How we write your posts

_{Day 177/100}

Hey—It's Tim. 

I want to tell you why we’re re-writing how Penfriend writes.
To the point that our lawyers call me for giving up IP.

Why I’m saying this out loud

I’m sick of the mystery act.

You’re asking a tool to speak for you.

If I’m writing in your voice, you deserve to know how the voice is built—what we read, what we ignore, why we changed the knobs last week.

When other companies go quiet, you feel it in your metrics and your reputation. They call it “IP.” I call it cowardice.

I want Penfriend to be different because I grew up in SEO where you have to show receipts. If rankings dip, if conversions wobble, if tone slides off brand, you need causes, not comforting adjectives. That’s why I’m telling you:

• What we feed the model and why those inputs beat scraping-for-scraping’s sake.

• How the structure is chosen (intent templates, word caps, flow rules) so sections don’t go feral.

• Where the human bits live - stories, stakes, examples—and how we force them to show up.

• What changed, what we expect to happen because of it, and how to override it if you need to.

I don’t remember the last time Surfer told me how they write blogs, or why ByWord chose to make this change or that. It’s frustrating.

Competitors might copy the what. I’m okay with that. We’ll win on the how—taste, discipline, and the willingness to ship changes with a plain-English note.

If we’re going to write for you, we owe you transparency you can actually act on. That’s the whole point of walking you through it today.

How we used to write

We started simple - honestly, better than most, but still simple.

It looked something like this.

Scrape SERP titles
→ propose your title
→ you approve an outline
→ one prompt for the intro
→ a few for the body
→ one for the conclusion.

It shipped. It worked. It wasn’t enough.

How we write today

We feed the model more brains up front:

• SERP titles, People Also Ask, and focused lookups for quotes, studies, and specifics.

• Detect search intent and snap the piece to a matching template (how-to, comparison, teardown, playbook, etc.).

• Build each section with composable prompt fragments so it knows its place in the story and what came before/after.

Result: comprehensive articles that often outrun competitors.
Side effect: sometimes… novels.

Brains up front

The change you’ll feel: front-loaded intelligence, capped output.

Instead of re-researching every section (which multiplies fluff), we do the heavy thinking once, then write each H2–h3 with one decisive prompt and hard word targets.

We read page-one results (not just scrape), map topic coverage, note exact phrases competitors lean on, spot human elements (stories, wins/fails), and surface gaps for a 2025 reader.

That intelligence powers the draft; the draft doesn’t go hunting mid-paragraph.

What this means for you

• Pieces get leaner without losing authority.

• Flow improves because sections aren’t independently overfed.

• Token cost drops; quality signals go up (specificity, order, human bits).

• You can see where we diverge from competitors - on purpose.

See you tomorrow.

✌️ Tim "One Prompt To Rule Them All" Hanson
CMO @Penfriend.ai

Same brain, different platforms: X, Threads, LinkedIn.

P.S. Hold me to this. Every time Google makes an update, I’ll tell you what we’re doing to keep your Penfriend content aiming for the first page.

Penfriend.ai
Made by content marketers. Used by better ones.