Introduction — Two Hundred and Fourteen Posts, Six of Which Were the Blog
Twenty-one clusters started, two finished, and a sixty-two per cent collapse three years in the making.
⚠ Kit published two hundred and fourteen posts in thirty-four months. Six of them produced seventy-eight per cent of the traffic. One hundred and thirty-eight of them had never, in their entire existence, produced a single session worth counting — and it was those hundred and thirty-eight that eventually cost Kit the other six.
What the numbers looked like
⚠ Kit had been publishing weekly, and then twice weekly, for the best part of three years.
| ⚠ Measure | ⚠ Month 34 |
|---|---|
| ⚠ Posts published | ⚠ 214 |
| ⚠ Posts that had ever earned meaningful traffic | ⚠ ⚠ 31 |
| ⚠ Posts producing 78% of sessions | ⚠ ⚠ 6 |
| ⚠ Posts with no meaningful traffic, ever | ⚠ ⚠ 138 |
| ⚠ Clusters started | ⚠ 21 |
| ⚠ Clusters actually finished | ⚠ ⚠ 2 |
| ⚠ Hours per month | ⚠ ⚠ ~46 |
⚠ Five of the six earning posts were in the two finished clusters. The sixth was a fluke and Kit knew it.
⚠ Twenty-one clusters begun. Two completed. Kit had been publishing constantly and finishing almost nothing, and there was no number on any dashboard that said so.
The month it stopped
⚠ Month thirty-one. Sessions fell about sixty-two per cent over three weeks.
⚠ Nothing had been done wrong that week. Kit had published two posts, as usual, and both were fine.
⚠ There was no notification, no email and no explanation, because there is never one.
⚠ Kit spent six weeks looking for the cause among the recent posts, which is where everybody looks. Section 26.6.
⚠ The cause was not recent. The cause was a hundred and thirty-eight posts written to keep a schedule, published across three years, each of which had been individually defensible and collectively fatal. Rule Three.
What Kit had actually built
⚠ Kit had believed the site was two hundred and fourteen posts and that more posts meant a bigger site.
⚠ A search engine had read it as one object — and the object was mostly filler with six good pages in it. Section 5.2.
⚠ A reader had read exactly one page, once, arriving from a search, and had never seen the other two hundred and thirteen. Section 6.1.
⚠ Both of those are true at once, and they pull in opposite directions. That is the whole difficulty of this business, and almost nothing written about blogging acknowledges it.
The pruning
⚠ Month thirty-three. Kit deleted or merged a hundred and thirty-eight posts in four weeks.
| ⚠ Action | ⚠ Posts |
|---|---|
| ⚠ Deleted outright | ⚠ ⚠ 91 |
| ⚠ Merged into a stronger post | ⚠ ⚠ 34 |
| ⚠ Redirected to a cluster pillar | ⚠ 13 |
| ⚠ Kept | ⚠ ⚠ 76 |
⚠ "It took four weeks and it felt like dismantling three years of work, because it was." Section 25.5.
⚠ Sessions recovered over about seven months and then exceeded the previous peak, on a site with a third of the posts. Section 25.6.
⚠ Nothing was added. The recovery was entirely a matter of removal, which is not a thing anybody plans for and is the most common single fix in this trade.
The eighteen months
⚠ Then Kit measured something nobody had suggested measuring: the gap between when a post was published and when it first earned anything.
| ⚠ Post | ⚠ Published | ⚠ First meaningful traffic | ⚠ Lag |
|---|---|---|---|
| ⚠ A | ⚠ Month 14 | ⚠ Month 31 | ⚠ ⚠ 17 months |
| ⚠ B | ⚠ Month 15 | ⚠ Month 32 | ⚠ ⚠ 17 months |
| ⚠ C | ⚠ Month 16 | ⚠ Month 36 | ⚠ ⚠ 20 months |
| ⚠ D | ⚠ Month 16 | ⚠ Month 33 | ⚠ ⚠ 17 months |
| ⚠ E | ⚠ Month 19 | ⚠ Month 38 | ⚠ ⚠ 19 months |
| ⚠ F | ⚠ ⚠ Month 9 | ⚠ Month 22 | ⚠ 13 months |
⚠ Median lag: about eighteen months. Section 8.2.
⚠ Which means every judgement Kit had made about what was working had been made about a site that no longer existed, using evidence about posts written a year and a half earlier. Rule Four.
⚠ It also means the frantic twice-weekly publishing of months 28 to 34 had produced no information at all yet, and would not for another year — so it could not possibly have been the thing to correct. Section 8.5.
What Kit does now
⚠ Publishes less than half as often, and finishes clusters. Chapter 18.
⚠ Counts completed clusters rather than published posts, because that number is visible today and traffic is not. Section 8.4.
⚠ Keeps a cluster board on which nothing counts as done until every planned post exists, is linked from the pillar, and links back. Section 18.9.
⚠ Runs a publishing gate: no post goes live outside a cluster, without sources, or written to fill a slot. Section 19.8.
⚠ Prunes twice a year, deliberately, as a scheduled task rather than as a crisis response. Section 25.4.
⚠ And writes every post as though it is the only page that reader will ever see, because it very nearly always is. Rule Two.
The four rules this book is built on
⚠ One — you publish posts; you are judged on a site. ⚠ Every post changes what every other post means, including making it worse. Chapter 5.
⚠ Two — readers arrive at one page and leave. ⚠ The object you are building is invisible to the people whose behaviour decides its fate. Chapter 6.
⚠ Three — a site is judged on its weakest pages. ⚠ Not its best. Publishing something merely adequate has a cost, and the cost is charged to everything else. Chapter 7.
⚠ Four — the feedback is eighteen months late. ⚠ You are always steering by an instrument reading a site that no longer exists. Chapter 8.
What this book is not
⚠ It is not a search-engine guide, and it states no ranking rule, because nobody outside those companies knows them and the ones who claim to are selling something. Section 15.7.
⚠ It quotes no advertising rate, no earnings-per-thousand figure and no traffic projection. Section 2.9.
⚠ It names no platform, host, theme, plugin, ad network or email service.
⚠ And it does not tell you that this is passive. Kit's site, at its most efficient, needed about eighteen hours a month, forever, and most of those hours were spent on posts that had already been published. Chapter 25.
⚠ Kit had not written a blog. Kit had written two hundred and fourteen posts and hoped they would add up to one — and the six that worked were the only six that had ever been part of a finished cluster.
⚠ That sentence took Kit three years and a sixty-two per cent collapse to arrive at. The rest of this book is an attempt to get you there in an afternoon.
©2026 James Henderson / https://localhandyman.work