BUSINESS ROADMAP · FY2027–FY2031

The operating model behind the $100M

Same five years as the plan, read forwards: who we sell to, what we sell them, how the revenue mix layers on, and the gates between stages. Figures are planning targets, not forecasts.

THE MODELFIVE-YEAR LEDGER

The whole plan in one table

Take rate compresses by design — 4.4% on early anchor accounts down to 3.5% blended at scale — while fintech and software lines grow into the gap. Fiscal years run Jul–Jun; FY2027 = Jul 2026 – Jun 2027. ARR and GMV are fiscal-year exit run-rates; the financial plan reconciles them to recognized revenue. New buyers ramp to steady state in ~2 quarters; churn and NRR are planned in the sales capacity model, not discovered later.

Y1 · FY27Y2 · FY28Y3 · FY29Y4 · FY30Y5 · FY31
Exit ARR$2M$8M$22M$48M$100M
GMV$45M$170M$450M$1.0B$2.0B
Active buyers20702005001,000+
Product families810162540
Blended take rate4.4%4.1%3.8%3.6%3.5%
Avg GMV / buyer$2.3M$2.4M$2.3M$2.0M$2.0M
Logo retention · NRR90% · 105%91% · 112%92% · 115%95% · 118%
Facts auto-approved40%60%75%85%90%
GMV per ops FTE$15M$28M$38M$42M$50M
Revenue lines live11–2333
REVENUE MIXHOW THE LINES LAYER ON

Take rate opens the door — the ledger data monetizes it twice more

Transaction revenue funds the wedge. Embedded finance arrives in Y3 once the payment ledger can underwrite; software & data scales as buyers run their own queues on our seats.

Transaction (take rate) Embedded finance Software & data
UNIT ECONOMICSONE BUYER, FULLY LOADED — Y3 TARGETS

The math behind "<6-month payback"

The average buyer runs ~$2.3M of GMV through the ledger per year. At Y3 pricing that's a ~$110K annual revenue account — and the payback claim is a target the G3 gate must prove on two consecutive cohorts, not an assumption.

REVENUE / BUYER / YR

$110K

$87K transaction (3.8% × $2.3M) + ~$23K fintech & software attach.

CONTRIBUTION

$66K

~60% contribution margin after supply-chain ops, review labor, and model costs.

CAC TARGET

≤$30K

Playbook-led sale anchored on a savings report from the buyer's own uploaded data.

PAYBACK / LTV

~5.5 mo

$5.5K contribution per month; 4-year LTV with category expansion >$300K → LTV:CAC > 8x.

Every figure above is a planning target. The G2 and G3 gates exist precisely to verify contribution margin and payback on real cohorts before scale capital is spent.

GO-TO-MARKETWHO WE SELL TO, WHEN

Segment by segment, category by category

Y1 · FY27
Y2 · FY28
Y3 · FY29
Y4 · FY30
Y5 · FY31
Segments
Anchor buyersRegional distributors and multi-unit operators sourcing private label. Founder-led sales.
Expand the wedgeMid-market distributors; first multi-site chains via savings-report proof.
Playbook salesRepeatable motion; expansion revenue passes new-logo.
National accountsLarge chains and buying groups; channel partnerships.
Category standardThe default way private-label foodservice sourcing gets done.
Categories
8 disposable familiesHinged containers, cups & lids, cutlery, gloves, bags, portion cups.
10 familiesRound out disposables; schemas + auto-matching default-on.
16 familiesEnter jan-san and packaging film.
25 familiesPPE and broader packaging; <30-day category onboarding.
40 familiesAdjacent non-food supplies; catalog breadth as retention.
Monetization
Take rate only~3–4% on sourced GMV. Simple, aligned, provable savings.
+ Software pilotsLedger seats for buyer ops teams running their own queues.
+ Embedded financeNet terms in-quote, underwritten from our payment history.
Mix shiftsFintech and data revenue outgrow take-rate line.
$70M + $18M + $12MTransaction + fintech + software & data.
Team
~10 peopleFounders + engineers + one ops lead. Everyone reviews facts.
~22First AEs; ops headcount 2x while GMV ~4x — the leverage proof.
~45Sales pods, credit lead, category managers.
~90Regional coverage; risk & compliance function.
~150Revenue per head keeps climbing — agents absorb the volume.
STAGE GATESWHAT UNLOCKS THE NEXT YEAR

No gate, no go

The path, gated

STAGE GATES · FLOW
flowchart TB
  Y1["Y1 · Prove
$2M ARR"] -->|"G1 · Trust gate"| Y2["Y2 · Platform
$8M ARR"] Y2 -->|"G2 · Leverage gate"| Y3["Y3 · Repeat
$22M ARR"] Y3 -->|"G3 · Repeatability gate"| Y4["Y4 · Scale
$48M ARR"] Y4 -->|"G4 · Scale gate"| Y5["Y5 · Compound
$100M ARR"]
CAPITAL AND HEADCOUNT UNLOCK AT GATES — NOT AT DATES
G1

Y1 → Y2 · Trust gate

20 buyers with repeat orders, quote-dispute rate near zero, evidence chain holds in every audit. If buyers don't trust the ledger, nothing else matters.

G2

Y2 → Y3 · Leverage gate

GMV per ops FTE up ~1.9x, 60% auto-approval, buyers running their own review queues. Proves the platform scales sub-linearly with people.

G3

Y3 → Y4 · Repeatability gate

CAC payback under 6 months across two consecutive cohorts, expansion > new-logo, fintech losses within model. The playbook works without founders in every deal.

G4

Y4 → Y5 · Scale gate

Fill rate and OTIF hold at $1B GMV, new categories land in under 30 days, credit book performing at portfolio scale.

CAPITAL PLANSCENARIO — SUBJECT TO PERFORMANCE

Raise against gates, not against hope

NOW

Seed — $6M

Funds Phases 0–5 and the first 20 buyers through the trust gate, reaching G2 with a ~$1.2M buffer on plan. Hiring gates protect the floor.

RAISING
AT GATE G2

Series A — ~$18M

Raised on leverage proof: $8M exit ARR, $170M GMV run-rate. Covers the Y3–Y4 growth trough (peak cumulative burn ~$11M) with ~2x coverage.

SCENARIO
AT GATE G3

Series B — growth

Raised on repeatability: $22M ARR and a performing credit book. Funds national scale and the category machine.

SCENARIO
RISK REGISTERWHAT COULD KILL THIS — AND THE DESIGNED ANSWER

Named risks, engineered mitigations

Credit losses FINTECH

  • Terms extended only against our own payment-ledger history — never bureau-only
  • Exposure caps per account, live in the financial sub-ledger
  • Capital partner keeps the book off our balance sheet at scale

Buyer concentration REVENUE

  • No buyer above a set share of GMV before G3
  • Expansion motion widens the base faster than whale-hunting

Supply shocks TARIFFS · FREIGHT

  • Landed-cost engine reprices from versioned assumptions in days, not quarters
  • Multi-country supplier bench per family; substitution via the equivalency graph

AI accuracy plateau DESIGNED FOR

  • Auto-approval thresholds are policy, granted and revoked by regression evals
  • Human review absorbs any rollback — the ledger never depends on model perfection

Channel conflict DISTRIBUTORS

  • Positioned as sourcing infrastructure for private label, not a branded-goods marketplace
  • Distributors are customers of the ledger, not disintermediated by it

Take-rate compression IN THE MODEL

  • Compression is declared — 4.4% → 3.5% — not discovered
  • Fintech and data lines grow into the gap by design