How should a solo founder prioritize many product bets?
Fifty properties cannot all matter. The arithmetic is unforgiving: if everything deserves attention, attention is worth nothing, and the property that could actually pay for the whole portfolio starves alongside forty-nine that never will. The last piece of the operating system is the one that feels least like building and matters most: triage. Deciding, on the record, which properties are engines, which are experiments, and which are corpses that haven't been told yet.
The priority ladder
Every property on my server carries a priority label from P0 to P4, assigned during periodic audits and written into a portfolio index that agents read before touching anything. P0 means cash engine — confirmed or near-confirmed revenue, the properties whose outages are emergencies. P1 means active growth: real products with live checkouts that haven't proven themselves yet. P2 is the gray zone of maintained experiments.…
Clone hygiene
Several chapters in Part II share a shape: a product that worked-ish, cloned into adjacent markets — another country, another currency, another language — because cloning is nearly free when agents do the work. Free to create, that is. Never free to own.…
Killing things well
Most portfolio advice is about starting; almost none is about stopping, which is strange because stopping is the half I had to learn from incidents. A property is never just its homepage — it's a process on the server, credentials in environment files, a database, DNS records, a renewal invoice, and an entry in every fleet-wide sweep.…
Doubling down
Triage exists to fund its opposite. The point of labeling forty properties "maintain" or "park" is to free the concentrated attention that the two or three real engines deserve — the deep SEO work, the pricing experiments, the trust fixes, the email sequences that only ever happen on properties someone is actually watching.…