CE-N compares the eighteen bottlenecks using three research judgments: whether customers have budgets, whether a solution is feasible, and whether suppliers can retain value. The score expresses a research priority. It does not measure a company’s return on invested capital or expected stock return.
Six ratings, three calculation inputs
The current records use a 1–5 scale. Higher values indicate more of the property described, not equal units of economic value.
| Axis | Question | Current range | Calculation role |
|---|---|---|---|
| Budget | Is there an identified buyer with money available to spend? | 3.5–5.0 | One third of the standardized composite |
| Solvability | Can the task be solved reliably under its intended conditions? | 2.6–4.2 | One third of the standardized composite |
| Value capture | Can the supplier retain economic value after costs and competition? | 3.4–4.8 | One third of the standardized composite |
| Maturity | How established are the technology and deployments? | 2.0–4.0 | Context |
| Pain | How costly or consequential is the customer’s problem? | 4.0–5.0 | Context |
| Timing | What makes adoption feasible or urgent in the stated period? | 4.1–5.0 | Context |
What supports an input rating?
The bottleneck pages describe the research reasoning and show the stored ratings. The current records do not provide a complete, source-linked rubric explaining every decimal rating. A value such as 4.1 should therefore be read as an analyst judgment with unresolved measurement precision, rather than an observed physical quantity.
The scoring review sets out the evidence requirements and rating anchors approved for the pilot. Replacement ratings still require approval. Each bottleneck’s score panel makes the remaining evidence requirements visible.
Why Pain, Timing, and Maturity remain context
The current calculation excludes these three axes. Pain and Timing cluster toward the high end of the scale, but that alone does not demonstrate that they contain no useful information. Selection of already-important bottlenecks, overlapping definitions, or insufficiently specific rating anchors could explain the clustering.
Maturity describes progress already achieved. Whether it adds useful information beyond Solvability is also a review question. We keep the current calculation unchanged while examining these issues.
The calculation behind the stored scores
The documented v0.2 calculation uses the following steps:
- Cap each of the three input axes at its 10th and 90th percentiles across the eighteen problems.
- Subtract the axis mean and divide by its population standard deviation. A constant axis contributes zero.
- Average the three standardized values with equal weights.
- Map the composite to 0–100 using
(composite + 3) / 6 × 100, clipped to that range. - Multiply by 0.85 if any of the three raw input ratings is below 2.5, then round to one decimal place.
A read-only audit reproduced all eighteen stored scores using this rule. The website displays the current approved score; historical calculations and candidate comparisons remain in the source repository.
A worked example
For P18, Vertical workflow ROI, the stored inputs are Budget 5.0, Solvability 4.1, and Value capture 4.8. Applying the calculation to the full board produces the stored CE-N score of 71.2.
The full board is needed because the percentile cutoffs, means, and standard deviations depend on all eighteen records. A change to one record can affect other records’ scores. Fixed 0–100 endpoints do not make scores from different research cuts directly comparable when the reference distribution changes.
The audit script and input hashes accompany this revision in tools/research-review and operations/reviews of the source repository. The review page explains the evidence and rating rules.
From a score to a portfolio
Company valuation, eligibility, and position sizing are separate steps. The website displays one current approved CE-N score. Internal evidence-confidence analysis is not displayed as a second CE-N score; this presentation rule does not change portfolio calculations.