Configure the editorial boundaries Publisher applies to briefs, section plans and long-form writing.
Publisher’s built-in system instructions and quality safeguards remain authoritative. Additional instructions are additive; they should describe brand, reader and format needs rather than attempt to disable safety, source or quality behaviour.
Core fields
| Field | Default/values | Effect |
|---|---|---|
| Final post status | Draft by default; Draft, Pending review or Publish | Controls the normal result outside the Scheduler’s own workflow rules |
| Author | Default 0/automatic | Uses the selected eligible author or an eligible fallback |
| Comments | Default, Open or Closed | Inherits WordPress default or sets the generated post explicitly |
| Tags | Auto, Existing only or Disabled | Controls automatic tag assignment |
| Tag count | Default 5 | Maximum target for automatic tag selection |
| Minimum words | 1,000–10,000; default 1,800 | Lower target boundary and a quality input |
| Maximum words | 1,000–10,000; default 2,800 | Upper planning boundary; keep it at or above the minimum |
| Brand voice | Free text | Describes tone, point of view and reader relationship |
| Additional instructions | Free text | Adds stable editorial requirements |
Write a useful brand voice
Describe observable choices: sentence density, level of technical detail, point of view, treatment of uncertainty, vocabulary and how calls to action should sound. Avoid vague labels such as “professional and engaging” without examples.
Example: “Write for experienced WordPress site owners. Use direct second person, short paragraphs and precise UI paths. Explain trade-offs before recommendations. Avoid hype, invented statistics and guaranteed outcomes.”
Use additional instructions well
- Include required terminology, audience assumptions, units and region-specific spelling.
- State claims that must be qualified and topics that require a source.
- Specify desired practical elements, such as examples, checklists or decision tables.
- Do not paste one article’s complete outline into a site-wide field.
- Do not repeat conflicting length, link or image requirements that already have dedicated settings.
Choosing 1,000–10,000 words
Use 1,000–1,800 words for a narrow operational answer, 1,800–3,000 for a normal comprehensive guide, 3,000–5,000 for multi-step cornerstone content and longer ranges only when the subject genuinely supports distinct sections. More words do not create more value automatically.
Publisher plans adaptive sections and checks the result against the configured range. Provider response limits, topic breadth and quality constraints still matter; do not request 10,000 words from a topic that supports only a short answer.
How Publisher varies structure
For each job Publisher selects an article archetype and a combination of content modules, then creates a topic-specific brief and section plan. Defaults target at least 7 and at most 10 modules. Recent structural signatures help reduce repeated layouts across articles. Variation does not mean randomness without purpose: required headings, evidence and quality rules still apply.
Recommended configuration sequence
- Choose Draft and an eligible author.
- Set comments and tags according to the site policy.
- Choose a realistic word range.
- Write a concise, testable brand voice.
- Add stable editorial constraints.
- Keep the quality threshold and module range at defaults for the first comparison.
- Generate two different topics and compare voice consistency and structural variety.
Automatic vs administrator-controlled
You set the boundaries. Publisher determines the brief, archetype, modules, headings and section order for each topic. There is no separate brief-approval screen in 3.7.6; control occurs at topic approval, final workflow and post review.
Common problems
- Every article sounds identical: replace generic voice adjectives with observable rules and avoid over-prescriptive templates.
- Structure repeats: check whether additional instructions force the same outline; keep module range broad enough.
- Article misses length: inspect topic scope, provider limits and quality report before widening the range.
- Conflicting output: remove duplicate instructions that oppose dedicated link, image, comment or status settings.
- Wrong author/status: verify the selected user’s capability and the Scheduler workflow.
Verification checklist
- Voice is recognisable across two unrelated test topics.
- Claims remain qualified and evidence-led.
- Each article has a topic-appropriate structure rather than a copied template.
- Word count sits within a defensible range.
- Native post attributes match the policy.