Six years, one method:read the site before you pitch it.

I was writing whatever came through the door — e-commerce product copy, the occasional blog series, a landing page if someone was launching something. The work was fine. The briefs all ended the same way: “and make it rank.”

Nobody could tell me what that actually required. Not the clients, and not the agencies subcontracting to me. So I started reading the sites that were ranking, and the answer was uncomfortable: the writing was rarely better than mine. What they had was coverage — other people's sites pointing at theirs.

That was the year the job changed from writing pages to understanding why anyone would link to one.

The model was volume. Forty pitches a day, templated, personalised only in the first line — a mail-merged compliment about a post the sender had not read. We tracked sends, not placements, which in hindsight tells you everything.

The placement rate hovered somewhere around four percent and nobody was allowed to say so out loud. Editors were not ignoring us because they were busy. They were ignoring us because the pitch made it obvious we had not looked at their site.

I stayed long enough to learn the machinery properly. It works, in the sense that a slot machine works — feed it enough and something comes out. It is just an appalling trade for everyone involved.

The pitch to leadership was simple: let me cut sends by half and judge me on placements instead. It was an easy sell only because the existing numbers were bad enough that there was little to lose.

What replaced the volume was research. Every site got read before it got pitched — what the editor actually publishes, what they have covered recently, where a gap genuinely exists. A pitch went out only if there was a real answer to “why this site, why now.”

Placement rate went from single digits to a little over a third. The team sent half as much mail and landed three times the coverage, and the part that convinced everyone was that the placements stuck — nothing to clean up six months later.

Agency economics and careful outreach pull in opposite directions. The agency needed billable hours to scale; the method needed fewer clients, done properly. That argument does not resolve — it just repeats every quarter until someone leaves.

So I went out on my own and took the constraint seriously: a small number of retainers, each with enough room to research properly. No packages, no monthly link quotas, no reselling the same placement to three clients in the same vertical.

Four retained clients in the first year. All four are still on the books, which is the only retention figure I find worth quoting.

The practice runs at six retainers across SaaS, fintech and health. That is not a step toward an agency — it is the ceiling. Past six, the research stops being mine and the thing clients are paying for quietly disappears.

120+ placements landed to date, typical DR range 50 to 90, and every one of them on a site with an audience that was already looking for what the client sells. I turn down more sites than I pitch, which is the part of the job that actually protects the result.

The waiting list runs about a quarter ahead. If that sounds like a sales tactic, it is really just what happens when you cap the client count and refuse to raise it.