Greenbox is a produce-box startup with 200 subscribers in Perth, racing to reach 1,000 within six months. They’ve discovered their customers care more about convenience than local sourcing – and now they need to figure out whether the business model actually works at scale.
Maya has a board meeting in three weeks. The agenda: present a credible path from 200 to 1,000 subscribers. If she can’t make the case, the money stops.
She’s been working on the pitch deck in the evenings. Good slides. Compelling narrative. But on Wednesday morning, she stares at slide nine – the financial projections – and realises she’s been avoiding the hard question. Not “can we grow?” but “can we grow profitably?”
Reaching for the familiar
Tom suggests Impact Mapping. The team spends thirty minutes on it. Useful – it shows the path to 1,000 involves both reducing churn and expanding acquisition. But Maya shakes her head.
“This tells me how to grow. It doesn’t tell me whether we can afford to.”
Lee recognises the gap. “What you need is a picture of the whole machine – how money comes in, where it goes out, and whether the engine runs at the scale you’re targeting. The Business Model Canvas maps that out. We can do that this morning.”
He pauses. The team is watching him. Lee has been their guide through the entire discovery journey. He’s the person who always has the next technique, the calm voice that says “let’s try this.”
“What I can’t do,” Lee says, “is read it for you once it’s mapped. CAC, lifetime value, what your moat looks like against a competitor with sixty times your funding – I’ve been around those questions, I’ve never run a subscription business through the wall they put up. If I try to interpret the canvas for you, I’ll be doing exactly what we tell teams not to do: guessing at the answers instead of finding someone who knows.”
The room is quiet. It’s a harder thing to say than it sounds. Admitting a limit feels like stepping off a cliff. But he feels something he hasn’t felt in twenty years of consulting: relief.
His phone buzzes in his pocket. He glances at it – a text from Yuki: Dad, can you call me this weekend? He puts the phone away. Maya notices.
“You can take that,” she says.
“She’ll call back,” Lee says.
Maya looks at him. “Will she?”
Lee doesn’t answer. He turns back to the whiteboard.
“We’ll map it this morning. Then I’m going to call someone. Charlotte Wong – she’s scaled two subscription businesses past Series A. Once we’ve got the picture, she can read it.”
What a Business Model Canvas is
The Business Model Canvas was created by Alexander Osterwalder. Nine building blocks on a single page describing how a business creates, delivers, and captures value.
The power is that it forces everything onto one page. The connections – and contradictions – become visible.
Filling it in
Lee facilitates. The team takes a morning. Maya brings the business knowledge, Sam brings customer data, Tom and Priya bring operational reality, Jas brings the product perspective. Tom jokes about buying shares in 3M. Nobody laughs, which tells you something about the mood.
Customer Segments: Two segments from the JTBD and assumption mapping: Convenience seekers (60%) who hire Greenbox to eliminate dinner stress, and Local food advocates (40%) who believe in supporting local farms and eating seasonal produce.
Value Propositions: For convenience seekers: “Dinner decided.” For local advocates: “Know your farmer.” Maya writes both on the board and steps back. “We’ve been marketing one value proposition to two segments. That’s a problem.”
Channels: Word-of-mouth (31%), Google search (28%), Instagram (19%), local press (14%). Delivery via local courier. Customer communication by email.
Customer Relationships: First-box discount for acquisition. Recipe cards, pause/skip, box preview emails for retention. Referral programme for growth.
Revenue Streams: $25/week subscription. Potentially $20/week for a mixed-sourcing box.
At 200 subscribers, all on the $25 box: $5,000 per week. $260,000 per year. Sounds decent.
But Maya hasn’t looked at the other side yet.
Cost Structure:
This is where the room goes quiet. Maya pulls up the numbers on the projector. She hasn’t shared them with the full team before.
| Cost component | Per box |
|---|---|
| Produce (farm gate price) | $14.00 |
| Packing (materials + labour) | $3.50 |
| Delivery (courier) | $4.50 |
| Total variable cost | $22.00 |
Revenue per box: $25.00. Margin per box: $3.00.
