Business Model Classification — Crackerjack NZ
Current model vs AI-enabled target model + exit optionality. April 2026.
Current Model — Relationship-Dependent Discount Retail
Classification: Opportunistic buying + planned category retail. Founder/management dependent. Margin driven by sourcing advantage. Risk unmanaged.
How Value is Created Today
| Layer | How it works | Dependency |
|---|---|---|
| Sourcing | Buying team hunts clearance and planned category deals globally | Buyer relationships + Craig Faulkner network |
| FX management | CFO reconciles quarterly; hedging ad hoc if at all | CFO time + manual processes |
| Pricing | Set below market at point of buying decision | Buyer judgment + margin estimation |
| Competitive intelligence | Informal — sales decline is the signal | None systematic |
| Inventory management | Store-level + buying team judgment | Manager experience |
| Reporting | CFO manually assembles monthly board report | CFO time + accounting system exports |
Structural Ceiling
At ~70 staff across 15 stores, the business is running near peak efficiency for its current operating model. Growth requires either:
- More stores (capital intensive, lease market tightening)
- Higher margin per store (requires better buying, better pricing, lower FX exposure)
- Online channel expansion (Head of eCommerce role exists — this is the growth path)
The current model cannot improve margin without improving the quality of buying decisions and reducing the FX cost of import commitments. Both require analytical capability that does not currently exist.
Fragility Points
- CFO departure: All FX knowledge, financial model logic, and cash flow intelligence leaves with the CFO. No documentation, no system, no backup.
- Buying team turnover: Supplier relationships and category knowledge are personal. No formal knowledge base.
- NZD sustained weakness: If NZD/USD remains below 0.58 for 24 months, the accumulated unhedged margin cost compounds significantly.
- Temu category penetration: If Temu reaches 35%+ of NZ adults in Crackerjack's core categories, pricing pressure becomes structural rather than competitive.
Target Model — Intelligence-Led Discount Retail
Classification: Systematic buying advantage with AI-augmented treasury and competitive intelligence. Documented, transferable, scalable.
How Value is Created in the Target Model
| Layer | How it works | Dependency |
|---|---|---|
| Sourcing | Buying decisions scored by AI: landed cost, FX risk, Temu competitive exposure | System + buyer judgment |
| FX management | Real-time exposure dashboard. Weekly model. Hedges placed against policy, not timing | System + CFO sign-off |
| Pricing | Category margin model informs price setting — FX position + Temu benchmark built in | System + buyer judgment |
| Competitive intelligence | Weekly Temu monitoring. Quarterly landscape update. Alerts before margin erodes | System + buying team review |
| Inventory management | AI ageing alerts + markdown triggers based on sales velocity + Temu price signals | System + manager judgment |
| Reporting | Auto-generated monthly CFO report. Narrative reviewed, not assembled | System + CFO editing |
What This Changes
For the CFO: From data assembler to intelligence consumer. Hedging decisions made with full information. Board reporting time: 4 hours → 45 minutes.
For the buying team: Every new sourcing opportunity evaluated with an AI-generated scorecard alongside the commercial negotiation. FX risk and Temu competitive exposure are visible before the commitment is made.
For Craig Faulkner: A picture of the business he has never had. Category-level margin performance. Store-level inventory health. Competitive position vs Temu by product segment.
Revenue Impact (Conservative Estimate)
| Source | Value | Basis |
|---|---|---|
| FX hedging (Stage 2) | NZD $200K–600K/year | 25–75% protection of $800K–2.4M exposure on average 5–10% NZD move |
| Better buying decisions | NZD $100K–300K/year | Avoided clearance buys in Temu-displaced categories |
| Reduced markdown | NZD $80K–200K/year | Earlier identification of slow-moving stock |
| CFO time reclaimed | 40+ hours/year | From reporting + FX modelling automation |
| Total conservatively | NZD $380K–1.1M/year | First full year of target model operation |
Model Transition Path
TODAY MONTH 6 MONTH 18
Relationship-dependent FX-hedged Intelligence-led
Manual FX management → Automated exposure model → Real-time buying intelligence
No competitive intel Temu category monitoring Integrated buying dashboard
CFO assembles reports Reports auto-generated CFO edits, not builds
15-store North Island Same footprint, better margin Platform for growth decision
Exit Optionality
Current Exit Profile (No AI)
Acquirer type: Larger NZ discount retailer, private equity, or Australian retail group.
Valuation basis: Revenue multiple on thin margins. Relationship-dependent buying advantage discounted heavily — it is not transferable without the people.
Multiple estimate: 0.3–0.5x revenue (NZD $6–20M) — discounted for private company, thin margins, and key person risk.
Target Exit Profile (With AI — 18–24 months)
Acquirer type: Same universe, plus strategic acquirers who want the buying intelligence system.
Valuation basis: EBITDA multiple on improved margins. Intelligence infrastructure valued as a replicable system asset.
Multiple estimate: 4–6x EBITDA — standard for a systematised, growing retail operation with documented processes and improving margins.
Exit improvement from AI investment: If the AI investment costs NZD $50–80K and improves EBITDA by NZD $400K/year, the exit multiple improvement alone (4x EBITDA × $400K) = NZD $1.6M additional exit value. Against an $80K investment, that is a 20x return at exit — before the annual operational value is counted.
The Classification in One Sentence
Current: A good buyer's business. Strong sourcing, thin processes, unmanaged financial risk, no competitive intelligence.
Target: An intelligent buyer's business. The same sourcing advantage, with a systematic model that protects the margin, monitors the threats, and makes the business valuable beyond its relationships.
The transformation does not change what Crackerjack is. It changes how durable it is.