A G2 Grid plots each product on two scores, Satisfaction and Market Presence, both scaled from 0 to 100 against other products in the same category. Satisfaction comes mainly from review form answers. Market Presence leans on weighted review count plus company signals. G2 says reviews are subjective experiences, not expert opinions.
What goes into Satisfaction
According to G2's published scoring documentation, Satisfaction is built from the answers on its review form. Questions aimed at users, such as ease of use, whether the product meets requirements, and quality of support, carry the most weight. Questions aimed at administrators come next, and general questions such as likelihood to recommend follow.
The answers are then weighted by review volume, recency, review quality (G2 uses a readability score for this) and review source. Reviews gathered without an incentive count for more than incentivized ones.
Notice what is missing from that list. Nothing in it measures detection accuracy, false positive rates or fix quality. A security tool can score well on support and ease of use and still miss findings in your stack.
What goes into Market Presence
The most heavily weighted input is the weighted review count for the category. G2 also lists employee counts, estimated revenue, web presence, vendor growth and year founded. Product level metrics get more weight than vendor level ones.
So Market Presence describes how visible and established a product is. It does not describe how well it works. A newer product with fewer reviews can sit low on that axis while being a good technical fit.
Scores are relative
G2 normalizes both scores within a category. A product's number depends on who else is in the category at that moment, and it can differ between categories. A score of 80 in one grid is not the same claim as 80 in another.
This matters when you compare across grids. Two scanners listed in different categories, for example a dependency scanner and a broader application security platform, are measured against different peers.
What to do with a Grid
Use it as a way to build a long list, not a short one. It tells you which products real users write about and how those users felt about support and usability.
Then check the things the Grid cannot see:
- Does it support your languages, package managers and build systems?
- How noisy are its findings on your own repositories?
- Can you export your data if you leave?
- What happens to a finding after it is raised?
A trial on two or three real repositories answers those. Our post on evaluating a dependency scanner in one afternoon shows one way to run it.
G2's documentation carries a note that readers should confirm current details with G2 before relying on it, so check the live page if the exact weighting matters to a decision.
Where Safeguard fits
Safeguard offers SBOM and dependency scanning, reachability analysis, and policies and gates for pull requests. You can run a free scan on your own code at gold.safeguard.sh/scan and judge the output yourself.
Quick check
Does a high Satisfaction score mean a tool finds more vulnerabilities?
No. The inputs G2 describes are survey answers about usability, fit and support. They say nothing direct about detection quality.
Can I compare scores across two different G2 categories?
Not safely. Scores are normalized within each category, so the same number can mean different things.
Sources
Last reviewed: 19 October 2026.
