Score the page
Seven dimensions, a hundred points, a gate on each. Stop guessing whether the page is good — measure it.
Your job: replace the argument about whether the page is good with a number.
Assess my landing page Here's my landing page: [paste the URL or the copy]. Score it on the seven dimensions — clarity, targeting, problem, unique value, proof, offer, story — gate each at 7/10, and name the one dimension to fix first. claude mcp add --transport http leanspark https://leanspark.ai/mcp Connect first →
What you get back
Fetches your live page and scores it across the seven dimensions, flags every one below the gate, and names the single dimension to fix first — so you spend your next cycle on the thing that's actually costing you conversion.
A 0–100 score with a per-dimension breakdown and the weakest dimension named. You decide what to change.
You can argue about a landing page forever. A scorecard ends the argument. The Landing Page Assessment reads your live page and scores seven dimensions, each out of ten, each with a gate it has to clear:
- Clarity & Positioning — is it instantly obvious what this does and who it’s for?
- Targeting — is the audience specific, and does the language match it?
- Problem & Insight — is the problem articulated with a real, earned insight?
- Unique Value — is the value proposition differentiated and specific?
- Proof & Credibility — are there outcomes, authority, and social proof?
- Offer & Objections — is the CTA clear, the friction low, the objections handled?
- Story & Flow — does the sequence carry a reader from top to bottom?
The seven roll up to a score out of 100, with bands — excellent (85+), good (70–84), needs work (50–69), poor (under 50). The gate matters as much as the total: a page can look fine and still fail on the one dimension that’s quietly costing you the conversion.
A real example: the LEANSpark homepage scored 77. “Good” — all seven cleared the gate. But two dimensions only just did: Targeting (the segment was “founders,” too broad) and Proof (one anecdote, no stacked outcomes). That’s the value of the gate — it doesn’t let a strong average hide the two weak rooms. The fix list writes itself: name the segment, stack the proof.