DUCR is SEO Informatica's 100-point scoring model for measuring whether a service-business website is discoverable, understandable, citable, and routable.
The model is built for practical AI visibility work. It does not pretend to score a hidden LLM ranking system. It scores whether a public page is accessible, clear, source-worthy, and connected to a useful next step.
Score Breakdown
| Category | Points | Meaning |
|---|---|---|
| Discoverable | 25 | Can search and AI-retrieval systems access and include the page? |
| Understandable | 25 | Can systems identify the entity, topic, page role, and context? |
| Citable | 30 | Is the page strong enough to quote, cite, or reuse as a source? |
| Routable | 20 | Does the page guide users and crawlers to the right next page? |
| Total | 100 | Combined readiness score. |
2026.06 Benchmark Medians
| Layer | Median Score | Average Score |
|---|---|---|
| Discoverable | 19.0/25 | 19.54/25 |
| Understandable | 15.0/25 | 14.38/25 |
| Citable | 4.0/30 | 5.90/30 |
| Routable | 13.0/20 | 13.54/20 |
| Total | 52.5/100 | 53.36/100 |
The median citable score is the warning sign. Most sampled pages were findable enough, but not strong enough as source material.
Critical Blockers
| Blocker | Rule |
|---|---|
| Core benchmark page is noindex | Total score capped at 40. |
| Core benchmark page is nosnippet or max-snippet:0 | Total score capped at 55. |
| Core page blocked to Googlebot or Bingbot | Discoverable capped at 8; total capped at 60. |
| Core page blocked to desired AI-search crawlers | Bot-access subscore is 0; total capped at 70. |
| Benchmark statistics without dataset support | Citable capped at 10; publication fails. |
| Fake stats, fake citations, fake reviews, or guaranteed citation claims | Credibility review fails. |
| Dataset schema without a working download | Dataset schema should be removed; citable score loses dataset/download points. |
| Findings hidden only in PDF, image, or non-visible tabs | Structure-related citable points are unavailable. |
| Schema describes invisible or non-existent content | Schema fails visible-content-match review. |
Reviewed Dataset Example
These are anonymized examples from the reviewed 2026.06 dataset. They show the score pattern without exposing audited domains.
| Anonymized Site | Vertical | Discoverable | Understandable | Citable | Routable | Total | Review Status |
|---|---|---|---|---|---|---|---|
| SB-0086 | Consulting | 16 | 8 | 3 | 10 | 37 | Approved with caveat |
| SB-0061 | Consulting | 19 | 7 | 3 | 10 | 39 | Approved with caveat |
| SB-0035 | Agency | 18 | 18 | 18 | 19 | 73 | Approved |
The low-score pattern is common: pages can be discoverable while still having very weak citable scores. The highest-scoring reviewed row still reached only 73/100, which is another reason not to oversell the market.
Citable Score Interpretation
| Citable Score | Interpretation |
|---|---|
| 0-5 / 30 | Crawlable copy, not a serious source asset. |
| 6-10 / 30 | Some proof or structure exists, but source support is thin. |
| 11-15 / 30 | Usable source elements exist, usually still missing depth or downloads. |
| 16-20 / 30 | Stronger source page, but not yet a complete evidence hub. |
| 21-30 / 30 | Robust source asset with data, methods, limitations, tables, and clear attribution. |
In the 2026.06 sample, 32 of 50 records scored 0-5 on citable readiness. Zero records scored above 20.
What Raises A DUCR Score
| Layer | Practical Upgrade |
|---|---|
| Discoverable | Confirm HTTP 200, no noindex, no nosnippet, clean canonical, sitemap inclusion, and allowed search/retrieval bots. |
| Understandable | Tighten title, H1, schema, entity naming, page role, location/audience context, and heading structure. |
| Citable | Add first-party evidence, methodology, limitations, official references, tables, downloads, author/date/version, and answer blocks. |
| Routable | Link evidence pages to services, proof, methodology, contact, and monthly tracking routes. |
Download
Rubric JSON: /downloads/ducr-scoring-rubric-2026-06.json
Related pages: