- Home
- How we rank
Methodology
How every product gets its score
Six criteria per category. A cited scorecard for every product. Penalties in the open. A rank set by the score and nothing else.
The six criteria
Four apply to every category. Two are chosen for the category, because a payroll tool is not judged like a design tool.
An 8.5 in CRM equals an 8.5 in cybersecurity.
Scores are normalised inside the category. They only compare products on the same page.
A real scorecard
Live from the scoring tables. This is what sits under every product on a ranking page.
Open the ranking
Klaviyo
Marketing Automation Platforms for Marketing Agencies · scored Apr 2026 · klaviyo.comHow to read it
- Amber bar = the strongest criterion. Red = the weakest.
- Each criterion carries what was looked for, what was found, and its sources.
- Click a source chip to see what was taken from it before you leave the page.
Penalties on this product
Penalties
27,900+ documented problems, each with a severity and a source.
| Severity | Tier | Typical example |
|---|---|---|
| 0.00 to 0.10 | Negligible | A missing minor integration |
| 0.11 to 0.25 | Minor | Limited chat hours |
| 0.26 to 0.45 | Moderate | No phone support on standard plans |
| 0.46 to 0.70 | Significant | Auto-upgrades that raise the bill |
| 0.71 to 1.00 | Major | Billing for contacts you never use |
Penalties are subtracted, not hidden. A product can score high on capability and still lose ground for a billing policy or a support gap. Each penalty names the criterion it touches and links to the evidence.
They are the reason two products with similar feature depth can sit a full point apart.
Two tiers of research
Nobody credibly hands-on tests 8,000 products. The site says which treatment each product got.
Deep Lab
Category leaders and finalists · a minority of products
- Live account, real workflow
- Support response tested
- Native vs middleware integrations checked
- Screenshots and walkthrough kept
Forensic Spec
Every other product · the default
- True monthly cost: seats plus required add-ons
- Security and compliance from the legal pages
- Hidden limits: API caps, storage, contacts
- Standard spec matrix for side-by-side comparison
Workflow fit
A 9 for a five-person agency can be a 4 for a company with 500 reps. Weights follow who the category is for.
This is why the same product can hold different ranks in different categories.
Numbers we compute
"From $19 a month" tells you nothing. These do.
- Price to seat
- Per-user and flat fees standardised to the cost of a five-person team.
- Integration weight
- 500 shallow connectors do not beat 50 native two-way syncs. Depth counts.
- Skip if
- The specific buyer who should avoid the tool. Shown on every product.
- Support to price
- Whether the support you get (email, chat, phone) matches what the tier costs.
True monthly cost, example
Independence
Plainly.
Limits
Where the method has edges.
- Private features and internal processes vendors do not disclose cannot be scored.
- Adoption signals can lag, especially in fast-moving categories.
- Two category-specific criteria cannot capture every nuance of your own use case.
- Most products get the Forensic Spec audit, not a hands-on test.
- Scores are a starting point for comparison, not a final verdict. Trial the shortlist yourself.
Questions people ask
What does a score mean?
A score out of 10 that only compares products inside the same category. Six criteria, weighted for that category, each backed by cited evidence. An 8.5 in CRM is not the same as an 8.5 in cybersecurity.
Can a vendor pay for a higher rank?
No. Rank is set by the overall score. Affiliate commissions never touch the order. Sponsored rows, where they exist, are labelled and sit outside the organic ranking.
How are penalties applied?
Documented problems such as surprise billing or missing support channels are recorded with a severity from 0 to 1 and an evidence link, and they pull the overall score down. Every penalty is shown on the product.
Is the research done by AI?
Research and drafting use AI tools against public documentation, pricing pages and reviews. Criteria, weights and category decisions are set by a person. Anything marked hands-on was tested by a person.
Builds the scoring pipeline, sets the criteria for each category, and audits the data behind every ranking. Runs the hands-on tests personally.
About the site