> ## Documentation Index
> Fetch the complete documentation index at: https://docs.synscribe.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Content production & Definition of Done

> How content is produced each cycle, and the checklist a page must pass before it counts as done.

How the cycle's content gets produced, and the bar every page has to clear before it counts as
shipped. Production turns the tagged keyword plan into live pages; the Definition of Done is the
checklist that keeps quality consistent across every account and whoever's driving.

> **Two source pages merged here:** *Content Production Procedures* and *Definition of Done for
> Content*. Cross-links point to the Part 2 how-tos for each step.

## Content production

Production runs the tagged keyword plan through the generators (landing pages from the landing-page
pillar, blog posts with the account's refiners), then publishes and indexes them. The worked,
step-by-step version lives in the Month-1 flow:

* Landing pages: [3.1 Week 2 — pSEO pages](/sops/onboarding-month-1#week-2-—-publish-&-collect-data)
  and [how-to: 2.3.2 Generate landing pages](/platform/generate-landing-pages).
* Blog posts: generate with the account's refiners (typical defaults: Brand positioning, CTA, FAQ,
  TLDR), starting with listicles or BOFU how-tos. See
  [3.1 Week 1 — prep for blog posts](/sops/onboarding-month-1#day-4/5-—-strategy-kickoff) and
  [how-to: 2.1.7 Set up refiners](/platform/set-up-refiners).

> **Why the format matters:** we produce for two search surfaces, Google and the AI answer engines.
> The mix of what ChatGPT cites shifts over time (listicles, comparisons, how-tos, and other
> formats), so format choice is a real production decision. See
> [Part 1 §1.2 Two search surfaces](/theory/two-search-surfaces). Reference data on
> how citation types shift: [promptwatch.com/data/chatgpt-citation-types-over-time](https://promptwatch.com/data/chatgpt-citation-types-over-time).

### Production order and who does what

Landing pages lead, and they're front-loaded. The first batch is the first 5 landing pages, picked by
the product's sharpest edge and the lowest cannibalization risk; blogs start at 3, listicles or BOFU
how-tos first. Generate the whole cycle's content in Week 1 so the publisher never runs dry partway
through the month, then publish at the account's set cadence (some accounts drip a few pages a day,
others batch a couple of times a week). A standard cycle produces a set number of pages, split between
landing pages and blogs per the client's strategy.
❓ \[needs Raymond / \[DECIDE]: standard-tier output per cycle — pricing rework in progress]

The account manager drives Pi to generate everything, then runs each page's content check and
technical SEO audit. Pi self-audits and flags as it goes; the AM signs off, so the person who
generated the page isn't the one who declares it done. Publishing is a separate handoff: for several
accounts the BPO teammate publishes on the agreed schedule (coordinated in `#bpo`), and where the CMS
belongs to the client (Framer, Shopify) the client reviews and publishes their own landing pages. See
[3.14 Roles & handoffs](/sops/roles-handoffs).

## Definition of Done

The bar is a 15-item must-pass gate. A page (and a first batch) is done only when all 15 must-pass
checks clear. Miss one and the pipeline isn't done: fix it, re-verify, and nothing publishes as done
until it passes. Softer "nice-to-have" items don't block; they get logged for the next batch. The
escalation rules for a repeated gate failure live in [3.6 The Quality Gate](/sops/quality-gate).

Ten of the must-pass checks carry a hard number. This is the quantitative bar:

| Check                                                | Must-pass target                                         |
| ---------------------------------------------------- | -------------------------------------------------------- |
| First-batch post count                               | 3 or more                                                |
| Indexing lag per post                                | Within 48 hours                                          |
| Primary keyword in the meta title                    | Within the first 100 characters                          |
| Meta description length                              | 160 characters or fewer                                  |
| Primary or secondary keyword in the meta description | Within the first 100 characters                          |
| Org mentioned and linked in the summary              | Within the first 100 characters                          |
| Primary keyword in the H1                            | First word of the H1                                     |
| Broken links                                         | Zero                                                     |
| Competitor links in a blog                           | Zero, unless a reference carries a genuinely useful stat |
| Slug matches the primary keyword                     | Exact match, not derived from the title                  |

The rest of the gate is the structural and qualitative checklist below. Pi self-audits and flags;
the AM signs off.

* [ ] Keyword selection: the page targets a validated keyword from the cycle plan (right intent,
  winnable SERP, no cannibalization). See [3.4 Organise keywords](/sops/organise-keywords)
  and [Part 1 §1.4 SERP cliff](/theory/serp-cliff).
* [ ] DOM, everything present: all content elements are DOM-readable (they may be visually hidden on
  hub pages, but must be in the DOM).
* [ ] Heading structure: one H1 with the primary keyword first, and a sensible H2/H3 hierarchy.
* [ ] JSON-LD
  * [ ] Primary structured-data type present for the page kind, e.g. `ProductList` (pSEO / listing
    pages), `BlogPosting` / `Article` (blog posts).
  * [ ] FAQ JSON-LD: good to have, and required on pSEO pages.
* [ ] Qualitative self-audit: keyword intent is BOFU or high-intent MOFU (no pure informational fluff
  in batch 1), each post fits a declared pillar (`Core`, `Fringe`, `Alternative`), brand voice
  passes any regulated-industry banned-term check, product claims trace to the product bible
  rather than a model guess, and client link sensitivities are respected (default: no Reddit links
  unless the client has approved them).

> This checklist is the acceptance bar for the technical SEO audit run at publish time. The
> fully-expanded audit (canonical link, OG image, table of contents, Ahrefs toolbar checks) lives in
> [3.1 Week 2 — technical SEO audit](/sops/onboarding-month-1#week-2-—-publish-&-collect-data) and
> [how-to: 2.5.2](/platform/technical-seo-audit). This page is the concise
> Definition of Done; that page is the full audit procedure.

> Full theory on the Definition of Done: [Part 1 §1.11](/theory/definition-of-done).

***

## Related

* [3.3 — Weekly planning flow](/sops/weekly-planning): where production runs each week.
* [3.6 — The Quality Gate](/sops/quality-gate): the ranking outcome the DoD feeds.
* [how-to: 2.5.2 Technical SEO audit](/platform/technical-seo-audit): the full audit.
* [Part 1 §1.11 Definition of Done](/theory/definition-of-done): the theory.
