Sample report · unredacted

This is the whole report.
Here's how to read it.

A real score on a real listing — nothing redacted, nothing held back for a demo. Every check below shows its working: what it looked for, what it found, and what would have earned the points.

Score my listing
Report provenance computed live
Listing
Acme Observe
Category
Monitoring · Observability
Input
Pasted copy · 114 words
Checks run
16 across 7 dimensions
Raw points
51 / 330
Scorer
v2 · on page load

Scored on Acme Observe, a fictional observability listing written the way most listings are written.

39
Out of 100
Weak

Below the category leaders — start with Buyer fit.

Scored on your listing copy across seven buyer dimensions — mostly on what a buyer could actually check (numbers, named integrations, certifications, stated pricing), and partly on structure. Backlink authority, AI visibility, and rich-media agent legibility are not measured here.

0
checkable claims
10
superlatives
114
words of copy
94
structure score

How to read this

Look at the last two numbers together. This listing scores 94/100 on structure — every field is filled in, at a sensible length, with paragraphs and bullets. By the old measure it was a great listing. But it contains 0 claims a buyer could check and 10 superlatives. That gap between "well-formed" and "says anything" is the whole report.

The seven buyer dimensions

Don't read the bars as a report card — read them as a work order. Hover a row to isolate it; open one to see the checks that produced it.

Opening copy is plain, not hype 0/22
Says plainly what the product is 18/18
Names who the product is for 0/20
Names the systems it works with 0/22
States how it deploys 0/14
Highlights carry specifics 0/14
Claims a buyer can check 0/34
Enough proof to carry the page 0/16
Title names the category 18/18
Copy is legible to an AI assistant 0/26
Description is scannable 10/10
Named certifications and security posture 0/32
Support is a commitment, not a mention 5/20
Pricing is on the page 0/26
A buyer can start without talking to sales 0/20
No "contact us" gate on the purchase path 0/18

How to read this

The lowest bar is where a buyer stops reading, and the dimensions are weighted — a low Differentiation costs more than a low SEO. A dimension marked "not measured" is not a zero: it means we genuinely could not see it from public copy, and we would rather say so than guess.

What this report cannot see

Here is what is genuinely outside what public listing copy can tell us:

  • Backlink authority — how much domain trust points at your listing
  • AI visibility — whether assistants actually surface you in a shortlist
  • Rich-media legibility — whether your video has a transcript an AI can read

The score above is calculated only from the dimensions we can actually read. Nothing here is estimated to fill a gap.

Every check, and what it found

This is the full working. Each row is one check, the points it earned, and the exact thing in the copy that earned or lost them. Open a row to see what would have scored.

51/330
raw points, weighted to 39/100
Check
Dimension
Points
What the check looked for

Sentences carrying a number, a named system, a certification, or a stated price — anything a buyer could independently verify — measured against the count of unbacked adjectives.

What would earn the points

Nothing in this listing can be verified. Adjectives like "revolutionary", "best-in-class", "powerful" +7 more describe every competitor equally, so they carry no information. Replace each one with the number behind it, the certification that proves it, or the system it integrates with. This is the single highest-leverage change on the page.

What the check looked for

Density, not volume — checkable claims per 100 words, so a long listing cannot pass on length alone.

What would earn the points

The proof that is here is thin against the length of the copy — a buyer reads several paragraphs before hitting anything concrete. Aim for roughly one verifiable fact every two or three sentences.

What the check looked for

The first sentences a buyer reads, scanned against a superlative list every competitor also uses.

What would earn the points

The first thing a buyer reads is adjectives, not what the product is. Superlatives are the fastest way to read as undifferentiated — every competitor uses the same ones. Say what it does in the first sentence and let the numbers carry the persuasion.

What the check looked for

A single sentence that names the product and its category in plain terms — the "X is a Y for Z" an assistant extracts verbatim.

Why it scored

Nothing in the copy says "X is a Y for Z" outright. Human buyers skim for it and AI assistants extract it verbatim when they summarise you in a shortlist — if it is not there, both have to guess.

What the check looked for

A named role, team, industry, or company shape — not "modern teams".

What would earn the points

The copy never says who this is for. "Modern teams" is not an audience. Name the role, team, or industry — a buyer self-qualifies in seconds, and an AI assistant needs it to decide whether to surface you for a given query.

What the check looked for

Named products and services — AWS services, data stores, CI systems, identity providers — rather than "your existing stack".

What would earn the points

A buyer is asking "does this fit my stack?" and the copy does not answer it. Name the actual services — AWS Lambda, EKS, Snowflake, Okta — not "your existing tools". Named integrations are also one of the few claims a buying agent can verify on its own.

What the check looked for

A stated delivery model — SaaS, in-VPC, self-hosted, AMI, container — plus the prerequisites a buyer needs before day one.

What would earn the points

SaaS or AMI? Agent or agentless? In their VPC or yours? A buyer cannot get budget for something they cannot describe to their architect, and this is the first question procurement asks.

What the check looked for

Each marketplace highlight field, checked individually for a number, a named system, or a concrete capability.

What would earn the points

Highlights are the most-read part of a listing and these are adjectives. Each one should carry a number, a named system, or a concrete capability — "Cut p99 latency by 40%" instead of "Best-in-class performance".

