Cyber risk rating
Also called security rating, cyber security score
A cyber risk rating is a score of an organisation's security posture, built from data observable outside its network. Rating providers scan for exposed services, unpatched systems, leaked credentials, misconfigured email and signs of malware. Firms use ratings to screen and continuously monitor suppliers between formal assessments, and to benchmark their own exposure.
How it works
A third-party risk team loads its supplier list into a rating service, which matches each supplier to its internet domains and IP ranges. The service scores findings by severity and tracks changes daily. A drop below an agreed threshold, or a new critical finding, triggers an alert. Analysts contact the supplier, request remediation and, for critical providers, escalate to the relationship owner and the security function.
Ratings fill the gap between annual questionnaires and audits. Regulators expect ongoing monitoring of ICT providers rather than point-in-time checks, and ratings offer a scalable signal across thousands of suppliers. They see only what is visible from outside, so firms treat them as one input. Mismatched assets and false positives are common, and suppliers can dispute findings with the rating provider.
Main rating providers
SecurityScorecard groups findings into 10 factor scores, such as network security, DNS health, patching cadence, endpoint security and leaked information. The overall score is a weighted average of the factor scores on a 100 point scale and maps to a letter grade from A to F. Scores update daily.
Bitsight rates on a scale of 250 to 900, where a higher number is better. It calculates ratings daily from two classes of externally observable data, configuration and security events, and normalizes for company size. It groups scores into Basic (250 to 630), Intermediate (640 to 730) and Advanced (740 to 900) bands.
UpGuard uses a subtractive model. Each domain or IP address starts at 950 points and loses points for every failed check, with the deduction set by the severity and weight of the risk. Organization scores are built from the scores of all its domains and IPs, and scans run every 24 hours.
Mastercard's RiskRecon uses a numeric scale of 0.0 to 10, with 10 the best, applied to security domains, individual criteria and an overall rating, with letter grades on top. Its model weighs the prevalence and severity of issues and the value of the systems where they sit.
The methods differ in scale, direction and construction. SecurityScorecard averages weighted factor scores, UpGuard deducts from a fixed maximum, and RiskRecon factors in the value of the affected systems. The same supplier can therefore look different across providers, and firms compare grades and trends within one provider rather than raw numbers across providers.
Using ratings in third-party risk management software
Third-party risk management platforms pull ratings into the vendor record through APIs and connectors. The Bitsight integration for ServiceNow Vendor Risk Management, for example, brings in daily and historical ratings and risk vector grades, and lets teams set alerts for significant rating changes. ServiceNow documents integration of scores from risk intelligence providers that can create an assessment, issue, task or email when a rating is updated. OneTrust lets customers using SecurityScorecard configure workflows that act automatically when a third party's score changes.
Platforms use the rating alongside inherent risk data to place each vendor in a tier. The tier sets the type and frequency of assessment, so a critical vendor gets a deeper review and closer monitoring than a low-risk one. A rating that falls below the agreed threshold, or a drop in a risk category, can then push a new questionnaire or assessment to the vendor and open a task for the risk owner. Teams usually set thresholds per tier, with tighter limits for critical suppliers.
Limitations and disputes
Ratings rely on outside scans, so they can flag issues that do not exist or miss controls that sit inside the network. Attribution is the most common source of error. Shared hosting, cloud infrastructure, content delivery networks and legacy IP blocks can cause findings to land on the wrong organization, a problem SecurityScorecard describes in its support documentation. A finding tied to an exposed endpoint stays on the rating until the underlying issue is fixed.
Rated companies can challenge a score. SecurityScorecard says any rated organization can dispute a finding at no charge, customer or not, by submitting evidence that an issue was wrongly associated with it. A company that has fixed an issue files a remediation request, and once approved the score usually updates within 48 hours. Bitsight lets rated organizations challenge the assets, findings and interpretations behind their rating through its ombudsman process, with appeals to its Policy Review Board, and says it typically resolves disputes within 7 to 10 business days.
The Principles for Fair and Accurate Security Ratings, published by the US Chamber of Commerce in 2017, call for transparency on methodology and for rated organizations to be able to dispute, correct and appeal their ratings.
The rules
Requires firms to manage ICT third-party risk across the contract life cycle, including monitoring of providers supporting critical or important functions.
EU financial entitiesExpects ongoing monitoring of third parties, with depth matched to the risk of each relationship, including information security performance.
US banking organisationsAdds a Govern function with cybersecurity supply chain risk management outcomes, including monitoring suppliers over the relationship.
Organisations of all sizes and sectorsCommon questions
How accurate are cyber risk ratings?
Ratings reflect only data visible from the internet. They can misattribute assets or miss internal controls entirely. Firms validate significant findings with the supplier and combine ratings with questionnaires, certifications and audit evidence.
Do regulators require cyber risk ratings?
No rule names ratings specifically. Rules in the EU, UK and US require ongoing monitoring of third-party providers, and many firms use ratings as one way to meet that expectation at scale.
Can a company improve its cyber risk rating?
Yes. Fixing exposed services, patching visible systems, correcting email security records and removing outdated assets usually lifts a score. Companies can also challenge findings they believe are wrong.
Are ratings used outside third-party risk?
Yes. Cyber insurers use them in underwriting, boards use them to benchmark the firm against peers, and acquirers use them in due diligence on targets.
Our coverage
News and analysis on third-party risk from RegTech Post.
Read the coverage →RegTech PostThe Compliance Officer’s Toolkit: Building Cybersecurity into Corporate GovernanceRead the article →RegTech PostFive Strategies to Contain Third-Party RiskRead the article →