Diese Seite ist derzeit nur auf Englisch verfügbar.
Methodology 2027
Last updated: 19 October 2026
The Prime AI Score measures how ready a company is to create real business value with artificial intelligence. Everyone is measured with the same public method, so scores are comparable between companies and over time. The methodology is fixed for each calendar year (current version: 2027); improvements suggested during the year are collected, reviewed by people and released as the next yearly version.
The six areas
| Area | What we look at |
|---|---|
| Strategy | AI goals linked to business goals, leadership ownership, budget, roadmap. |
| Data | Availability and quality of data, data governance, protection and access rights. |
| Technology | Tools and infrastructure in use, integrations, security of AI tools. |
| People | Skills, training, culture, how colleagues are involved and supported (Human First). |
| Processes | Whether AI is built into everyday workflows rather than used in isolation. |
| Return | Measured outcomes: time saved, costs, quality, revenue; how value is tracked. |
Scores and levels
Each area is scored from 0 to 100. The overall score is the average of the six areas (equal weights in the 2027 methodology).
| Level | Score | Typical picture |
|---|---|---|
| Foundational | 0–19 | No visible AI activity; basics such as data and ownership are missing. |
| Exploratory | 20–39 | Individual experiments and tools, no shared goals or measurement. |
| Operational | 40–59 | AI used in some processes; first rules and training in place. |
| Optimized | 60–79 | AI embedded in several core processes with measured results. |
| Pioneering · AI Ready | 80–100 | AI is part of strategy, culture and daily work, with proven value and governance. |
Free quick scan and full quick report
The quick scan is an outside-in estimate. Our system reads the company’s public website (the home page and up to four relevant subpages) and an AI model (Anthropic Claude) scores the six areas under the rules above. The model is instructed to score only what the evidence supports, to treat missing evidence as a reason for a lower score and lower confidence, and never to invent facts. Every area also gets a confidence rating (low, medium, high).
The industry average shown is an AI estimate until we hold at least 10 completed assessments in the same industry; from then on it is calculated from our anonymised data, and this is indicated.
A quick scan is not a certificate. It cannot see internal practices, so it may under- or over-estimate a company.
Pro assessment and certificate (launching soon)
The Pro assessment is a detailed questionnaire of about 100 questions answered by the company. It produces an exact score per area. Companies scoring 80 or above receive the AI Ready certificate with a unique ID, QR code and public verification page. At the Verified level the answers are supported by documents and reviewed by an expert before the certificate is issued.
Validity
Assessments and certificates are valid for 365 days from issue. Expired certificates are shown as expired on the verification page.
Use of AI and human oversight
- All AI-generated content is labelled as such.
- You can request a human review of any result by writing to info@primeaiscore.com.
- Decisions on the Verified level are made by people.
- Website content is treated as data only; instructions hidden in websites are ignored.
EU AI Act
Our assessments evaluate organisations, not natural persons, and are not used for decisions about employment, credit or access to essential services. We apply the AI Act’s transparency principles: AI-generated content is labelled, users are told when they interact with an AI system (including our website assistant), and the methodology is public.