Flourishing
LifeBuild
← All verdicts

Topic 5 · Conversation 1 · 2026-05-28

Where does Modus actually change a decision Martin makes?

Topic 5 — Verdict 1

1. Crux of the debate

The question was simple. Where does Modus actually change a decision Martin makes? Not "is Modus cool." Not "what should the wedge be." Where does it flip a real call on a real day. The council started scattered. The Scout pointed at backend picks. The Steward listed three flips. The Platformist pitched a morning to-do list. By call four, they all landed in the same spot. Modus changes one call: "do I send this answer to a paying customer without reading it first." That call only shows up when a buyer is sitting there waiting. No buyer, no call. The Platformist then added a second call: the kill date itself. Two decisions. Zero this week.

2. Idea ranking

  • Morning triage list (Platformist, opening) - A daily "what to touch first" sorter for Martin's inbox and calendar.
  • Backend pick for project two (Scout, opening) - Modus vs pgvector (Postgres with vector search added on, like bolting a roof rack to a normal car) for the one project where graph is the main event.
  • Rollback snapshot before a risky ship (Steward, opening) - Show the data map before the change so Martin hits go.
  • Onboard a messy client in a day not a week (Builder, opening) - Use the graph to chew through new client docs fast.
  • The kill date (Platformist, late) - Friday in ten days, shelve Modus or keep going, based on whether anyone said yes on the phone.
  • The trace flip with a 95% pass bar (Steward and Builder, locked) - "Send this answer to a paying customer unread, yes or no," gated on the trace pointing back to a real source doc 95 times out of 100 in under two seconds.

    • 3. Winning move

    The winning move: Modus changes one real customer call (send an answer unread, yes or no) plus one founder call (shelve or keep going on a fixed Friday), and both depend on Martin dialing the phone this week instead of writing code. Owner: Martin. Trigger: Martin makes five cold calls a day starting 2026-05-29 to lawyers, accountants, and compliance leads at small firms drowning in PDFs.

    4. Losing moves

  • Morning triage list (Platformist) - Killed because a calendar and a small script do the same job. No graph needed. The Platformist conceded this himself by call four.
  • Rollback snapshot as a Modus feature (Steward) - Killed because nightly Postgres backups already do this. Steward conceded.
  • Kill-the-feature usage tracking (Steward) - Killed because a spreadsheet column does this. Steward conceded.
  • One day of plumbing before the calls (Steward) - Killed because the first call asks "do you have this pain," not "look at my demo." Building a demo before anyone says hi is the museum trap the council named twice.
  • Open the 95% trace bar on call one (Steward) - Killed because cold call one is "where does it hurt," not a spec sheet. Like talking about a prenup on a first date. The lawyer hangs up.
  • Sell Modus as a platform (Builder, his own opening framing) - Killed by Builder himself. Sell one feature: answers a reviewer can trace to the source doc. Delete every other slide.

    • 5. Next 7 days

    • Build a list of 50 target buyers (small-firm lawyers, accountants, compliance leads) - by 2026-05-30
    • Make 5 cold calls per day asking "would you pay for AI answers your reviewer can trace to the source doc" - by 2026-06-04
    • Write the one-sentence pitch and the one follow-up question for call one - by 2026-05-29
    • Set the kill date on the calendar (Friday 2026-06-12, shelve or keep going) - by 2026-05-29
    • Log every call outcome in a single spreadsheet (yes, no, maybe, callback) - by 2026-06-04

    6. 30-day milestone

    Milestone: At least one paying customer (signed invoice or paid pilot) for "auditable answers traced to source doc" by target date. Target date: 2026-06-27 Inside View confidence: 20% Outside View base rate: 10%

    7. Hidden assumption

    Nobody named who Martin actually is on the phone. The whole locked answer rides on Martin being able to run a sales call. Five calls a day for ten days is fifty cold calls. That is grunt sales work. It needs a voice that does not crack when a lawyer says "not interested" on call three. It needs follow-up notes, a CRM habit, and the stomach to dial again on call four. The council treated "phone first" like flipping a light switch. It is not. It is a muscle most builders do not have. If Martin freezes at the dial pad, or sends emails instead, the whole verdict collapses. Modus stays at zero decisions changed, and the Friday kill date never gets a real test because the calls never really happened. The variable is Martin's cold-call stamina. If it is high, the plan works. If it is low, he needs a co-founder or a sales contractor by week one, or the answer to the literal question stays zero forever, just like the Builder warned on call fourteen.

    ---

    Full transcript: [[Flourishing/Builders Council/Debates/2026-05-28_where-does-modus-actually-change-a-decision-martin-makes_c1|Debate transcript]]

    Read the full debate transcript →