Crackerjack / One-page plan

Approve Stage 1. Make the Monday number prove itself.

The investable question is narrow: can Stage 1 land the Monday number automatically, prove saleId ROI, and stop before Stage 2 if the evidence fails?

Blueprint: One-Page Plan

$54.5K

Stage 1 bet

source: Financial Impact + Decision Summary | date: 2026-05-26 | confidence: medium | status: projected

payback

source: Benefit Ledger | date: 2026-05-26 | confidence: medium | status: projected

$70K+

wait cost / qtr

source: Financial Impact | date: 2026-05-26 | confidence: medium | status: inferred

$30K

kill cost

source: Decision Summary + Critical Path | date: 2026-05-26 | confidence: medium | status: projected

The bet

A bounded data-foundation bet, not an open-ended transformation programme.

Every claim is tagged as fact, estimate, assumption, or unknown so the CFO can inspect exactly what is proven and what is being bought as evidence.

What must be true

[FACT] Company purpose: put Monday's merchandising number in front of leadership on Monday, not Wednesday.
[FACT] Audience + decision: CFO and Owner decide within 14 days whether to approve Stage 1 only: a NZD $54,500 bounded data-foundation bet.
[FACT] Problem: the Monday report takes 7-13 manual hours and arrives Wednesday, creating a 2-4 day decision lag on every weekly call.
[FACT] Customer: CFO, Owner, Finance team, Buying team, and Store Managers who need weekly merchandising, catalogue, stock, and FX decisions to use the same substrate.
[ASSUMPTION] Solution: a Unified Merchandising Intelligence substrate plus saleId ROI dashboard can eliminate the weekly reconciliation drag and make catalogue learning visible.

Source trail

Decision Summary: Stage 1 approval, 14-day decision, named GO/NO-GO conditions.
Critical Path: Monday number target, API access dependency, Week 8 kill switch.
Financial Impact: fee, Year 1 cost, cost of waiting, walk-away cost.
Benefit Ledger: payback, benefit:cost ratio, 24-month scenarios.
Prompt Deck: outside-in and inside-out business signals for the board conversation.
Downstream: pitch deck, data room, implementation roadmap, board approval note.

The case

The pain is proven. The lift is still an estimate.

The page should make the investment logic inspectable: Crackerjack already feels the lag, and Stage 1 buys a dated proof that the lag can be removed.

[FACT] Why now: discount tier demand is growing while competitors build data infrastructure; The Warehouse Group is estimated 18-24 months ahead.
[ESTIMATE] Market / opportunity: the immediate value pool is not generic retail TAM; it is 15 stores, weekly catalogue cycles, FX-exposed COGS, and decision lag in Crackerjack's current operating loop.
[FACT] Proof / traction: the live operating evidence is 2 active saleIds with zero attribution data, 7 platform hops in the weekly analysis cycle, and 5 of 7 hops classified as artifact work.
[FACT] Competition / alternatives: direct competitors are Look Sharp and other discount retailers; indirect alternatives are DIY BI, internal spreadsheet process, and big-4 consultancy; status quo is Wednesday reporting and intuition-led catalogue planning.
[ASSUMPTION] Team / edge: the advantage is bounded-bet discipline: Stage 1 creates the substrate, names the kill switch, and prevents Stage 2/3 spend until the Monday number proves itself.

The economics

The upside is attractive because the downside is named.

Every number carries source, date, confidence, and status in the metric notes or source pages. The decision is not whether AI is good; it is whether this bounded bet is worth the named walk-away cost.

[ESTIMATE] Stage 1 transformation fee: NZD $54,500.
[ESTIMATE] Year 1 cost if Stage 1 succeeds: NZD $108,000.
[ESTIMATE] Payback period: Month 9 in the conservative scenario.
[ESTIMATE] Benefit:cost ratio: 2.3x, calculated as NZD $250K conservative floor divided by NZD $108K Year 1 cost.
[ESTIMATE] 24-month cumulative net benefit: NZD $362K conservative, NZD $662K base, NZD $1.36M optimistic.
[ESTIMATE] Cost of waiting: NZD $70.5K-$234K per quarter, already above the Stage 1 build cost.
[FACT] Walk-away cost if Week 8 kill switch fires: NZD $30K-$40K.

The decision

Approve Stage 1 only if the stop rules are signed first.

The ask is commercial, dated, and reversible: sign the budget, confirm access, name owners, and stop if the Monday number cannot prove itself.

Ask + next action

[FACT] Ask: approve Stage 1 within 14 days, capped to the NZD $54,500 bounded bet with a written kill-switch agreement.
[FACT] Use of resources: build UC1 Monday auto-report and UC2 saleId ROI dashboard; confirm POS access; name Documentation Sprint owners; select BI substrate.
[FACT] Milestone unlocked: Monday number auto-delivered by 8am for 4 consecutive weeks, then Stage 2 can be considered.
[FACT] Owner/terms: CFO + Owner sign budget and kill switch; IT contact confirms POS access; Dream team and CFO select substrate.
[FACT] Deadline: Day 7 budget decision, Day 14 GO/NO-GO conditions, Week 8 kill-switch review.

Kill signals

Monday number

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

Pause Stage 1 at NZD $30K-$40K sunk cost before Stage 2 spend begins.

Owner: CFO

Access

POS or e-commerce API access cannot be obtained in writing within 30 days.

Stop integration path at NZD $15K-$25K and redesign around export-based reporting.

Owner: CFO + IT contact

Budget

Total spend exceeds NZD $65K at Week 12.

Stop, review scope, and do not unlock Stage 2 until the budget variance is explained.

Owner: CFO

Put this to work

Build the board pitch with your own AI assistant

For the Owner / Board

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 need to present this AI/data investment case to my business partner or board in plain language and defend it in one screen.

THE BET:
Crackerjack is a 15-store NZ discount retailer. The Monday merchandising report arrives Wednesday after 7-13 hours of manual compilation. Weekly decisions lag 2-4 days behind reality. Catalogue saleIds have zero attributed revenue. 60-70% of COGS is FX-exposed and the hedge model sits in the CFO's head.

THE PROPOSED STAGE:
Approve Stage 1 only: NZD $54,500 over 90 days for the Unified Merchandising Intelligence substrate and saleId ROI dashboard. Continue only if the Monday number lands automatically by 8am for 4 consecutive weeks.

THE ECONOMICS:
Year 1 cost if Stage 1 succeeds: NZD $108K. Conservative payback: Month 9. 24-month cumulative net benefit: NZD $362K conservative to NZD $1.36M optimistic. Cost of waiting: NZD $70K-$234K per quarter. Kill-switch walk-away cost: NZD $30K-$40K.

ALTERNATIVES:
Status quo keeps Wednesday decisions and manual catalogue intuition. DIY/internal likely takes 18+ months without AI/data expertise. Big-4 style delivery costs NZD $200K-$500K with less bounded-bet discipline.

ASK:
Approve Stage 1 within 14 days if four GO conditions are met: CFO + owner sign the budget and kill switch; POS access is confirmed in writing; Documentation Sprint owners are named; BI substrate is selected with quote in hand.

Write me a 5-sentence elevator pitch for this board decision. Then give me the 3 most likely objections from a skeptical owner and how to answer each honestly without overselling AI.

Next →Compare AI chat providers: costs, strengths, fit