← The story

Chapter 08

The Decision

Build this first as a productised commercial-recovery service with software underneath, not as standalone venture SaaS.

— strategic-review-memo-v2.md (One-line VERDICT)

The Decision

Verdict: SERVICES-FIRST, with Australia as the preferred first beachhead. Build this first as a productised commercial-recovery service with software underneath, not as standalone venture SaaS.

Three folded decisions: what kind of company, where it starts, how much to fund.

Services-first, not SaaS. Pure SaaS dies early here, for three reasons:

  1. The field-capture wedge is already commoditised — the buyer won’t pay serious money for another capture UI.
  2. The highest-value output is judgement-heavy (entitlement, causation, notice compliance, quantum) and needs commercial/legal review. “An unreviewed AI answer is not a product; it is a liability generator.”
  3. The moat depends on data permissions buyers won’t grant a cold signup.

First vehicle = productised recovery service: fixed diagnostic fee + account retainer + optional success-linked upside, with software underneath that automates as repeatability is proven. Honest label: “a strong specialist services-first company with a possible SaaS/data moat later… not yet a venture SaaS company.”

Australia-first. The de-biasing scan killed the “empty market” pitch in every geo, but produced a ranking. Criteria weighted 30% willingness-to-pay / 30% recovery regime / 20% competitive density / 20% cold-start ease:

GeoScore
Australia4.35
UK4.05
US3.55
Canada3.35

AU wins on its statutory Security of Payment regime → the cleanest “commercial event → payment claim/adjudication → recovered cash” loop of any English-speaking market. Half the capture loop is legally mandated; rapid adjudication turns substantiated events into near-term cash, not multi-year litigation; public state adjudication datasets (Victoria, Queensland, NSW) are ready-made fuel for the cross-firm benchmark others keep private.

Detail: geo comparison, Australia, US, Canada.

Fund Stage 0 only. Not “raise and build for a year.” Fund exactly one 30-day stage — beachhead validation — and let real evidence release the next tranche. Stage 0 gate: 20+ senior AU buyer interviews, 5+ firms providing redacted project packs, 3+ agreeing paid diagnostic terms in principle, AU counsel confirming the service boundary is legal. Fail (no paid intent / no records access / can only operate as a regulated legal practice) → stop cheaply.

Hypothesis chain

Each hypothesis has a validate-and-advance or fail-and-exit gate.

HypothesisWhat it claimsGate
H1Real demand for recovery existsValidate or exit
H2Firms hand over the messy dataValidate or exit
H3Diagnostic produces genuine value (recovered cash, not nice summaries)Validate or exit
H4Work compounds into reusable structured outcomesValidate or exit
H5Model generalises beyond bespoke archaeologyValidate or exit
H6Cross-firm data loop becomes a real moatValidate or exit
H7Model travels to next jurisdiction without a rebuildValidate or exit

Concrete kill criterion: if “fewer than 3 of the first 5 paid pilots allow your pack into a live payment claim, adjudication prep, final account or negotiated recovery,” the thesis is failing — stop or reshape. Full staged plan: entry plan.

Corrected positioning vs the seed call

Same sentence the work started circling — capture what happened, prove why it matters, turn it into commercial action, reuse the history to price the next job — now with a country, a vehicle, a buyer, a moat, and the exact list of things that would prove it wrong.