Research investigation R1101 / claim audit
Four Claude customer accounts report benefits, but none makes net ROI calculable
The four sampled accounts support different kinds of benefit accounting, not a common return metric. None discloses sufficiently matched benefits and attributable costs to calculate net ROI or a numerical net-ROI bound.
Early sample. Read individual values rather than treating the pattern as a market trend.
Version ledger
Frozen public editions
Each edition preserves the records, method, sources, and downloads available at publication time.
The October 30–31 treatments supplied the coverage hook, not independent validation of four outcomes. The underlying documents are Anthropic-published customer accounts, including an edited Airtree interview; their publication dates are unverified. The contribution here is the accounting comparison, rather than another adoption or savings headline.
- Dataset ID
- spd:which-claude-customer-savings-claims-can-be-converted-into-net-roi-1653e465
- Stable URL
- /research/which-claude-customer-savings-claims-can-be-converted-into-net-roi-1653e465
- Version
- v1
- Coverage
- Live collection
- Records
- 4
- Fields
- 7
- Updated
Measurement technique
How to read this report
- 01Use only the supplied evidence for the four named customer documents. Preserve the nominal cutoff of November 1, 2026, 16:30:54.294 UTC. The supplied snapshots were retrieved at 16:31:39.995 UTC and their URLs checked live; exact cutoff versions were not established. Reusing this evidence is not a new collection or experiment.
- 02Build an evidence_matrix with one row per case and columns for reported benefit, baseline, measurement period, supported arithmetic, attributable costs, overlap risks and net-ROI calculability. Keep estimates, annualized rates, staff time and adoption distinct.
- 03LG CNS: record roughly half-cost delivery against an estimated conventional rebuild and the reported seven-month migration, November 2025–June 2026. The implied approximately 50% reduction is relative counterfactual cost avoidance, not reproducible monetary savings or net ROI. Exact expenditure and cost inclusions are undisclosed.
- 04Duvo: record €1.45 million annualized savings in the first week and more than €2.8 million annualized across three months as successive deployment descriptions, not additive benefits. Annualization workings, persistence assumptions, purchasing baseline and customer costs are absent; whether expenses were deducted is unspecified.
- 05Cyera: subtract roughly 30 minutes from the reported five–six daily triage hours to obtain approximately 4.5–5.5 hours released daily. Keep 88% weekly usage across roughly 1,500 employees in an adoption column. Neither measure supplies labor valuation, demonstrated expenditure reduction or priced implementation and review effort.
- 06Airtree: subtract two minutes from 30 minutes to obtain 28 minutes released per board summary; retain the separately reported two hours saved monthly on marketing reporting. Do not monetize research falling from two days to unspecified minutes or extrapolate without task volumes and review-time accounting.
- 07Classify net ROI as publicly indeterminate for every sampled document. The buyer checklist is matched workload and baseline definitions, dated observation windows, realized-versus-annualized reconciliation, overlap exclusions, attributable software/model, integration, training, governance, review and maintenance costs, and evidence connecting released capacity to financial value.
Sources
Evidence
1 publisher supporting 4 records. Expand a publisher to inspect its cited pages.