Cycle 1
Instead of building outright, this cycle is focused on distribution, intercepting demand from existing audiences, and validating new hypotheses.
Objectives and key results
Enginery
Prove whether any free product can reach a stranger, not just ship to one
- Combined downloads across all free products reach 30 by Sept 13
- One ecosystem registry listing published by Sept 7
Italian (Paroletta)
Get one Italian product in front of someone who wasn't already looking for it
- Paroletta iOS downloads reach 10 by Sept 13
- Paroletta Android downloads reach 10 by Sept 13
Salotto and Filotto carry no objective this cycle. Filotto is parked until Oct 1. Salotto has an open bug worth fixing, but it's quality work, not a distribution objective.
Pet Med Reminder
Prove the App Store search channel actually works, since PMR has the cleanest data to test it on
- PMR downloads reach 26 by Sept 13
Attribution tool (AI-native engineering lane)
Find the job in this audience that your experience can serve and a tool can't
- One job named by Sept 13 that clears the job-served check and the experience filter: could a sharp AI-fluent engineer with no leadership scar tissue credibly do this job? If yes, it fails the filter
Show Notes
Get someone else using it
- Per-account processing cap enforced by Sept 7, with a test that trips it
- One audience and channel named for Show Notes that clears the Aug 29 gate, by Sept 13
Vine Trails
No objective. Re-staged personal on 2026-08-30 after the intercept check found the job closed at ten shipped incumbents.
The retro
Focus on growth over revenue and intercepting existing demand where it exists.
How I chose this work
I start with three questions:
- Can I intercept demand? I’ll optimize my pet med reminder app’s store listing to test ASO, and publish one tool into an ecosystem registry to test how well an indexed entry compounds over time.
- How do my existing users find out about my products? Strangers arrived at my Italian word game. One came back the next day and finished a test he had abandoned three times, and I don’t know how they got there. I need instrumentation.
- What should get built for engineering leaders? I have a real advantage with this audience: the know-how and the skills to transition engineering teams to the AI-enabled PDLC. I’ll produce a hypothesis for six candidate jobs-to-be-done, checked against the market.
Because I’m adopting a fully AI-native PDLC, I’d like to test whether my playbook survives contact with reality. It contains an ordered list of moves, ranked by how much is learned per hour spent, and I’m crafting my OKRs for this cycle from those.
I’m picking the two cheapest moves in the playbook: the registry listing and reading a message that already went out.
No key result may commit build hours until someone has searched for the competitors and written down what they found. Zero or one is open. Two or three is contested. Four or more is closed, and the search costs about a minute.
No-Gos
- Revenue. I’m focusing on growth, not income. No monetization.
- Any channel that needs an audience to already exist. No newsletter, no group announcement, no share-loop target.
- Features. This cycle is about distribution, not building.
You can read my retro for Cycle 0 here.
Get updates
Occasional notes on what's happening at Enginery. No spam, no marketing.