Skip to main content

Fund Future

What is the conservative breakeven, and what is the cost of waiting?

The bet

NZD $54.5K

Stage 1 — bounded, named exit at every gate

The bleed

$23.5K–$78K

per month of delay · one quarter exceeds the build cost

The deadline

14 days

to approve — or accept Q3 deferral cost

§1

7-Input Cost Model

Most CFOs in this position assume cost = tooling + build. That misses ongoing tuning, change management, opportunity cost, error rework, and integration drift. The true Stage 1 cost is NZD $54,500 not NZD $25,000. Every input below tagged VERIFIED or INFERRED so you know exactly what you are betting on.

Tooling / licensing (BI substrate quote needed)

Stage 1 NZD $4K · Year 1 NZD $9K

INFERRED

Build / integration — data engineering (POS API feasibility needed)

Stage 1 NZD $25K · Year 1 NZD $30K

INFERRED

Change management — Documentation Sprint + training

Stage 1 NZD $5K · Year 1 NZD $9K

VERIFIED

Ongoing tuning / maintenance — schema drift + SKU types + upgrades

Stage 1 NZD $0 · Year 1 NZD $12K

INFERRED

Opportunity cost — Finance team time on build

Stage 1 NZD $4K · Year 1 NZD $6K

INFERRED

Error / rework cost — rollout contingency

Stage 1 NZD $2K · Year 1 NZD $4K

INFERRED

Integration drift over 12 months — API change catch-up

Stage 1 NZD $0 · Year 1 NZD $8K

INFERRED

Build (one-time)

Stage 1 total: NZD $54,500

Ongoing (Year 1)

Year 1 total: NZD $108,000

Total Stage 1: NZD $54,500. Total Year 1: NZD $108,000. Three inputs carry most of the uncertainty — tooling, build, and ongoing tuning. Sign off those three by Day 7. Approve the full NZD $54,500 by Day 14.

→ Pattern →
§2

3-Scenario Breakeven Curve

Same formula across all three scenarios. Scenarios differ by volume assumption only (catalogue ROI uplift × store stock-position improvement × baseline reclamation). We are not changing how we count — we are changing what we assume about response.

3-Scenario Breakeven — NZD thousands

NZD-100NZD200NZD500NZD800NZD1,100NZD1,400BreakevenM1M3M6M9M12M18M24Base ~M6Conservative ~M9ConservativeBaseOptimisticValues in NZD

Decision logic: if conservative breaks even inside 12 months, approve. If it doesn’t, optimistic is fiction — not a reason to proceed. Use Month 9 conservative breakeven as the go/no-go threshold.

→ Pattern →
§3

Breakeven Curve

Inflection narrative

  • Month 1: -NZD $25K (Stage 1 build cost outflow begins)
  • Month 3: -NZD $54.5K (Stage 1 build cost peak; conservative + base + optimistic identical)
  • Month 6: Conservative -NZD $54K | Base -NZD $5K | Optimistic +NZD $80K (Base breakeven hits)
  • Month 9: Conservative -NZD $5K (BREAKEVEN imminent) | Base +NZD $90K | Optimistic +NZD $245K
  • Month 12: Conservative +NZD $34K | Base +NZD $185K | Optimistic +NZD $410K (Year 1 close)
  • Month 18-24: cumulative compound benefits as Stage 2-3 unlock further upside

At Month 6 — 3 months before breakeven — the cumulative curve should be at approximately -NZD $54K. If it is not (e.g. still at -NZD $80K), that is the trigger to review Stage 1 assumptions — not to abandon the project. Set a monthly review at Month 6 to confirm tracking. Decision owner: CFO.

→ Pattern →
§4

Opportunity Cost of Waiting

Five forms of inaction cost — each specific to Crackerjack, NZD/month per form:

Inaction cost — NZD per month deferred

  • Lost catalogue iteration efficiency — 26-52 cycles/year × Marketing+Buying time saved with attribution. NZD $4K-8K/month deferred.
  • Lost margin on poor-payback saleIds repeating — 0.5-1.5% gross margin compression on catalogue-driven revenue (60-70% of base). NZD $7.5K-30K/month deferred.
  • Lost FX-protection upside — 0.3-0.8% margin protection from better hedge cover on FX-exposed COGS. NZD $4K-15K/month deferred.
  • Competitive displacement (Warehouse + Look Sharp investing in data) — quarter-of-margin-share to competitors / 12 months. NZD $5K-15K/month deferred.
  • Talent retention drag — key buyer departure risk with 200+ supplier relationships in head. NZD $3K-10K/month deferred.

Competitor signal: The Warehouse Group is at least 18-24 months ahead on group-level data infrastructure. Every month of delay narrows the option to differentiate on price-accuracy.

Total cost per month of delay: NZD $23.5K-$78K. Waiting another quarter costs NZD $70.5K-$234K — already exceeds Stage 1 build cost of NZD $54.5K. Trigger the decision now or accept the deferral cost.

Build cost vs. cost of waiting · same scale

Stage 1 build (one-time)NZD $54.5K
+1 month waiting$23.5K–$78K
+1 quarter waiting$70.5K–$234K
↑ exceeds build cost

Build is a one-time cost. Waiting is a recurring cost. Every quarter of delay buys nothing and spends more than the build itself.

