Content about content, written by the product
Every post here was drafted by a RankMill agent and edited by a human. If you like how it reads, that’s the point.
Introducing RankMill
Agents that learn your business, suggest your topics, write the drafts in your voice, and hand you markdown you would actually publish.
Two posts that disagree: running a project audit before a batch goes out
What the project audit reads, the three kinds of problems it writes down, how each finding becomes a fix, and when to run one before your next batch of posts goes out.
Declining a topic with a reason: teaching the topic hunt what your blog is not
Why every decline reason reaches later topic hunts, how to write one that steers future suggestions, when to fix the website profile instead, and how declining keeps the 20-post limit from blocking your hunts.
Letting the calendar fill itself: publish cadence, the 24-hour fallback and human in the loop
How the publish cadence keeps your calendar two weeks full, when a topic on a slot gets written, what the 24-hour fallback spends, and what Human in the loop changes.
Answering a topic's questions: what happens to your answers, and when to leave one blank
What the questions on a suggested topic are, how your answers reach the writer as facts, what happens when you leave one blank, and when to switch on Human in the loop.
Rewriting a post that already ranks: importing it into RankMill and making it your own
How to import a post by pasting it or giving its URL, when to pick As is, Restructure or Rewrite, what a rewrite keeps and changes, and what to check before the new version goes out.
A frontmatter block that fits your static site, built once per project
Set a project's frontmatter list once, pick the right kind for each field, and every downloaded post drops into your Jekyll, Hugo, Astro or Next.js repo with no hand edits.
One writer, five clients: a solo agency operator's weekly blog workflow with agents
A weekly routine for running five client blogs alone: research on Monday, drafting midweek, approval and export on Friday, and the setup that keeps each client's voice rules from leaking into another's posts.
The banned words list: a small lever for killing AI slop in your blog
A starter list of common AI tells, how to add the offenders from your own field, and a simple loop for keeping the list up to date as new ones slip through.
Why post-based pricing beats per-seat for small content teams
Per-seat tools charge small content teams for who can log in, not for what they publish. A worksheet for your true cost per post, and where each pricing model breaks down.
Answer engine optimization: writing for AI search without losing Google
How ChatGPT, Perplexity and Google AI Overviews pick the pages they cite, the structure, passage, citation and phrasing habits that get a post quoted, and a pre-publish checklist that keeps your Google rankings intact.
Reading the agent log: what a good research run looks like before you approve it
What you can actually read after a topic hunt, the signs of a solid topic and a shallow one, and a five-check routine to run before you approve a topic.
Editing an AI draft by conversation: prompts that get real revisions instead of reworded fluff
Sort every edit prompt into structure, argument, evidence or tone, then name the target and the new state, and the editor agent hands back a real revision instead of a reworded paragraph.
The four-agent loop, mapped: what research, writer, editor, and human each own
How RankMill splits a blog across a research agent, a writer agent and an editor agent, the two approval gates a human owns, and what passes between each stage.
Writing the research prompt: how to tell your research agent what a ranking opportunity actually looks like
The three levers that shape what a research agent returns: intent filters, SERP shape rules and a short competitor allow-list, plus the hard rules a topic list must obey.
The content brief is the prompt: what to hand an AI writer before it starts
The five inputs that change an AI writer's draft the most, why a thin brief produces a generic post, and a fill-in brief you can copy into any agent.
The banned words list is not enough: configuring brand voice rules that survive a full draft
Why a banned words list cannot make a draft sound like you, the four kinds of positive writing rule that can, and a before and after on the same topic.
The approval checkpoint checklist: what to actually look at when you sign off on an AI post
A two-pass checklist for the human approval step: skim for title, voice, layout and intent, then read every claim, number, name, quote and link line by line before you approve.
Turning a topic hunt into a month of posts without a plan doc
Run a topic hunt, decline what does not fit, answer the questions that matter, and let the publish cadence fill the next two weeks of slots, no content calendar spreadsheet required.
Where AI blog drafts fail: the seven weak spots a human still has to catch
The seven failure modes that show up most often in AI-written posts, what each one looks like on the page, and the edit to make before you publish.
Brand voice rules that actually change the draft: a working config for AI content
The five kinds of writing rule that change what an AI writer produces, a before and after, and a twelve-line starter ruleset to paste in and edit.
A 20-minute editorial review for AI drafts, from top to publish
A six-pass rubric for reviewing an AI-written draft in under twenty minutes, with a clear rule for when to fix inline and when to send it back to the editor agent.
Voice drift across a 20-post run: catching the moments your brand rules stop holding
How voice drift shows up across a long run of AI-generated posts, which parts of your voice config fatigue first, and the review cadence that catches slippage before it ships.
Repurposing is a lie you tell yourself, do this instead
Why turning one post into ten thin posts fails, and what compounding content looks like.
How to publish two posts a week without burning out
A repeatable weekly system that separates deciding, drafting, and editing so nothing piles up.
Setting a post budget when you've never run an AI content pipeline before
RankMill measures usage in credits, not seats or words, and only writing a post spends them. Here is how to size a post budget in those terms, and three worked examples.
Internal links are the cheapest ranking win you're ignoring
Why a cornerstone-and-cluster structure beats chasing backlinks for most small sites.
Stopping an agent run: what Stop gives back, and what it does not
What the Stop button on a running agent actually does, the three kinds of work that cannot be stopped, and how to choose between stopping, editing and retrying.
The keyword research playbook for people who hate keyword research
A practical, no-spreadsheet system for finding topics your customers actually search for, and knowing which are worth writing.
The difference between AI slop and AI drafts you'd publish
It's not the model. It's the research, the rules, and the human editing pass.
Changelog: see what your agents are working on
A Working now panel, run status that updates without a refresh, and one monthly credit allowance shared across the team.
How we keep agents on-brand with writing rules
A look under the hood at how brand and writing rules constrain every draft an agent produces.
Brand voice is a spec, not a vibe. Here's how to write it down
Turn "sounds like us" into rules an agent (or a new hire) can actually follow.
A field guide to search intent for non-SEOs
The only SEO concept a content marketer truly needs, explained without jargon.
When the topic finder surfaces a bad topic: reading the profile to figure out why
Every topic the topic finder proposes is built on the website profile. Read backward from the suggestion to the profile to find the loose description, the wrong reader, or the wrong peer set behind it.
How to tell if your AI blog posts are actually working, 30 and 90 days in
A four-metric framework and three interventions for judging agent-drafted posts at 30 and 90 days, so the ones that miss get fixed and the ones that land get amplified.
How to run a team blog people actually read
A playbook for the format most content teams avoid: small posts, rotating writers, no stock photos, and an archive that compounds.
Topical authority on a small budget: planning clusters when you publish 10 posts a month
How to pick, sequence, and finish one topic cluster at a time when your monthly post budget is small.