We publish comparable metrics for AI, with room to grow into the internet economy
Lyell Data provides standardized datasets on AI adoption and GenAI/ChatGPT usage—each with a point estimate and a modeled range. Our framework extends to upcoming verticals: Internet Economy and Digital Marketing.
Make metrics comparable—and honest about uncertainty
Cross-country and cross-industry stats are often apples-to-oranges. We standardize definitions (e.g., “% of firms using AI in ≥1 process”), apply SME vs Large splits, and always show a modeled range. We prefer reproducibility over black-box claims, and document revisions like features—not bugs.
Principles & guarantees
- Modeled + Public Data: open datasets, directional proxies, documented anchors.
- Ranges, not absolutes: uncertainty bands on every estimate.
- Revisions documented: semantic versioning and a public changelog.
- Privacy & ethics: no PII; aggregated signals only; respect for national data laws.
- No sponsored metrics: vendor telemetry appears only if public or licensed, with labeling.
Authors & domains
Maarja Veskimägi
Leads country×industry adoption modeling, SME/Large splits, ROI signals, and cross-sector calibration.
Siti Norhayati Omar
Builds MAU and usage models from traffic, app signals, and panels; extends to internet-economy indicators.
Quality assurance & revisions
QA checks
- Outliers: |z| > 3.5 → flag
- Trend breaks: |ΔYoY| > 15 pp → manual review
- Coverage < 60% → widen band / reweight
- SME–Large sanity: |Δ| > 30 pp → recalibrate
Revision policy
- Semantic versioning on datasets and methodology
- Back-revisions documented with reasons
- Stable CSV URLs; metadata includes
method_version
How to cite
License: CC BY 4.0. Please attribute as: “Lyell Data (2025). AI Adoption in <Country> — 2025. https://example.com/ai/industry/<slug>”
Press kit
- Short bio & logo (PNG/SVG)
- Color tokens & usage
- One-paragraph boilerplate
Publications over time
Talk to us
General & data inquiries. We respond within 2–3 business days.
Weekly snapshot
A short weekly update with new datasets and notable revisions.