PLAYBOOK
More content,
and it still sounds like you.
Publishing more is easy now, and that's the problem: most of what gets produced at volume reads the same. This playbook is about raising output without losing the voice, the accuracy or the editor's judgement that made your content worth reading.
Three times the output, same team.
The plan calls for far more content than the team can write by hand. AI drafts are quick, and the first few looked fine. By the twentieth they all sound alike, a couple contain claims nobody can source, and your best editor is spending the week rewriting them.
"This doesn't sound like us."
The draft is competent and anonymous. It could have come from any company in your category.
"Where did this number come from?"
A confident claim with no source. At volume, nobody catches them all.
"We're publishing more and ranking for less."
Output went up. Nobody checked whether each piece matched what the searcher wanted.
Teach it once, then keep a person in charge.
The gain comes from doing the setup properly. A voice profile built in ten minutes produces ten-minute content.
Give it your voice
Feed in the pieces you're proudest of: documents, URLs, files. Adjust the tone until a sample reads like something you'd publish.
Brand voice →Pick topics worth writing
Start from content opportunities found in your own rankings and your competitors' coverage. Volume for its own sake is how you end up with forty articles nobody searches for.
Content opportunities →Decide before you draft
Pick the template, choose the keywords and edit the outline. Ten minutes here saves an hour of rewriting.
Content brief →Outline generator →Let it write, and check itself
The draft goes through three proofing passes and comes back with five scores and a list of fixes.
Content score →Free: GEO-friendly content checker →A person signs it off
Someone reads it, checks the facts and decides whether it's good enough. This step doesn't get automated.
Content pipeline →SEO content optimization guide →Check a draft's structure before it goes live.
Paste the draft into the GEO-friendly content checker, or open the published version with the extension, to see whether the structure came through as planned.
Where the time goes, and where it shouldn't.
Ledgerly is our demo account, a B2B accounting tool for agencies. Its two-person content team needed to publish weekly.
Your voice comes from your own writing.
A tone setting like "friendly" or "professional" gets you a generic version of that word. What works is showing it how you write.
- Learned from your materialDocuments, URLs and files you choose.
- Tone you can adjustTune it until the sample sounds right.
- Writer personasA profile for each person behind your content, with their own experience and way of putting things.
Most bad articles were bad briefs.
If the brief is vague, the draft will be too, and you'll spend the time you saved on rewriting it.
- Topics from your own dataEach opportunity is tagged as a competitor gap, a ranking opportunity or a high-volume topic.
- A template that fitsA how-to guide, a comparison, a product page. The opportunity suggests one.
- Keywords you choosePrimary and secondary, with volume, competition and intent beside each.
- An outline you can editWith the keyword placement and schema planned.
The draft arrives already checked.
A first draft is cheap. What costs time is finding out what's wrong with it, so that part is done before it reaches you.
- Three proofing passesThe draft, then E-E-A-T and intent, then schema and meta.
- Five scoresSEO, readability, E-E-A-T, GEO and intent.
- Fixes for each issueAccept one, edit it, or ask for another version.
- Any section, redoneRegenerate, expand, shorten or simplify a single section without touching the rest.
Somebody still has to say yes.
We'd be wary of any process that publishes without a person reading the piece. Software doesn't know which of your claims are true.
- Facts checked by a personEvery figure and every claim about your product.
- One view of everything in progressThe content pipeline shows each article's status and names the ones that are stuck.
- Export in the format you needHTML, Markdown, plain text, DOCX or PDF.
Who does what
| The task | Who does it | Why |
|---|---|---|
| Choosing topics | You, from a suggested list | Only you know what the business wants to be known for. |
| Choosing keywords | You, from suggestions | The suggestions come with data. The choice is a judgement. |
| Writing the first draft | Content Studio | It's the slow part, and the part that follows a brief. |
| Keeping the voice | Content Studio, checked by you | It learns from your writing. You'll still hear when it's off. |
| Checking the facts | You | Software can't tell which claims about your product are true. |
| Deciding to publish | You | Somebody has to put their name to it. |
Where the team's time went, before and after.
This is the demo team's month. More got published, and more of the week went on editing and checking.
Writing everything by hand
- –Four articles a month
- –Most of the week spent on first drafts
- –Topics chosen from a keyword list
- –Little time left to check or improve anything
With drafts handled
- ✓Eight articles a month, each one read by an editor
- ✓Most of the week spent on briefs, facts and edits
- ✓Topics chosen from gaps in their own rankings
- ✓One article sent back, because it wasn't good enough
The numbers are from the demo account and are there to show the shape of the change. Yours will depend on your team and your standards.
For teams that publish more than they can write.
How do you scale content without losing your brand voice?
Teach the software your voice from your own writing, and build a persona for each writer. Choose topics from the gaps you found, and settle the template, keywords and outline before any draft is written. Let the drafting and the first round of checks run by themselves. Then have a person check the facts and decide whether each piece is good enough to publish.
Key facts
- A voice profile is only as good as the writing you feed it. Use your best pieces
- Most weak AI drafts trace back to a vague brief
- Drafting is the step to hand over. Fact-checking and approval are the steps to keep
- Publishing more only helps if each piece matches what the searcher wanted
- Content Studio learns brand voice from your documents, URLs and files, and scores every article on five measures
- No process should publish without a person reading the piece first
Last reviewed 15 October 2026 by the SEORCE team.
Frequently asked questions.
Yes, if it learns from your own writing. A tone label produces generic copy. A profile built from your best documents and pages gets much closer, and an editor closes the rest of the gap.
No. It replaces the slowest part of their week, which is the first draft. Choosing topics, checking facts and deciding what's good enough stay with people.
It depends on how much checking each piece needs. Technical or regulated subjects need more. We'd set the pace by what your editor can read properly, and no faster.
Google's stated position is that it rewards helpful content however it is produced, and acts against content made mainly to manipulate rankings. Thin, unchecked articles at volume are the risk, whoever wrote them.
Use writer personas, vary the templates, and edit the outline before drafting. Sameness usually comes from identical briefs.
Yes. Set up each client or brand as its own project and teach each one its own voice, so one client's content doesn't read like another's.
You can check a draft for free with the GEO-friendly content checker, the readability score analyser and the AI tell detector. Checking it against your brand voice is part of Content Studio.
Start with your voice, then write one article.
Add your site, feed in a few pieces you're proud of, and brief a single article. Judge it the way your editor would.
Not ready for an account? Use the free browser extension or any of the free tools.