MARSTIVO SCORE v1 · Updated 23 October 2026
Useful software recommendations need evidence, context and clear limits. This page sets the standard for future MARSTIVO reviews and comparisons. No product testing results or scores have been published yet.
How we evaluate software
We start with a real user, a practical workflow and realistic alternatives. Where practical, we use the software ourselves, record repeatable tasks and explain what happened. Research supports the evaluation: we check official documentation, pricing, plan limits, security information and relevant third-party evidence. Research-only coverage is labelled and is not presented as a hands-on review.
AI may assist research, organisation and drafting. A published factual claim still needs verification, and AI-generated text is not evidence that a task was tested. We never fabricate experience, testing results or scores.
What each review should disclose
The reviewer and publication/update dates; the dates of testing; the software version, deployment and plan where relevant; the tasks and integrations actually tested; the device or environment where it affects results; and important exclusions or limitations. Free access, review copies, sponsorship and affiliate relationships are disclosed when applicable.
A short test is a snapshot, not a long-term uptime study. We do not imply an independent security audit, penetration test or privacy guarantee from reading a vendor policy. Untested claims are attributed to their source. If access or evidence is insufficient, we say so and withhold a numerical score.
MARSTIVO Score v1
Each criterion is scored from 0 to 10. The final score is the weighted average below; weights total 100%. We retain calculation precision and round the final result once to one decimal place. The display is “MARSTIVO Score — X.X/10”. Scores summarise an evaluation for a stated use case; they are not universal guarantees.
| Criterion | Weight | What we assess |
|---|---|---|
| Ease of use | 15% | Onboarding, clarity, task effort and accessibility in the tested workflow. |
| Features & capabilities | 20% | Depth, usefulness and limitations of relevant features. |
| Reliability & performance | 10% | Observed failures, consistency, responsiveness and task completion. |
| Integrations/ecosystem | 10% | Relevant integrations, API access, extensibility and portability. |
| Pricing & value | 15% | Cost at realistic usage, billing terms, limits and alternatives. |
| Documentation & support | 10% | Accuracy and usefulness of documentation; support only if actually tested. |
| Privacy & security | 5% | Documented controls, permissions, data handling and evidence limitations. |
| Category-specific capability | 15% | The core capability for the software category, defined before testing. |
Formula: (Ease × 0.15) + (Features × 0.20) + (Reliability × 0.10) + (Integrations × 0.10) + (Value × 0.15) + (Documentation × 0.10) + (Privacy × 0.05) + (Category capability × 0.15).
Use the same 0–10 scale for every criterion: 0 means the capability is absent or unusable for the defined task; 2 indicates severe limitations; 5 means adequate with material compromises; 8 means strong with limited drawbacks; 10 means exceptional, well-supported performance against the defined requirements. Intermediate values need a written rationale. A 10 does not mean perfect or risk-free.
All eight criteria need sufficient evidence before an overall score is published. We do not silently remove a criterion or redistribute its weight. Where a criterion cannot be evaluated, mark it “not assessed”, explain why and withhold the overall score. Published reviews should show the criterion breakdown and rationale so readers can assess the total.
The category-specific criterion
| Software category | Criterion | Example evaluation focus |
|---|---|---|
| Developer software | Developer Experience | Setup, APIs/SDKs, debugging, local workflows and deployment. |
| AI software | AI Quality/Capability | Task-specific output quality, consistency, failure modes and controls. |
| Automation software | Automation Flexibility | Branching, transformations, error recovery and maintainability. |
| Productivity software | Workflow/Productivity effectiveness | Whether it improves the defined task without excessive friction. |
| General SaaS | Core workflow effectiveness | Fit and completeness for the primary business workflow. |
We choose the primary category and define the tasks before scoring. The category criterion measures the core result; general features measure breadth and supporting capabilities. This helps avoid counting the same benefit twice. Comparisons use the same rubric, equivalent tasks and clearly stated plan and usage assumptions.
Pricing, alternatives and editorial independence
We check pricing dates, currencies, billing periods, usage assumptions, free-plan restrictions and any relevant add-on costs. We compare realistic alternatives and explain who should choose something else. Vendor prices and features can change; confirm important details before buying.
Affiliate availability, commission rates, free access or sponsorship do not determine MARSTIVO scores or recommendations. Commercial arrangements do not buy a higher score. See our Affiliate Disclosure for the policy that applies when monetized links are used.
Updates and corrections
Scores can change when software, pricing or our evidence materially changes. Material reassessments should state the updated test date, affected plan/version and reason for the score change. Changes to the rubric will receive a new version; scores from different rubric versions may not be directly comparable.
If a claim cannot be supported, we correct, qualify or remove it. Review conclusions should make uncertainty visible rather than conceal it.