Personalized outbound at campaign scale
Personal outreach, written one prospect at a time — at campaign scale
Every prospect gets their own landing page and their own message — written by name and to their actual situation, not merged from a template. One run produced hundreds of genuinely personal pages, each one fact-checked so nothing about the prospect is invented.
Outbound makes you choose: genuinely personal but only a handful at a time, or high-volume but obviously a form letter. You couldn't have both.
An engine that writes each prospect their own landing page and message — by name, to their real situation, in the company's voice — and ships hundreds in one run, with hard guardrails so nothing about the prospect is invented.
| Before | After | |
|---|---|---|
| How personal each message is | a first name merged into a template | a page written to that prospect's real situation |
| How many you can do | a handful a week, by hand | hundreds in a single run |
| What stops a made-up claim | whoever proofreads it, if anyone | every fact tied back to the research brief |
| Quality control | spot-checked, or not at all | voice, links, and emails checked on every page |
| Knowing what worked | a guess once replies trickle in | tracking on every link, read in one place |
The whole point of personalization — that the recipient feels written-to, not processed — survives being done hundreds of times in a single pipeline. Hand-writing pages this personal at this volume isn't a slower version of this; for a team it simply doesn't happen. Mail-merge gets the volume but loses the personal. This is the combination that wasn't on the table before: every prospect gets something made for them, and the run still finishes in one pass.
What I built
An engine that produces personalized outbound at campaign scale — one tailored landing page and one tailored message per prospect, generated individually rather than merged from a template. It's standalone software with its own deployment, not a workflow that runs inside a chat window.
- A page per prospect, written to them. Each prospect gets their own landing page and message, addressed by name and written to their actual situation, in the company's own voice. These aren't slots filled into a shared template — each page is composed for the person it's going to.
- Grounded in real research. Every page is built from a researched brief about that prospect, and the engine can only use facts that came from that brief. It can't reach for a flattering detail that isn't there — which is what keeps a "personal" page from quietly becoming a false one.
- Voice rules as a hard constraint. The company's voice is enforced, not suggested. A page that drifts off-voice is caught and corrected before it ships, so a hundred pages still sound like one company wrote them.
- A full quality harness. Before anything goes live, an automated pass lints the voice, checks every link works, and validates every email address. A page that fails a check doesn't ship. This is what makes volume safe: you don't have to hand-read every page to trust it.
- Built and deployed for real. In one run, the engine generated 463 pages from 677 researched lead briefs in a single pipeline, deployed them as live static pages, and put tracking on every link so the team could see what got opened and clicked.
The numbers are about effort and reach, not results: this is what one pipeline produced, not a claim about replies or revenue.
Why it matters
The payoff is having both things outbound usually forces you to choose between: messages that genuinely feel written to the person, and enough of them to run a real campaign. Hand-writing pages this personal, this many times, is the kind of work a team never actually does — it's too slow, so it doesn't happen. Mail-merge does the volume but every recipient can feel the template. This engine does the personal part individually, hundreds of times, in one run.
And it does it without the risk that usually comes with letting a model write at scale. The guardrails — facts tied back to research, voice enforced, every page checked — mean the volume doesn't come at the cost of trust. That's the real unlock: not faster outreach, but a kind of outreach a team couldn't produce at all before, made safe enough to actually send.
The hard part isn't writing a personal page — it's writing hundreds of them without one of them inventing a flattering detail that isn't true. At scale, one hallucinated claim about a prospect's business turns a thoughtful message into an embarrassing one, and you can't read every page by hand if the whole point is volume. So invention is treated as the thing to engineer against: the engine can only use facts that came from the prospect's research brief, the company's voice is enforced as a hard rule rather than a suggestion, and a quality harness lints the voice, checks every link, and validates every email address before a page is allowed to ship. A page that can't be backed up doesn't go out. That restraint is what makes sending hundreds of personal pages safe rather than reckless.
The shift is that personalization stops being something you can only afford a few times a week. Each prospect gets a page made for them, hundreds of pages come out of one run, and the guardrails mean nobody has to choose between "personal" and "trustworthy." It's the combination outbound never let you have — written-to, at scale, and safe to send.
The anti-invention guardrails — how the voice rules are enforced, how every claim is tied back to a researched fact so the model can't flatter a prospect with something untrue, and how the QA harness catches a bad page before it ever ships — are what make personal-at-scale safe to send. That discipline is the part clients pay for.