Tom does the arithmetic. “Three dollars margin per box. Two hundred boxes a week. That’s $600 a week. $31,200 a year.”
The room is silent. $31,200 doesn’t cover Maya’s salary, let alone the team, warehouse rent, software costs, or marketing.
“What about at 1,000 subscribers?” Priya asks.
Maya updates the spreadsheet. Some costs improve with volume. Produce costs are relatively fixed – farms don’t offer bulk discounts at this scale.
| Cost component | Per box (at 1,000) |
|---|---|
| Produce (farm gate price) | $13.00 |
| Packing (materials + labour) | $2.50 |
| Delivery (courier, volume rate) | $3.50 |
| Total variable cost | $19.00 |
Margin per box: $6.00. $6,000 per week. $312,000 per year. Barely covers operations. No money for growth.
Tom stares at the projector. “We’re building a charity.”
The two-tier question
The canvas is showing contradictions. 60% of subscribers would accept mixed sourcing at $20. But the cost structure assumes 100% local at $25. Maya is paying the premium for local produce, but the majority of her customers wouldn’t notice if she didn’t.
“What if we offered the mixed-sourcing box?” Jas asks.
Maya runs the numbers. If produce cost drops to $8 per box with mixed sourcing:
| Model | Revenue/box | Cost/box | Margin/box |
|---|---|---|---|
| 100% local, $25 | $25.00 | $19.00 | $6.00 |
| Mixed sourcing, $20 | $20.00 | $14.00 | $6.00 |
The margin per box is the same. But the subscriber ceiling changes. With a $25-only model, the addressable market is the 40% who value local sourcing enough to pay the premium. With two tiers, the team can serve both segments.
Priya adds: “And mixed sourcing means we’re not dependent on local farms scaling up in six months. Dave told you he can’t increase supply until next growing season.”
Charlotte
Lee sets up a video call for Friday. Charlotte Wong joins from her home office in Perth’s northern suburbs. She’s 41, short grey hair, a bookshelf behind her stuffed with business books and, inexplicably, a small collection of wooden ducks.
Charlotte grew up in Penang, Malaysia. Moved to Australia at fifteen. Engineering degree from UNSW, then a career in the specific kind of companies that either scale or die: a meal kit company, a SaaS platform, a logistics startup. The SaaS platform was acquired. The logistics startup is still running. The meal kit company – the one she doesn’t talk about unless you ask directly – folded eighteen months after she joined. She’d done everything correctly, or thought she had. The unit economics were wrong from the start and nobody caught it until the cash ran out. She keeps a spreadsheet of every business she’s ever worked with. Row 47 is Greenbox. She added it yesterday, after Lee’s call.
Lee gives Charlotte a ten-minute summary. He shares the canvas. Charlotte listens without interrupting. Her face is still – not hostile, diagnostic. She’s reading the canvas the way a mechanic reads an engine.
Then she asks three questions.
“What’s your customer acquisition cost?”
Silence. Nobody knows.
“You don’t know,” Charlotte says. “That’s the most important number in a subscription business. If you can’t tell the board what it costs to acquire a customer, you can’t tell them whether growth is profitable or just expensive.”
“What’s your subscriber lifetime value?”
Maya starts: “Well, the average subscriber stays for…” She trails off.
“At 5% monthly churn, average lifetime is about twenty months,” Charlotte says. “At $25 a week, that’s roughly $2,000 lifetime revenue. Minus variable costs, about $480 lifetime margin at 1,000 subscribers. If your acquisition cost is more than $480, you lose money on every subscriber you add. Growth makes you poorer, not richer.”
She says this without emotion, but behind the flat tone is the meal kit company. They’d grown to 4,000 subscribers before anyone realised the CAC was higher than the lifetime margin. She’s never fully stopped carrying that one.