What the check looked for

The title field, matched against the category terms buyers actually search in Marketplace.

Why it scored

The title has no term anyone would actually search. Buyers filter by category and AI assistants match on it — a brand name alone is invisible to both. Lead with what it is ("API Gateway", "Data Warehouse"), then differentiate.

What the check looked for

Sentence-level self-containment — an assistant quotes sentences, not paragraphs, so each one has to stand alone with a fact in it.

What would earn the points

When an assistant builds a shortlist it lifts individual sentences out of context. Sentences that need the paragraph around them to mean anything get dropped. Most of this copy is that kind of sentence — rewrite so each one stands alone with a fact in it.

What the check looked for

Structural markup in the long description — bullets, numbering, short sections.

Why it scored

The description is an unbroken block. Buyers skim and parsers segment on structure — bullets and short sections make both work.

What the check looked for

Named certifications, security controls, and AWS validations — matched against a known register, not the word "secure".

What would earn the points

Nothing here tells a buyer their security review will pass. Name the certifications you hold (SOC 2, ISO 27001, HIPAA) and the posture behind them (encryption, SSO, audit logging). Security review is where marketplace deals die, and this is the field that survives it.

What the check looked for

Three parts — channel, response time, and coverage window. A channel on its own earns partial credit.

What would earn the points

"Email support is available" is not a commitment — it is a sentence. State the channels, the coverage window, and a response time a buyer could hold you to. A number here is worth more than a paragraph of reassurance.

What the check looked for

Any figure or unit a buyer could budget against, anywhere in the listing copy.

What would earn the points

There is no price on the page. Buyers who cannot price you compare you against vendors they can, and an AI assistant building a shortlist will simply omit you — it has nothing to put in the price column. Even a starting rate or a unit ("$0.10 per 1,000 requests") is enough.

What the check looked for

A stated free tier, trial, or self-serve plan that a buyer can act on unaccompanied.

What would earn the points

Every route into this product goes through a person. A free tier, a trial, or a pay-as-you-go subscription lets a buyer evaluate you on a Friday afternoon — which is when most marketplace evaluation actually happens.

What the check looked for

Gate phrases sitting on the purchase path, where a price or a start button should be.

What would earn the points

The copy routes the buyer to a demo or a quote instead of a price. That is a dead end for a self-serve buyer and a hard stop for a buying agent — it cannot book a meeting. Keep the demo offer, but put a price beside it.

Diagnostic · not scored

How far in the first checkable claim sits

No checkable claim anywhere in the copy

A retrieval bot fetching this page gets a finite budget and spends it in reading order, so everything ahead of the first checkable claim is what it buys first. This measures position, not length. Tooling vendors sell token efficiency as a saving, and it is the model provider's saving — the token thing is not your bill, it is your odds.

There is no claim here to reach at any depth. That is the same finding as the zero above, read from a different angle: a bot that read every character of this listing would still leave with nothing it could quote.

What to fix first

Ranked by how much the gap costs, not by how easy it is. Do them in this order.

01

No checkable claims — replace the adjectives with numbers

High impact

Nothing in this listing can be verified. Adjectives like "revolutionary", "best-in-class", "powerful" +7 more describe every competitor equally, so they carry no information. Replace each one with the number behind it, the certification that proves it, or the system it integrates with. This is the single highest-leverage change on the page.

Points in play
+34
in Differentiation
02

Name your certifications and security posture

High impact

Nothing here tells a buyer their security review will pass. Name the certifications you hold (SOC 2, ISO 27001, HIPAA) and the posture behind them (encryption, SSO, audit logging). Security review is where marketplace deals die, and this is the field that survives it.

Points in play
+32
in Trust signals
03

Put a number on the pricing

High impact

There is no price on the page. Buyers who cannot price you compare you against vendors they can, and an AI assistant building a shortlist will simply omit you — it has nothing to put in the price column. Even a starting rate or a unit ("$0.10 per 1,000 requests") is enough.

Points in play
+26
in Conversion

The walkthrough is being recorded

Ron is cutting a short video that reads this exact report line by line — what each dimension measures, why this listing scored what it did, and which fix he'd make first. Until it's up, every part of the report above is annotated.

The same listing, rewritten

A different listing in the same category, written with numbers instead of adjectives. Same scorer, same checks, run at the same moment. Switch between them.

39
Out of 100
0
checkable claims
10
superlatives
Adjectives — the copy that was scored
“The innovative, best-in-class monitoring, logging, and tracing solution for modern engineering teams.”

no checkable claim to reach

Well-formed and completely unfalsifiable. Every competitor in the category could publish this sentence unchanged, which is what makes it worth zero points.

Proof — the copy that was scored
“Cut p99 latency by 40% at the edge. SOC 2 and ISO 27001 certified, with a free tier and usage-based pricing.”

first checkable claim at character 55

53 points separates them, and both are about the same length. The difference is that one says "40%", "SOC 2 Type II", and "$0" — things a buyer can check and an AI assistant can quote.

Now run yours.

Paste your listing URL, or paste the copy straight in. It takes about fifteen seconds and you get the same report you just read — same checks, same working.

Score my listing No email. No signup. Nothing is saved against your name.
16
checks, every one shown with its working
7
buyer dimensions, weighted
~15s
from paste to full report
0
fields required before you see it