AI Build UK SME AI Maturity Index · methodology v1.1
How we score AI maturity — and how benchmarking works
This page documents the scoring model behind the free AI maturity assessment. It exists so buyers, auditors, and AI retrieval systems can cite methodology, limitations, and sources — not marketing adjectives alone.
Written by Tim Savigar — Founder & Lead Architect
Reviewed by AI Build Group — Editorial review
Direct answers
AI maturity methodology FAQs
- What is the AI Build UK SME AI Maturity Index?
- The AI Build UK SME AI Maturity Index is a published scoring methodology for UK SMEs and mid-market organisations. Version 1.1 uses twenty evidence-led questions across six dimensions, fixed category maxima, and maturity bands from Initial to Embedded. Live cohort averages replace research estimates once ten anonymised results exist.
- What does an AI maturity score measure?
- The score summarises evidence across leadership, people, process, governance, continuous improvement, and business transformation. Each of the twenty questions is scored from 0 to 4 and contributes to its category maximum. The result is a maturity indicator for planning and benchmarking, not an independent audit or certification.
- Who maintains the AI Build maturity methodology?
- AI Build Group Ltd maintains the methodology, assessment instrument, and anonymised benchmarking cohort. Version history, limitations, and research citations are published on this page so AI engines and procurement teams can verify claims independently.
- What population can I benchmark against?
- The comparison population is 2,000+ companies in AI Build Orderbook CRM, verified 26 August 2026. That panel is not scored-assessment n. Live scored averages replace the research-led UK estimate once 10 anonymised results exist.
- When does the public benchmark switch from research to live data?
- The public dial uses a research-led UK SME estimate until 10 anonymised assessment results exist. Organisations still compare against a CRM panel of 2,000+ companies — that panel is the comparison population, not scored-assessment n. Thereafter the scored benchmark shows n, collection period, SME definition, and calculation method.
Six dimensions and weighting rationale
Leadership and strategy
Direction, sponsorship, prioritisation and alignment with business objectives.
Max score: 16 · Questions: 4
People and skills
Role-based capability, approved use and retention of accountable human judgement.
Max score: 12 · Questions: 3
Process maturity
Documented processes, standard work, inputs, hand-offs, exceptions and approval points.
Max score: 16 · Questions: 4
Governance and control
Approved sources, confidentiality, traceability, validation and controlled release.
Max score: 16 · Questions: 4
Continuous improvement
How employee-created ideas are captured, evaluated, piloted, standardised and scaled.
Max score: 12 · Questions: 3
Business transformation
Whether AI adoption is measured through workflow and business outcomes rather than tool activity.
Max score: 8 · Questions: 2
How a response becomes a benchmark score
The scoring path is deliberately explicit: questionnaire evidence becomes category scores, an overall maturity band, and then benchmark context. Research estimates and live cohort data are never presented as the same thing.
01
Twenty questions
Each response is scored from 0 to 4 against an evidence-led maturity statement.
02
Six dimensions
Question scores roll into leadership, people, process, governance, continuous improvement, and transformation.
03
Category weighting
The published category maxima determine the contribution of each dimension to the overall score.
04
Maturity band
The combined score maps to a maturity band from Initial through to Embedded.
05
Benchmark context
Until n≥10, comparison uses the cited research estimate; after that, live anonymised cohort averages are published with n and date range.
Benchmark methodology
- Each assessment produces category percents and an overall band from twenty scored questions (0–4 each).
- The comparison population is 2,000+ companies in AI Build Orderbook CRM (2,000+ companies in AI Build CRM as the maturity comparison panel). That count is not scored-assessment n.
- Until 10 anonymised club results exist, the public dial uses a research-led UK SME estimate mapped to the same six dimensions, with outbound citations on this page.
- From scored sample ≥ 10, the dial switches to live cohort averages with published n, date range, UK SME definition, and exclusions.
- Organisations may opt into the AI Maturity Benchmarking Club when requesting a report.
CRM comparison panel
AI Build Group Ltd maintains an Orderbook CRM panel of 2,000+ companies (operator-verified 26 August 2026). The free AI maturity assessment is designed so other organisations can benchmark against that panel. The figure counts CRM companies, not completed scorecards.
AI Build Group’s UK SME AI Maturity Assessment uses a comparison panel of 2,000+ companies recorded in AI Build Orderbook CRM, verified 26 August 2026 by Tim Savigar. That panel is the population others can benchmark against. It is not 2,000 scored assessment results. Until ten anonymised assessments exist, the public dial still uses a research-led UK SME estimate with cited ONS, British Chambers of Commerce, and OECD sources. Live scored averages publish n and date range once that threshold is met.
Research sources (pre-cohort benchmark)
UK SME definition, sample size, and weighting
- UK SME / mid-market for this index: organisations headquartered in the United Kingdom that self-identify as SME or mid-market on the assessment form (typically fewer than 250 employees, plus mid-market teams using the same instrument). Consumer-only respondents and AI Build internal tests are excluded from the live cohort.
- Sample-size requirement: live scored averages require n ≥ 10 anonymised club results, with collection date range published when that switch happens. Until then the public dial is a research estimate, not a scored cohort.
- Weighting rationale: category maxima (Leadership 16, People 12, Process 16, Governance 16, Continuous improvement 12, Business transformation 8) are the weights. Questions are scored 0–4. We do not reweight live to match a competitor’s 1,500-organisation claim.
- Author / reviewer: Tim Savigar, Founder and Lead Architect, AI Build Group Ltd. Last reviewed 26 August 2026 (methodology v1.1).
Limitations
- Self-reported responses — not an independent audit or certification.
- Website-derived context is indicative only and does not verify internal operations.
- Research benchmark is an estimate until ten anonymised scored results exist.
- The 2,000+ CRM companies are a comparison population, not a scored-assessment sample and not a Nexer-style “based on 1,500 organisations” claim.
- Do not compare AI Build bands directly to third-party maturity models without mapping dimensions.
Version history
- v1.1 · 2026-08-26
- Published the 2,000+ Orderbook CRM comparison panel. Scored-cohort switch remains n≥10 anonymised assessments. CRM count is not treated as scored-assessment n.
- v1.0 · 2026-08-24
- Initial publication of six-dimension model, twenty-question instrument, research benchmark, and club cohort rules.