. In the 2025 round of the same survey, the average estimate fell to 2.8 times.

By
Last reviewed
Corrections
Report an error
Verdict Needs correcting

The source exists, but the claim gets the year, the population or the wording wrong.

Kind of figure Self-reported

What people said in a survey or interview, not measured.

Holds for What decision-makers at organizations already using generative AI estimated in August 2024, leaving out those who were not sure, with return left undefined. It is not a measured return, and the 2025 round of the survey gave 2.8 times.

The claim

For every $1 a company invests in generative AI, it gets $3.70 back, according to an IDC study commissioned by Microsoft.

How the claim travelled

Each step is a publication that passed the claim on, oldest first, with its exact words and what changed.

  1. Research

    IDC (Ritu Jyoti and Dave Schubmehl) · archived copy

    “For every $1 an organization invests in GenAI, they are realizing an average of 3.7x return. What would you estimate your organization’s ROI is for every $1 spent on generative AI projects or initiatives?”

    What changed: Origin. IDC asked decision-makers at organizations using generative AI to estimate their return on each $1, in answer bands from no return to “5x+”. 3,343 gave an answer; “not sure” was left out. The heading above the question already states the estimate as “realizing” a return. Return is not defined.

  2. Sponsor’s summary

    Microsoft (Alysa Taylor) · archived copy

    “For every $1 a company invests in generative AI, the ROI is $3.7x.”

    What changed: The sponsor’s blog drops the survey question, so the estimate reads as a result. “3.7x” gains a dollar sign: “$3.7x”. The blog cites the report by its IDC number but does not link it, and Microsoft’s landing page gives no report, date or method.

  3. Press

    ITPro (Nicole Kobie) · archived copy

    “Generative AI offers a return on investment (ROI) of $3.70 for every dollar invested, according to a report commissioned by Microsoft from IDC.”

    What changed: “$3.7x” becomes “$3.70”, a sum of money returned. Generative AI in general now “offers” it, not the organizations that answered. The attribution to IDC and Microsoft stays.

  4. Blog

    Radiant Institute (Maverick Foo) · archived copy

    “Organizations investing in GenAI see an average ROI of 3.7x per $1 spent. This means that every dollar invested in generative AI yields almost four times its value back”

    What changed: The title turns 3.7 times into “370% ROI”. The text reads it as money back, “almost four times its value”. If 3.7 times means money back, the gain is 270%, not 370%. In April 2025, Amity Solutions, an AI vendor, credited the $3.70 to “the Radiant Institute” instead of IDC.

  5. Research

    IDC (David Schubmehl and Kathy Lange) · archived copy

    “The lower ROI reported in this year’s survey — 2.8 times for organizations using GenAI compared to 3.7 times in the 2024 survey — most likely reflects the evolving maturity and complexity of GenAI deployments.”

    What changed: Follow-up. The 2025 round of the same survey, again paid for by Microsoft, gives 2.8 times from 2,688 answers. It reprints the 2024 average with a different base, 3,383 rather than 3,343, and calls 2024 a survey about “traditional AI”, though the 2024 question asked about generative AI.

  6. Vendor

    AmplifAI · archived copy

    “For every $1 invested in generative AI, companies see an average return of $3.70.”

    What changed: The current wording, on a statistics page by a seller of contact-centre AI software. It credits Microsoft’s landing page, not the report. “Companies” in general now “see” the return. The page does not mention the 2025 figure of 2.8 times, published eleven months earlier.

What the primary source says

The report is 2024 Business Opportunity of AI: Generative AI Delivering New Business Value and Increasing ROI, a 47-page IDC InfoBrief published in November 2024 (IDC #US52699124). IDC is a market research firm. Its custom research unit produced the report for Microsoft, which commissioned and paid for it. Microsoft sells generative AI, including Copilot and Azure OpenAI. The report says that its publication does not mean IDC endorses the sponsor’s products. Microsoft’s pages cite the report but do not link it; a public copy sits on a third-party file store.

How it was done. In August 2024, IDC surveyed “over 4,000 business leaders and decision-makers from around the world who are responsible for bringing AI transformation to life within their organizations”. The report does not list the countries, say how the respondents were found or give a response rate. Respondents whose organizations used generative AI were asked: “What would you estimate your organization’s ROI is for every $1 spent on generative AI projects or initiatives?” They picked one answer: no return, 1x, 2x, 3x, 4x, 5x or “5x+”. 3,343 gave an answer; those who said “not sure” were left out. The survey did not define ROI (return on investment). It did not say which costs count, over what period, or whether 3x means $3 back in total or $3 on top of the $1.

What it found. 31% picked 3x, 24% picked 2x and 24% picked 4x. 6% picked “5x+”, and 1% said there was no return. The average was 3.6x to 3.8x in the four regions, and from 3.2x in education to 4.2x in financial services.

Where the 3.7 appears. The key findings say: “For every $1 a company invests in GenAI, the ROI is 3.7x across industries and similar across regions.” The chart of answers is headed: “For every $1 an organization invests in GenAI, they are realizing an average of 3.7x return.” The survey question sits directly below that heading.

What is wrong with the claim

  • It is an estimate, not a measured return. Each respondent guessed the return for their own organization. Nobody checked the answers against accounts.
  • “Return” has no definition. “$3.70 back” reads 3.7x as money returned. If that is the meaning, the gain is $2.70 on each $1. The Radiant Institute’s “370% ROI” reads it the other way. The survey does not say which is right.
  • It covers only organizations already using generative AI, and only answers with a number. The people who answered were responsible for AI at their organizations, so they were judging their own programmes. “Every $1 invested” stretches their estimates to all spending on generative AI.
  • The report does not say how the average was calculated. On the published answers, the average comes to about 3.3 if “5x+” counts as 5. Reaching 3.7 needs the 6% who picked “5x+” to average about 12 times. Without IDC’s method, the 3.7 cannot be checked.
  • It is out of date. In the 2025 round of the same survey, the average estimate was 2.8 times. IDC wrote that the fall “most likely reflects the evolving maturity and complexity of GenAI deployments”.
  • The “10.3x” for top leaders, often repeated with it, is circular. IDC classed organizations as leaders partly on “Higher ROI from AI initiatives”, and top leaders are those “who realize more than 5x return”. Their high average follows from how they were picked.

What other sources found

This check found no source that measured the return on each dollar spent on generative AI from companies’ accounts. A larger survey run by researchers with no AI product to sell points the other way. Between November 2025 and January 2026, central banks and a university asked nearly 6,000 executives in the United States, the United Kingdom, Germany and Australia about AI at their firms. “89% report no impact on labor productivity (sales per employee)” over the past three years (Firm Data on AI). In a second round, run between May and July 2026, that share fell to 83%. That is also what executives said, not a measurement.

The IDC survey can be cited for what decision-makers at organizations using generative AI estimated, in a survey paid for by Microsoft: 3.7 times in 2024 and 2.8 times in 2025. It is not a source for what generative AI returns.

What to cite instead

In an August 2024 survey for Microsoft, which sells generative AI, IDC asked business leaders responsible for AI at their organizations to estimate the return on each $1 their organization spent on generative AI. The 3,343 who used generative AI and gave an answer estimated 3.7 times on average; the survey did not say what counted as return. In the 2025 round of the same survey, the average estimate was 2.8 times.

Sources

  1. Jyoti and Schubmehl (November 2024). 2024 Business Opportunity of AI: Generative AI Delivering New Business Value and Increasing ROI. IDC InfoBrief sponsored by Microsoft, IDC #US52699124. Public copy on a third-party file store. 143485449.fs1.hubspotusercontent-eu1.net
  2. Taylor (12 November 2024). IDC’s 2024 AI opportunity study: Top five AI trends to watch. Official Microsoft Blog. blogs.microsoft.com
  3. Microsoft. Business Opportunity of AI. Landing page, no date. info.microsoft.com
  4. Kobie (18 November 2024). AI is finally delivering bang for its buck, according to Microsoft. ITPro. itpro.com
  5. Foo (30 November 2024). 370% ROI on Generative AI Investments [Latest IDC 2024 Report]. Radiant Institute. radiant.institute
  6. Sirimaya (25 April 2025). From Idea to Impact: How Generative AI Transforms ROI. Amity Solutions blog. amitysolutions.com
  7. Schubmehl and Lange (November 2025). What every company can learn from Frontier firms leading the AI revolution. IDC InfoBrief sponsored by Microsoft, IDC #US53838325. marketingassets.microsoft.com
  8. AmplifAI (updated 14 October 2026). 100+ Generative AI Statistics You Need to Know in 2026. Statistics page by a seller of contact-centre software. amplifai.com
  9. Yotzov, Barrero, Bloom, Bunn, Davis et al. (February 2026, revised March and October 2026). Firm Data on AI. NBER Working Paper 34836. The 89% is quoted from the March 2026 revision; the October 2026 revision adds the second round. nber.org
  10. Yotzov, Barrero, Bloom, Bunn, Chawla, Davis, Foster, Jalca, Meyer, Mizen, Navarrete, Oh, Garcia Loredo, Reid, Smietanka, Thwaites and Wang (2026). Firm Data on AI. NBER Working Paper 34836 (February 2026, revised October 2026). What this study did and found