“One more thing. Freshly charges eighteen dollars a week. You charge twenty-five. They have sixty times your funding and a polished app. If your customers are convenience-driven – and your JTBD data says sixty percent are – and Freshly delivers convenience at a lower price with better technology, what’s your moat?”
Nobody answers. Charlotte doesn’t wait for one.
“Your canvas shows two segments. Have you modelled what happens to your farm relationships if you introduce mixed sourcing? If 60% of subscribers switch to the mixed box, your local farm orders drop by 60%. Dave and Rachel are suddenly selling you 40% of what they used to. Can their businesses survive that?”
Nobody had considered this. Charlotte saw the dependency that the canvas made visible – changing the cost structure could destroy the partnerships.
“I’m not saying the mixed box is wrong. I’m saying you need to model the second-order effects. You need to bring your farms along, or you’ll have a cheap box with no story and an expensive box with no supply.”
Maya writes furiously. Charlotte winds up the call.
“Lee told me about the discovery work. Event Storming, JTBD, assumption mapping. That’s genuinely impressive for a team this size. Most startups your stage are still arguing about what the product should be. You know your domain and your customers. That’s rare.” She pauses. “The next problem is different. You need to know whether the business works, not just the product. I can help with that.”
After the call, Charlotte sits in her home office. She picks up her phone and calls James.
“How was it?” he asks. She can hear the boys arguing in the background.
“I just told a founder her business model doesn’t work. The look on her face.”
“Is the business worth saving?”
Charlotte thinks about Maya’s eyes when the $3 margin appeared on the projector. Not defeat – recognition.
“I think so. But she has to decide that, not me.”
She opens her spreadsheet. Row 47. In the “First Impression” column: Strong discovery culture. Broken unit economics. Founder identity tied to local sourcing – biggest risk is emotional, not financial.
Maya’s draft
That night, Maya sits at the kitchen table in Fremantle. Nadia is in the other room reading. The house is quiet. Maya opens her laptop and starts a new email.
Dear Greenbox subscribers,
We’ve made the difficult decision to pause operations while we reassess our business model to ensure we can continue to deliver the quality you expect.
She reads the three sentences back. They’re corporate and bloodless and they sound nothing like her. She imagines Mrs Patterson reading them. She imagines Patrick reading them. She imagines Dave reading them and thinking: Another one.
She doesn’t delete the draft. She doesn’t send it. She closes the laptop.
Nadia appears in the doorway. “Come to bed.”
“Coming.”
She doesn’t tell Nadia about the email. She doesn’t tell anyone. The draft sits in her email, unsent, for the next six months.
When to use a Business Model Canvas
- Preparing to pitch investors. The canvas forces you to think about the whole business, not just the product.
- Considering a significant business model change. Launching a new tier, entering a new market – the canvas shows second-order effects.
- Post-revenue, pre-profitability. When the product works and people pay, but the model might not sustain itself.
When not to use it
- When the problem is execution, not strategy. If deliveries arrive late, fix logistics. The canvas is for strategic clarity.
- When you need detailed financial modelling. The canvas shows what the cost structure looks like. For exact numbers, you need a spreadsheet.
What comes next
Maya has three weeks to prepare her board pitch. She has JTBD data, validated and invalidated assumptions, the canvas, and Charlotte’s framework for calculating the numbers that matter.
She’s also preparing to propose something that would have been unthinkable three months ago: a two-tier product that partially abandons the 100% local sourcing she built the company around. The data says it’s the correct move. Her gut says it’s a betrayal.
Charlotte told her, on that first call: “The founders who scale are the ones who fall in love with the problem, not the solution. You fell in love with local sourcing. Your customers fell in love with not thinking about dinner. Those aren’t the same thing.”
Maya is still thinking about that.
But thinking isn’t a plan. The team has data, frameworks, and broken unit economics. They know what’s wrong. They can’t fix everything at once. The board meeting is in three weeks.
The question isn’t what to change. It’s what changes first.
The next chapter, Prioritisation: What Changes First, publishes around 26 May.