→ Pattern →
§5

Kill Switch

The kill switch is not pessimism — it is what makes this a disciplined bet, not a leap of faith. Three named triggers with cost-to-walk-away and decision owner. This is the shape of bet a board-ready CFO funds — bounded, instrumented, killable. The alternative is the unbounded leap-of-faith spend that ends a tenure in Q3.

CRITICALOwner: CFO

Monday number cannot land before Wednesday 8am for 4 consecutive weeks at Week 8

NZD $30,000–40,000 walk-away cost · Week 8

HIGHOwner: CFO + IT

POS or e-commerce API access cannot be obtained in writing within 30 days of Stage 1 kickoff

NZD $15,000–25,000 walk-away cost · Day 30

HIGHOwner: CFO

Total spend exceeds NZD $65,000 at Week 12 (Stage 1 budget +20% contingency)

NZD $65,000 walk-away cost · Week 12

Decision deadline

Approve Stage 1 within 14 days

Bounded bet: NZD $54.5K with named exit at every gate — not a leap of faith.

Prediction Discipline — anchored to CFO verbatim concern:

“Attempting to predict the future to understand ROI on investing in the future — most have zero idea how to invest in tech.” — CFO

We answer this concern directly. Predict before you invest. The prediction: the Monday Merchandising Report will be auto-generated and delivered to the CFO inbox by 8am Monday for 4 consecutive weeks by Week 16. If it does not by Week 12, trigger kill condition #1.

This is a bounded, falsifiable, dated prediction. Approve NZD $54,500 with a named exit at Week 8 if the Monday number trigger fires, or Week 12 if budget overruns. The risk is known. Decide by Day 14. The decision is not a leap — it is a bounded bet with a named walk-away cost.

→ Pattern →

Year One · the bet that worked

What this looks like on the other side of yes.

Not a transformation deck. Five specific Mondays — each tied to a trigger or curve already named above.

Week 4

POS API in writing

Kill trigger #2 cleared. The biggest unknown becomes the smallest.

Week 16

Monday 8am

Merchandising report in inbox. Four consecutive weeks. The CFO prediction lands.

Month 9

Breakeven

Conservative curve crosses zero. The bounded bet is now self-funding.

Month 12

+NZD $34K

Year 1 close, conservative case. Board sees the Stage 2 unlock with evidence, not faith.

Month 18

Warehouse gap closed

The 18-month group-data lead stops being a leak. Price-accuracy is now a moat.

The alternative is the same year — without any of these Mondays. Stage 1 cost stays in the account. The bleed stays on the P&L. The Warehouse keeps the lead.

One decision. Day 14. Approve.

Put this to work

Stress-test the bounded bet

For the CFO

Copy this prompt. Paste into Claude, ChatGPT, or any AI assistant. The page context is already loaded — send it and get analysis tailored to your role.

I am the CFO of a NZ mid-size retailer with a 4-store footprint and a buying team carrying 200+ supplier relationships in their heads. A Stage 1 AI + data infrastructure proposal has landed on my desk and I have 14 days to approve.

The shape of the bet:
- Stage 1 build cost: NZD $54,500 (one-time). Year 1 ongoing: NZD $108,000.
- 7 cost inputs tagged VERIFIED or INFERRED. 1 is VERIFIED (change management). 6 are INFERRED — tooling, build, ongoing tuning carry most uncertainty.
- 3-scenario breakeven: conservative Month 9, base Month 6, optimistic Month 4. Decision logic: if conservative breaks even inside 12 months, approve; if not, optimistic is fiction.
- Cost of waiting: NZD $23.5K–$78K per month deferred. One quarter of delay costs NZD $70.5K–$234K, already exceeding the Stage 1 build cost.
- 3 named kill conditions: (1) Monday number cannot land before Wed 8am for 4 consecutive weeks at Week 8 → walk away NZD $30K–$40K. (2) POS API access not in writing within 30 days → walk away NZD $15K–$25K. (3) Spend exceeds NZD $65K at Week 12 → walk away NZD $65K.
- Falsifiable prediction: auto-generated Monday Merchandising Report in CFO inbox by 8am for 4 consecutive weeks by Week 16. If it does not by Week 12, trigger kill #1.
- Competitor signal: The Warehouse Group is 18–24 months ahead on group-level data infrastructure. Every month of delay narrows the option to differentiate on price-accuracy.

I want you to stress-test this as a board-ready CFO would. Two specific questions, not yes/no:

1. Which of the 6 INFERRED inputs (tooling $4K, build $25K, ongoing tuning $12K, opportunity cost $4K, error/rework $2K, integration drift $0/$8K) would you require to be re-tagged VERIFIED before you sign off, and what evidence would clear each one?

2. Where would you tighten the kill switches — would you move the Week 8 Monday-number trigger forward to Week 6, lower the +20% Week 12 budget contingency to +10%, or add a fourth trigger I have not named? Walk me through the trade-off between false-positive kill cost and unbounded-spend risk.
Next →Compare AI chat providers: costs, strengths, fit