Five axes decide whether a visitor converts.
Convexa does not grade taste. It grades the five things that move a conversion rate, the same factors a studio checks before it ships a page. Each axis is scored independently from 0 to 100, then weighted by how much it actually swings outcomes, so a beautiful page with a buried call to action still reads as a low grade. The weights are fixed and public, listed in the right-hand column, so the composite is reproducible rather than a matter of opinion. Read the legend as the specification sheet for the instrument: this is exactly what the needle responds to, and nothing else.
How the conversion score is calculated.
The composite is a weighted mean of the five axes: CTA clarity counts for 25 points, the above-the-fold offer for 25, friction for 20, trust signals for 15, and load speed for 15. Each axis is scored from 0 to 100, combined into a single 0 to 100 composite, and mapped to a letter grade where 90 and up is an A, 80 a B, 70 a C, 60 a D, and anything lower an F. The same URL always returns the same reading, because the score is deterministic rather than sampled. Convexa is a demonstration grader: it scores from the URL and public signals, not from your analytics.
composite = (CTA×25 + Offer×25 + Friction×20 + Trust×15 + Speed×15) / 100
- Load the specimen. Paste any URL into the load tray and run the audit. Nothing is stored, and there is no account.
- The instrument probes six signals. Above-the-fold capture, CTA hierarchy, friction index, trust scan, a speed proxy, and host resolution all read in sequence.
- Read the grade and the fix list. The needle settles on a composite, each axis returns a plain-English finding, and the fixes are ordered by impact, lowest axis first.
A reading beats an opinion.
Most teams audit conversions one of three ways. Two of them are guesses.
| Capability | Eyeballing it | Generic CRO checklist | Convexa |
|---|---|---|---|
| Objective 0 to 100 score | No | No | Yes |
| Five-axis breakdown | No | Partial | Yes |
| Plain-English fixes | Maybe | No | Yes |
| Repeatable result | No | Yes | Yes |
| Prioritized by impact | No | No | Yes |
| Price | Free | Free | Free |
What the instrument actually checks.
Each probe is narrow on purpose. A conversion grader earns trust by measuring concrete, observable things and reporting them in units, not by hand-waving about best practices. These are the five readings the composite is built from, and the order Convexa fixes them in.
CTA hierarchy
Counts the competing calls to action above the fold and scores how clearly the single primary action stands apart from everything else on the page.
Above-the-fold offer
Reads whether the value proposition and the offer are legible before any scroll, in the first five seconds.
Friction index
Estimates the steps, fields, and decisions between intent and the click that completes the action.
Trust scan
Looks for the proof a buyer needs: secure transport, social proof, and credible specifics over adjectives.
Speed proxy
Approximates load behavior, since a page that arrives late converts late, or not at all.
Questions, answered.
Is Convexa free to use?
Yes. Convexa is free, with no signup and no install. It is a demonstration grader you can run on any URL to see how its conversion design holds up.
How does the Convexa conversion score work?
The score is a weighted mean of five axes: CTA clarity 25, offer 25, friction 20, trust 15, and speed 15. Each is scored 0 to 100 and mapped to one composite grade. The same URL always returns the same reading.
How is a conversion grader different from analytics?
Analytics tells you what visitors did. A grader diagnoses why a page underperforms and ranks the fixes by impact, before you have enough traffic to A/B test.
I have my score. What should I do with it?
The score is a diagnosis, not a cure. Fix the lowest axis first. Shape Meets Form rebuilds your highest-leverage page to score in the 90s.