Rowanstead

What it was

A web and iOS app for the adult child who lives far from an aging parent and is coordinating care through siblings, a home-care agency and a phone. Tasks, rides, a shared log, a digest, voice notes, and later a version aimed at the agencies themselves.

I built it solo: Next.js and Postgres on a VPS, an Expo mobile app, a Playwright test suite, and a small fleet of coding agents doing most of the typing. Version 1.0 went to Apple for review on September 2, 2026.

The thesis, and how it changed

The first thesis was that home-care agencies needed a sign-off record: proof, for the family and for compliance, of what was done on each visit. That died on a single call on August 10, 2026, when an agency owner explained, patiently, that this is not a problem they have. The repo kept building on it for three more weeks anyway, which is the mistake I would most like not to repeat.

The second thesis was absence. The real pain belongs to the person who is not in the room, and it was my own problem, which is why I believed in it. That thesis I still think is right.

Why it was set down

Because a correct thesis is not a business until someone with the pain also has a reason to pay.

• The agency has budget but not the pain. They were polite, and nobody bought.
• The family has the pain but no habit of paying. A one-time setup fee is not a business, and a monthly subscription for looking after your mother is the guilt tax that filled the consumer-caregiving graveyard before me (CareZone, Lotsa Helping Hands, Carely: all free, none of them a company).
• The employer, the way Wellthy and Cariloop sell, is the one shape where the payer has both budget and motive. It is also a six-to-twelve-month enterprise cycle for a solo, unincorporated, pre-revenue founder. I named it and did not start it.

What the numbers actually said

Almost nothing, and that is the honest part. Every discouraging number the project produced (224 prospects, 50 invitations, one acceptance, seven replies from 28 DMs, a 649-row dial list) was aimed at the thesis that had already died. The gate that mattered, ten real conversations with people far from a parent, stood at one of ten on the day I stopped.

What I would do first if I picked it back up

Ten conversations. Not a build.

What I kept

The kill-gate habit, written before the first call. The agent harness that built most of the product. And a working App Store submission that taught me more about Apple review than I wanted to know.