Method, read 13 Nov 2026

# Seven criteria, weighted for proof that an exposure is real, fixed before any score

**The short answer:** Each vendor is scored 0 to 10 on seven criteria: validation (weight 25), prioritization on evidence (15), discovery breadth (15), mobilization (15), works with the stack in place (10), scoping and reporting (10) and transparency and proof (10). The weights were committed on 12 Nov 2026, before the first score.

[See the index](https://ctemvendors.com/) | [Read every score](https://ctemvendors.com/index/)

## The question the index answers

Which vendors cover which stage of a CTEM program, and who can prove that an exposure is real in this environment? A security lead starting or maturing a CTEM program, who already has scanners and findings and needs to know which exposures can be exploited here and get them fixed. The criteria follow the five stages of CTEM as Gartner named them, and validation carries the most weight because it is the stage that separates CTEM from vulnerability management.

## The weights and the scoring anchors

| Criterion | Weight | What is scored | Anchors |
|---|---|---|---|
| Validation | 25 | Proof of which exposures can be exploited in this environment: attack simulation, automated penetration testing or red teaming, and testing of the controls in place. | Count the capabilities documented on a public page as part of the vendor's own product: (a) breach and attack simulation or attack scenario execution in the customer's environment, (b) automated penetration testing or automated red teaming against real assets, (c) testing of deployed security controls (EDR, SIEM, email, web, WAF) with results per control, (d) attack path analysis that shows reachable paths to critical assets, (e) continuous or scheduled re-testing after a fix, (f) a threat or attack library updated with new threats. Score: 0 capabilities 0; analysis only (d alone, or exploitability inference from threat intelligence with no testing) 3; one of a, b, c 5; two of a, b, c 7; all three of a, b, c 8; add 1 for each of e and f when at least one of a, b, c is present, to a maximum of 10. Validation that comes only through a partner integration scores at most 3. Vendor-only evidence caps the score at 7. |
| Prioritization on evidence | 15 | Ranking that uses validation results, threat intelligence and business context, not severity scores alone. | Inputs documented as used in the vendor's ranking of exposures: (a) exploitability evidence from its own validation or testing, (b) threat intelligence or known-exploited data (such as CISA KEV, EPSS, threat actor activity), (c) business context or asset criticality, (d) reachability or attack path context, (e) compensating controls in place. Score: severity scores only 2; one input 4; two 6; three 7; four 8; all five 10. Vendor-only evidence caps the score at 7. Nothing published 0. |
| Discovery breadth | 15 | Assets and exposures across cloud, on-premises, identity and the external surface, found by the product or taken in from other tools. | Four surfaces: cloud, on-premises and endpoints, identity, external attack surface. Per surface: found natively by the product 2.5; taken in only from other tools through integration 1.5; none found 0. Sum, rounded to the nearest whole number, maximum 10. Vendor-only evidence caps the score at 7. |
| Mobilization | 15 | Remediation guidance, mitigation rules for the controls already deployed, ticketing and automation integrations, tracking to closure. | Four capabilities documented on a public page: (a) remediation guidance per exposure, (b) mitigation rules or detection content for controls already deployed (such as EDR or SIEM rules, IPS signatures, configuration changes), (c) ticketing or workflow integration (such as ServiceNow, Jira) or automated response, (d) tracking to closure with re-validation or SLA tracking. 2.5 each, rounded to the nearest whole number. Vendor-only evidence caps the score at 7. Nothing published 0. |
| Works with the stack in place | 10 | Integrations with scanners, EDR, SIEM and cloud platforms, as published. | Four categories with at least one named integration published by the vendor: vulnerability scanners, EDR or XDR, SIEM or SOAR, cloud platforms (AWS, Azure, Google Cloud). 2 per category, plus 2 if the vendor publishes a public integrations list or catalog with 25 or more named integrations. Maximum 10. Vendor-only evidence caps the score at 7. This criterion is where breadth counts; it never exceeds 10 percent of a ranking. |
| Scoping and reporting | 10 | Business context, views by unit or crown-jewel asset, reporting a board can read. | Three capabilities documented on a public page: (a) scoping by business unit, asset group or crown-jewel asset, (b) executive or board-level reporting with a risk or exposure measure, (c) trend tracking or benchmarking over time. Score: none 0; one 4; two 7; all three 9; add 1 if a second source shows it in use. Vendor-only evidence caps the score at 7. |
| Transparency and proof | 10 | Public documentation, a published price, peer reviews as published, analyst placements. | Four items: (a) public product documentation readable without a login 3, (b) a published price or price list 3, (c) peer reviews as published on Gartner Peer Insights, G2 or PeerSpot with at least 20 reviews 2, (d) a named placement or mention in a third-party analyst report (Gartner, Forrester, IDC, GigaOm, KuppingerCole, Frost and Sullivan) 2. Sum, maximum 10. Items (c) and (d) are third-party by nature; for (c) and (d) a vendor page that reports them counts until the third-party page is read. |

Each criterion is scored 0 to 10 from the evidence files. The total is the weighted mean on a 0 to 10 scale; pages print it to two decimals.

## The rules every score follows

- A claim published only by the vendor itself (its site, documentation, press releases, customer statements it publishes) scores at most 7 of 10 on a criterion until a second source not controlled by the vendor confirms it.
- None found scores 0 and is written 'none found on <date>', never as a finding that the vendor lacks the capability.
- Breadth facts (module counts, integration counts, headcount, funding) carry at most 10 percent of any ranking: they enter only through 'Works with the stack in place' (weight 10), and there only as published integrations.
- No weight, criterion or anchor changes after the first score exists. A score changes only when a sourced fact changes, and the change is logged with the fact in data/scores/corrections.jsonl.
- Scoring is blind: scorers see vendors as codes, not names.
- Equal totals share a rank (shown as 1=) and are listed alphabetically; a head-to-head with equal totals is a tie.
- A vendor claim is labeled as the vendor's on every page.
- Each vendor's strength is stated on its own page in one sentence, and so is the point where it is thin.

