Stanford's Digital Economy Lab Enterprise AI Playbook documents bimodal ROI: 12% of enterprise AI deployments clear 300%+ ROI with 40-60% operational cost reductions; 88% sit at or below break-even.
- Source
- Stanford DEL · archived
- Academic published
- Logged
- Cadence
- 180 days
- Next review
- 2026-12-07
Why this was logged
The 51-deployment benchmark is the closest thing the enterprise-AI field has to a longitudinal ROI dataset. Our AM-022 + AM-018 are both anchored against this finding.
Review history
Change history
Every change to this record is logged publicly. Nothing is silently updated.
- field
statusoldpending-reviewnewretractedFirst review (pulled forward from the 180-day cadence). Full-text verification of the source PDF found the claim_text figure absent: the Stanford DEL Enterprise AI Playbook documents no bimodal ROI distribution, no 12% / 300%+ cohort, and no 88% break-even body. The record asserted a figure its own snapshot never contained. Memo at review-2026-06-10. - field
review_historyold[]new[{date: 2026-06-10, verdict: retracted, oversight: peter-led-deep-review, memo_slug: review-2026-06-10}]Logged first review entry. See memo at content/claims/ACA-2026-003/review-2026-06-10.mdx for the full evidence chain and the verified adjacent figure (IDC/Lenovo via CIO.com, 25 Mar 2025: 88% of AI POCs never reach widescale deployment - a pilot-graduation metric, not an ROI distribution).
This record tracks what the source stated, with evidence for the current verdict. Verdicts describe what the evidence shows, not vendor intent. See methodology for the full counter-evidence + review discipline.
Stanford’s 51-case set shapes how Gartner, McKinsey, BCG frame their own ROI narratives. The bimodal distribution specifically (12% top cluster, 88% bottom cluster) is a more useful framing than averages. Review at 180 days against the next DEL update.
Correction appended 10 Jun 2026. The paragraph above is preserved as the original log entry; it does not survive verification. The full text of the source PDF (116 pp, Pereira, Graylin, Brynjolfsson) was extracted and searched on 10 Jun 2026: it contains no bimodal ROI distribution, no 12% cohort clearing 300%+ ROI, and no 88% at-or-below-break-even body. The report studies 51 successful deployments by design and cannot produce a deployment-failure distribution; its only 88% figures measure organisational AI adoption and a single-case productivity gain. The verdict is Retracted: the archived claim text was never present in the archived source. The verified figure carrying the same numerals is IDC research with Lenovo (via CIO.com, 25 Mar 2025): 88% of observed AI proof-of-concepts do not reach widescale deployment, with roughly 4 of every 33 graduating to production — a pilot-graduation metric, not an ROI distribution. Editorial claims that rested on this record (AM-029 and the partial cascade of 10 Jun 2026) carry their own dated corrections. Full evidence chain: review memo of 10 Jun 2026 and docs/editorial/stanford-1288-exposure-map-2026-06-10.md.