Evaluation-led development
We don't
guess.
We grade.
Your site already has a grade, whether or not anyone has shown it to you. See it free across six lenses. Get the fixes ranked by what each one is worth, then re-grade after the work to see what it bought you.
Six lenses
Sample evaluation. These figures show how a Grader report reads. Live scanning of your URL is rolling out.
The Deep Report
Your grade is free.
The fix plan is $50.
The Deep Report ranks every fix by what it's worth and hands you the code to make it. It is not on sale yet. When it launches, the $50 comes back as a credit toward your build.
Watch one happen.
The Grader reads the site, settles on a score and hands back the fixes that matter. Scroll to step through it; on a phone it plays by itself. It is a sample walkthrough, not a live scan of your URL.
Sample scan…
Overall grade
Keep it sharp
Performance 94 → ~98. Shave the last milliseconds.
SEO 96 → ~99. Product pages become eligible for rich results.
Accessibility 92 → ~96. Make keyboard focus easier to see.
Sample walkthrough. Full live measurement with Lighthouse and axe-core is rolling out; until then this shows how an evaluation reads, not a live scan of the URL you enter.
Six lenses.
One grade, with the math shown.
Twenty-six signals, each weighted and tagged with how sure we are, rolled into one score. Every site is graded the same way, so your number means the same thing as your competitor's.
Speed & Performance
How fast the page loads and how quickly it responds when someone taps.
Accessibility (ADA / WCAG)
The WCAG failures that lock customers out and draw demand letters.
Findability (SEO & AEO)
Whether Google and AI answer engines can find you and quote you.
Security & Privacy
TLS, security headers and the trackers you load. The gaps attackers and regulators notice first.
Trust & Compliance
The policies and credibility signals a cautious buyer looks for, framed as business risk.
Reputation
What the web says about you: reviews and sentiment, each labeled by how sure we are.
The same four steps, in the same order, on every site.
Grade the same site twice under the same conditions and you get the same result. That repeatability is what lets us rank fixes by impact rather than by hunch.
Scan
We read your public pages and pull what can be observed. Nothing is copied or stored.
Score
A weighted model turns those signals into grades across 5 axes and 26 metrics, and says how sure it is of each.
Synthesize
The grades become one plain-language read on what is strong and what is at risk. One story instead of 26 numbers.
Recommend
Fixes ranked by impact, each with the score lift it is likely to move. Re-grade after the work and the lift is on the record.
What you get back looks like this.
Every lens scored and labeled measured or estimated, with the fixes in order. Your own site's report uses the same layout.
Overall grade
A solid base with room to climb. Accessibility is the weakest lens, and the quickest, cheapest one to fix.
Lens breakdown
Fix these first, in this order
A grade tells you where you stand today.
The Sentinel tells you when that changes.
Certificates lapse, software drifts, rivals rebuild and AI answers change, and nobody emails you about any of it. The Sentinel is our monitoring agent for that job, priced but not on sale yet. It would re-grade your site on a schedule, grade up to three competitors beside yours, track who's talking about you, and email you when a lens drops.
Know what to fix first,
and exactly how.
The free grade names what's holding your site back. The Deep Report hands you the fixes: copy-paste code, a dated roadmap and the tools to do the work. You decide whether that's worth $50 after you've seen your grade.
Every finding is labeled measured or estimated. We never promise a ranking, because a report that flatters you is worth less than one you can trust. If it says measured, we saw it.
Measure first. Fix in order. Prove it after.
An agency sells you taste. A free scanner hands you a wall of warnings with no order to them. We grade the site before anyone quotes you a price, rank the fixes by what they're worth to you, then grade it again on the same rubric, so the improvement is on the record and not in a slide deck.