Content writing
Most content shops will publish six pieces a week and tell you that’s the work. We publish fewer pieces that land harder — gated by a voice doc built from your founder’s actual writing, edited in three passes by a human, and indistinguishable from your best in-house writer’s output. The job isn’t to fill a calendar. It’s to write something worth reading.
How we work
Some pieces start from the research: we map the gap, brief the piece, write it. Others start from you — a launch piece, a founder essay, a specific brief you already have. Same pipeline, same voice gate, either way.
Driven by research
We do the research. We find the gap. We write the cluster. The map informs the brief; the brief informs the draft; the draft passes through the voice doc; the piece ships. Useful when you want one team handling research and execution end-to-end.
Driven by your prompt
You already know what you want written. A launch piece, a founder essay, the “really good explainer on X” your VP of marketing keeps asking for. Send us the prompt and the voice references; we write the thing. Same quality bar. Same voice gate. No SEO scaffolding required.
What “quality” actually means
Generic AI content is the floor of the industry. Beating it is mostly about process. Here’s ours, honestly described.
PASS 01
Does the piece have a real point of view? Are the claims defensible? Are the examples specific instead of vague? This is where most AI drafts fall apart and we rewrite, not edit.
PASS 02
Every client has a voice doc built from their founder’s actual writing — talks, blog posts, podcasts. The draft is rewritten against it sentence by sentence. Cadence, vocabulary, the specific things this brand does and doesn’t say.
PASS 03
Does the lede earn the next paragraph? Does every section open with a sentence that pulls the reader forward? Does the piece end somewhere — or just stop? This is the final read-aloud pass.
Sample chapter excerpt
Generic, not client-specific. The point is to show the texture — what shipping a piece through the three-pass pipeline produces, versus the default AI output that everyone has read too much of.
From a recent founder-voice essay
The honest version is that nobody was reading it. The flattering version is that we were “refining our channel mix.” The honest version is more useful, so let’s start there.
For three years we published a weekly newsletter because every B2B marketing playbook said we should. The open rate looked respectable, the click rate looked terrible, and the unsubscribe rate looked like a Manhattan rent ladder — one extra step every quarter. We told ourselves the metric to optimize was list growth, which is the metric you optimize when you don’t want to admit the thing isn’t working.
The week we stopped, three things happened. Our biggest client’s VP of marketing emailed us to ask where it had gone, which was nice. Two reporters DM’d to ask if we were okay, which was clarifying. And we wrote one long piece that took the full week instead of five short pieces that each took an hour, which was the part that mattered. The long piece got picked up by the trade publication that had been ignoring our newsletter for thirty-six months.
The lesson, if there is one, is that “more” is rarely the answer when you’re behind. The answer is almost always “less, but the thing that actually works.”
What lands in your CMS
We deliver in the format your CMS wants — clean HTML, Markdown, a Google Doc, or a direct push if you’re on a platform that supports it. Internal links cross-wired, meta and schema written, alt text and OG image briefed.
PER PIECE
Full draft, voice-gated, edited, ready to publish. Internal links to the cluster wired correctly. References cited where claims need backing.
PER PIECE
Title, meta description, slug suggestion, OG image brief, schema type recommendation. Not afterthoughts — part of the deliverable.
PER ENGAGEMENT
Built once, used forever. Cadence rules, vocabulary preferences, what this brand says and doesn’t say. Yours to keep if we ever part ways.
Pricing
Content writing pairs naturally with the SEO research that drives it — one shared map, one voice document. It also runs on its own when you bring the briefs (or your own existing research). Pricing reflects scope; we quote on the walkthrough.
SEO + Content
We build the map, do the audit, write the reconciliation, then execute on the writing. One shared voice document and one backlog — useful when you want one team covering both ends.
Standalone
You bring the briefs (or your existing SEO research). We write. Voice-doc gated, three-pass edit, ready to ship. Same quality bar; you supply the map.
Honest answers
Both, on purpose. Drafts are AI-assisted. Editing is human. The voice doc is the writer — the AI is the apprentice. If you want pure-human writing with no AI in the loop, tell us; it’s a different price point and we’ll be honest about it.
What you won’t get is generic AI output. The three-pass pipeline exists precisely so the finished piece doesn’t read like everything else on the internet right now.
Talks, podcasts, blog posts, founder essays, internal Slack threads if you’ll share them. We read them and extract: cadence patterns, vocabulary preferences, the specific things this brand says (and doesn’t). The doc is one of the deliverables — yours to keep, yours to evolve, yours to give to a different writer if you ever leave us.
Yes — including developer-facing pieces, fintech, healthcare, regulated industries. The voice doc handles tone constraints (no medical advice, FINRA-safe language, the specific things your legal team doesn’t want said). Where the topic requires source-of-truth review, we’ll set up a fast lane with your subject expert and build that into the timeline.
Cadence and volume vary by engagement scope. We size the program to your publishing rhythm, your category, and your operating reality — not the other way around. We’ll talk through what fits on the walkthrough.
Two rounds included on every piece. In practice, most pieces ship after one round because the voice doc front-loads the alignment. If we’re on round three on multiple pieces, something is wrong with the voice doc and we’ll rebuild it.
The chapter excerpt above is a real example, scrubbed. We’ll show more on the walkthrough call — live, because most published client work isn’t labelled as ours and we don’t want to break the illusion.
Whether that’s a competitor doing it well, an old founder essay you loved, or just a vibe — bring it to the call. We’ll show you how the voice doc would capture it.
Book the walkthrough