The rubric
Every teardown on this site resolves to one composite score out of 5, built from six sub-scores weighted equally. This page defines all six, states what each band means, and sets out what counts as evidence.
It exists so that a score can be argued with. If you disagree with a number, the disagreement should be about the evidence or about the band definition, not about what the number was supposed to mean.
What this scores
Six dimensions, each on a 1 to 5 band. Together they describe whether a publication is built to keep what it earns.
- Whether the publisher owns the ground the work sits on
- Whether the work can be found
- Whether it is well made to read
- Whether there is a surface for value to land on
- Whether attention becomes a relationship
- Whether the archive gains value as it grows
What this does not score
These are deliberately excluded. They are either not measurable from public evidence, or not the thing this rubric is for.
- Traffic volume. A large audience badly served scores worse than a small one served well
- Revenue. Not verifiable from outside, and not the same as monetization design
- Subject matter. Niche, topic and expertise are out of scope
- Taste. Whether the writing is good is not a rubric question
- Social following. Rented audience on a rented platform
Six scores, one number
The composite is the unweighted mean of the six sub-scores, rounded to one decimal place. No dimension counts for more than any other.
Equal weighting is a choice, not a finding. Weighting the dimensions would require evidence about which one moves outcomes most, and that evidence does not exist yet. Until it does, equal weighting is the honest default and the arithmetic stays checkable.
Owned hub strength
Does the publisher own the ground the work sits on?
Admissible evidence. Domain control, whether the archive is served from the owned domain, whether the subscriber list can be exported in full, and whether analytics are independent of any single platform.
| Band | What it means |
|---|---|
| 1 | No owned domain. Archive, list and analytics all live on a third-party platform and none of them can leave intact. |
| 2 | Owned domain pointed at a hosted platform. The archive is not portable and the platform mediates delivery to the list. |
| 3 | Owned domain and portable archive. The list can be exported. At least one critical function remains platform-dependent. |
| 4 | Domain, archive and list are all owned and portable. Analytics do not depend on the publishing platform. |
| 5 | All of the above, and the archive can be rebuilt and rehosted from source without losing a single URL. |
Discovery
Can people and machines find the work without being told it exists?
Admissible evidence. Feed presence and format, sitemap validity and freshness, whether the archive is crawlable without JavaScript, structured data, and whether topic routes exist above the level of individual posts.
| Band | What it means |
|---|---|
| 1 | No feed, no sitemap, and an archive that cannot be crawled without executing JavaScript. |
| 2 | One of feed or sitemap is present. The other is missing, invalid or stale. |
| 3 | Feed and sitemap are both present and valid. There is no hub or topic layer above individual posts. |
| 4 | Feed, sitemap, and a maintained hub layer covering the subjects that recur most often. |
| 5 | All of the above, plus structured data on hub pages and full-text rather than excerpt feeds. |
Craft
Is the work well made to read?
Admissible evidence. Measured load performance, consistency of the type scale across templates, line length, the mobile reading experience specifically, and whether the markup is semantic.
| Band | What it means |
|---|---|
| 1 | Slow, inconsistent, or genuinely hard to read on at least one common breakpoint. |
| 2 | Readable on desktop and materially worse on mobile, or with measurable performance problems on primary templates. |
| 3 | Consistent and readable across breakpoints, with performance around commonly cited thresholds. |
| 4 | A deliberate type scale, a controlled measure, and good measured performance on every template a reader meets. |
| 5 | All of the above, and the reading experience is itself one of the reasons people come back. |
Monetization design
Is there a defined surface for value to land on?
Admissible evidence. Whether inventory or a product exists at all, whether it is defined and priced, whether pricing is public, whether disclosure is present, and whether any of it interrupts reading.
| Band | What it means |
|---|---|
| 1 | No commercial surface of any kind. There is nowhere for existing demand to land. |
| 2 | A surface exists but is undefined or unpriced, so every transaction restarts from zero. |
| 3 | Inventory or a product is defined and priced, and sold case by case. |
| 4 | Defined, priced, published, and sold without interrupting the reading experience. |
| 5 | All of the above, packaged rather than sold per placement, with disclosure and a stated policy. |
Distribution loop
Does attention convert into a relationship the publisher controls?
Admissible evidence. Whether a capture unit exists and where it sits relative to reader intent, whether sends are authenticated, whether they are delivered, and whether the loop returns readers to the site.
| Band | What it means |
|---|---|
| 1 | No capture anywhere. A reader who finishes a post has no offered next step. |
| 2 | Capture exists in one low-intent location only, typically a footer, on every template equally. |
| 3 | Capture is present on the primary templates. Sends exist and are properly authenticated. |
| 4 | Capture is differentiated by page intent, sends are authenticated and delivered, and the loop returns readers to the site. |
| 5 | All of the above, and the loop is measured and tuned rather than assumed. |
Compounding
Does each new post make the existing ones worth more?
Admissible evidence. Internal link density and, more importantly, link direction. Whether hubs are maintained or generated. Archive depth. URL stability over time.
| Band | What it means |
|---|---|
| 1 | Write-once publishing. No structure, no internal links, and no guarantee that URLs survive. |
| 2 | Posts link sideways to each other. The archive is flat and there is nothing for links to point at. |
| 3 | Some structure exists. Internal links are present but are not routed through it. |
| 4 | Posts route into maintained hubs and URLs are stable over time. |
| 5 | All of the above, and the archive demonstrably earns links and traffic independent of new publishing. |
The evidence standard
Every score is produced from public evidence only. No analytics access, no interviews, no privileged data, and nothing a subject tells us in confidence. If a figure cannot be reproduced by a reader with a browser, it does not appear in a teardown.
Where a dimension cannot be assessed from public evidence, it is scored at the midpoint and flagged in the post rather than guessed at.
Every teardown records the rubric version it was scored against. Scores produced under different versions are not directly comparable, and any change to a band definition is published with the version that introduced it.
Right of reply. Subjects are contacted before publication and given seven days to respond. Any reply is appended in full and unedited. A subject that ships the fixes can request a re-score at no cost, and the result is published as an update rather than as a new post.