## How the scores were made

Evidence came first: 2190 records for 18 vendors, each with its source URL and read date, from each vendor's product pages, documentation, pricing page, integrations catalog, release notes and published research, and from at least two third-party sources per vendor. Each criterion was then scored for all 18 vendors by a scorer that saw the anchors and the evidence with vendor names replaced by codes. A second reviewer with fresh context checked every score against its anchor and its evidence. In review before the second evidence pass, 67 scores changed and 5 public reasons were reworded without a score change; each entry is listed on the [corrections page](https://ctemvendors.com/corrections/) with its reason.

## Second evidence pass, 13 Nov 2026

On 13 Nov 2026 every vendor went through a second evidence pass under one search plan: the same eight news queries, the same source types (review sites, integration listings run by other companies, analyst and reference pages, the vendor's own newsroom) and the same budget of 35 page reads per vendor, plus the same six-read supplement of integration listings for the three vendors whose researchers had not reached that source type. It looked for three things for every vendor alike: a second source for each score held at 7 because only the vendor's own pages supported it, anything published since the first pass, and evidence against each current score. It added 435 records from 228 sources (245 third-party, 106 from the vendors, 84 searches that found nothing, each logged as none found with its date). The new records were scored blind, by criterion, with the same anchors and weights; 7 scores changed and 5 more public reasons were reworded, each logged on the [corrections page](https://ctemvendors.com/corrections/). Weights, criteria and anchors did not change. The [changelog](https://ctemvendors.com/changelog/) lists what the pass changed on the site.

## Stage weights

Each stage page ranks the same scores on weights written for a buyer looking at that stage, committed with the default weights before any score.

Scroll the table sideways for every column.

| Stage | Validation | Prioritization | Discovery | Mobilization | Stack | Scoping | Transparency |
|---|---|---|---|---|---|---|---|
| [Scoping](https://ctemvendors.com/stages/scoping/) | 10 | 10 | 15 | 5 | 10 | 40 | 10 |
| [Discovery](https://ctemvendors.com/stages/discovery/) | 5 | 10 | 45 | 5 | 20 | 5 | 10 |
| [Prioritization](https://ctemvendors.com/stages/prioritization/) | 15 | 45 | 10 | 5 | 10 | 5 | 10 |
| [Validation](https://ctemvendors.com/stages/validation/) | 50 | 10 | 5 | 10 | 10 | 5 | 10 |
| [Mobilization](https://ctemvendors.com/stages/mobilization/) | 10 | 10 | 5 | 45 | 15 | 5 | 10 |

## Situation weights

Each of the five buyer situations ranks the same scores on its own weights, written for that buyer and committed with the default weights.

Scroll the table sideways for every column.

| Situation | Validation | Prioritization | Discovery | Mobilization | Stack | Scoping | Transparency |
|---|---|---|---|---|---|---|---|
| [Starting from vulnerability management](https://ctemvendors.com/situations/starting-from-vm/) | 25 | 20 | 10 | 15 | 15 | 5 | 10 |
| [Several scanners, no way to tell what matters](https://ctemvendors.com/situations/many-scanners/) | 15 | 25 | 15 | 15 | 20 | 5 | 5 |
| [Regulated, must prove controls work](https://ctemvendors.com/situations/regulated-evidence/) | 35 | 10 | 10 | 10 | 5 | 20 | 10 |
| [Lean team, no red team](https://ctemvendors.com/situations/lean-team/) | 35 | 15 | 5 | 20 | 5 | 5 | 15 |
| [Quarterly board reporting](https://ctemvendors.com/situations/board-reporting/) | 20 | 15 | 10 | 10 | 5 | 30 | 10 |

## The ranked views

- **Index**: all 18 vendors ranked on the weights above; ties share a position.
- **Stages**: each of the five stage pages ranks all vendors on that stage's own weights (table above).
- **Situations**: each of the five situation pages ranks all vendors on its own weights and names the top three.
- **Head-to-heads**: two vendors on the weights above; the higher total leads, equal totals are a tie. Published: Cymulate with each platform vendor, and every pair among the platform vendors (78 in all). Pairs between two validation vendors (Cymulate, Picus Security, AttackIQ, SafeBreach, Pentera, Horizon3.ai) are not published; their scores sit side by side on the validation stage page.
- **Alternatives**: none published.

## How the Stage Coverage Map is read

A stage is "native" when the vendor's own product does the work, "via integration" when the vendor documents the stage only through another tool or a partner, and "none found" when nothing was found on 12 Nov 2026. The map is evidence, not a score: the scores above decide every ranking. See the [map's methodology](https://ctemvendors.com/stage-coverage-map/methodology/).

## Evidence limits

- Gartner Peer Insights, G2 and TrustRadius refused automated reads from the research runner for most vendors on 12 Nov 2026; review counts are taken from the vendor's own report of them where the third-party page could not be read, and labeled so.
- Several documentation portals render only with JavaScript or sit behind a login, so their content could not be read as text; where that happened the score rests on product pages and the docs URL is recorded.
- Few vendors in the set publish a price for their CTEM product on their own site; marketplace listings are counted where a price is shown.

## Sources

- [https://web.archive.org/web/20260110234446/https://www.gartner.com/en/articles/how-to-manage-cybersecurity-threats-not-episodes](https://web.archive.org/web/20260110234446/https://www.gartner.com/en/articles/how-to-manage-cybersecurity-threats-not-episodes), read 12 Nov 2026
Last reviewed 13 Nov 2026.

Canonical: https://ctemvendors.com/method/
