For a while I stepped away from the internet and worked the land. Farming doesn't care about your launch date. You prepare soil, respect cycles, and earn the harvest. That patience changed how I build everything since.
The short answer
Why the usual approach quietly fails
The default failure mode is effort without leverage — doing more of the same and hoping volume rescues a weak mechanism. It rarely does. The work that compounds shares three traits: it is owned, it is measurable, and it gets better the more it runs.
- Owned: it lives on an asset you control, not a rented feed.
- Measurable: one north-star number you check weekly, not a dashboard you ignore.
- Self-improving: each cycle feeds the next with data, links, or distribution.
The mechanism, step by step
Start narrow. Pick the one query, channel, or surface where you can plausibly become the best answer within ninety days. Depth before breadth — a complete cluster outranks a scattered library every time.
- 01Map the real demand — the questions in the words people actually use.
- 02Build the spine: one pillar asset, deeply interlinked with supporting pieces.
- 03Instrument it: define the single metric that proves it's working.
- 04Compound it: refresh, expand, and redistribute on a fixed cadence.
Growth is not forced. It is designed. Isn't?
— An operating principle from the farm
What this looks like in practice
In production this means treating every asset — a page, a campaign, a product — as infrastructure. It has an owner, a job in the funnel, and a review date. When something stops doing its job, you fix it; you don't bury it under new work.
Info
Common mistakes to avoid
- Chasing volume before the mechanism converts a single unit of attention.
- Optimising for clicks instead of qualified intent.
- Letting AI replace your judgement instead of accelerating it.
Do the unglamorous version well and the compounding takes care of itself. That is the entire game.