Adelaide is live. The playbook works. The hiring pipeline is running. But the next expansion – Sydney – hits a wall that no playbook can solve. The promise that made GreenBox is the same promise that might stop it going national.
Dave Morrison has driven a lot of kilometres for GreenBox. The Margaret River farms. The Perth ring. Melbourne’s Yarra Valley producers. Brisbane’s hinterland growers. He’s put his hands in the soil of forty-two farms across three states. He can read a paddock the way some people read a balance sheet – what’s growing, what’s struggling, what the farmer isn’t saying about last season’s yield.
Adelaide was different, and he told Maya as much during the farm visits. The distances were wrong. The crop mix was narrower. The hub-and-spoke logistics model they invented during the Adelaide playbook was a good solution to a problem Perth never had. But Adelaide was still manageable. Seven farms, one hub, a workable radius.
Sydney is where the promise breaks.
The fifty-kilometre radius
GreenBox was built on a promise. Everything in the box comes from within fifty kilometres of the subscriber’s door. Maya made that promise sitting at the kitchen table in the Fremantle cottage three years ago, sketching the first version of GreenBox on a notepad while Nadia cooked pasta. It was the founding constraint. The thing that made GreenBox different from Freshly, different from wholesale delivery services, different from the supermarket. Local farms. Short supply chains. Produce that was in the ground yesterday and in the box tomorrow.
In Perth, the promise was easy to keep. Perth sits in the middle of one of Australia’s most productive agricultural regions. Fourteen farms within fifty kilometres. Dave’s farm in Margaret River is further – about two hundred and seventy kilometres – but the Perth ring farms are close. The original supply chain was built around a city that happened to be surrounded by arable land.
Melbourne worked because the Yarra Valley and Mornington Peninsula have dense farming communities. Brisbane worked because the hinterland is fertile and the farms are used to supplying local markets. Adelaide was harder – seven farms instead of fourteen, and the hub-and-spoke model was necessary – but the fifty-kilometre promise held.
Sydney has three farms within fifty kilometres. Three.
Dave discovers this on the second day of Phase 2. He and Maya are driving through the western suburbs, past the blue mountains’ foothills, looking for the produce farms that the playbook’s market assessment identified. Two of the addresses turn out to be hobby farms – small plots with a couple of vegetable beds, not commercial operations. One is a commercial farm, but it grows only Asian greens and supplies restaurants in Cabramatta.
“This isn’t a farming area,” Dave says. He’s standing in a car park in Penrith, looking west toward the mountains. “This is a city that ate its farms.”
He’s right. Sydney’s urban sprawl consumed most of the agricultural land that once surrounded the city basin. The farms that remain are either specialised (market gardens serving specific restaurant clients) or small (peri-urban operations on land that’s more valuable as real estate than farmland). The broad-acre mixed produce farms that GreenBox depends on – the kind Dave runs, the kind Kath runs in Adelaide – are eighty to a hundred and fifty kilometres from the city centre. The Central Coast. The Hawkesbury. The Southern Highlands.
The fifty-kilometre promise can’t work in Sydney. Not because there isn’t produce. Because the produce isn’t close.
The dilemma returns
Maya sits in the rental car in the Penrith car park and stares through the windscreen at the mountains. She knows this feeling. She’s had it before.
The two-tier pricing model – local boxes at $25 a week, mixed-source boxes at $20 – was born from a crisis in Series 3. When the Business Model Canvas work revealed that the original all-local model couldn’t sustain the business at scale, Maya faced a choice: stay local and stay small, or expand the sourcing and grow. It nearly broke her. She drafted the “pausing operations” email. She questioned everything she’d built. Charlotte and Lee and the JTBD insights pulled her through.
The two-tier model was the compromise. Local boxes for subscribers who valued provenance. Mixed-source boxes for subscribers who valued convenience and price. Both tiers kept the direct farm relationships. Both tiers delivered the same promise: dinner sorted, every week.
Now the dilemma is bigger. The two-tier model assumed that every city would have enough local farms to fill the local tier. Adelaide’s seven farms can just about manage it – they’re stretching, but the hub-and-spoke logistics make it work. Sydney’s three farms can’t. Not enough variety. Not enough volume. Not enough of the basic produce that fills a box: tomatoes, lettuce, zucchini, potatoes, onions, carrots.
Maya calls Charlotte from the car. Dave is outside, leaning against the bonnet, calling Helen.
“Sydney can’t do local-only,” Maya says. “We don’t have the farms.”
Charlotte is quiet for a moment. “What do you have?”
“Three farms within fifty kilometres. Two are viable – one Asian greens specialist and one mixed vegetables. Maybe eighty boxes a week between them.”
“How many subscribers are you targeting for launch?”
“Two hundred.”
“So you’re short by more than half.”
“For local-only, yes. If we go mixed-source – combining local and regional farms within a hundred and fifty kilometres – we can probably fill it.”
“Then Sydney launches as mixed-source only?”
“That’s what I need to decide.”
Dave’s assessment
Dave comes back to the car. He’s been on the phone with Helen for twenty minutes – longer than usual. He looks the way he looks when he’s thinking about something he doesn’t want to say.
“Talk to me,” Maya says.
Dave gets in the passenger seat. He doesn’t start talking immediately. He watches a truck pass on the road, carrying pallets of something under tarps.
“You can’t run this the way you run Perth,” he says. “The distances are wrong. The crops are different. The growing seasons – the coast is three weeks behind the inland. The water allocation is different. The soil is different.”
He pauses. “Perth works because Perth is a farming city that happens to have people in it. Sydney is a city that happened to push out its farms. You’re not going to find fourteen farms in fifty kilometres because they don’t exist anymore. The developers turned them into houses.”
Maya waits. Dave usually gets to the point on his own schedule.
“But that doesn’t mean there aren’t good farms. There are excellent farms on the Central Coast. Ninety minutes from the city. And the Hawkesbury – that’s an hour and a half, good soil, family operations. The Southern Highlands have people doing incredible things with regenerative grazing and mixed cropping.”
“They’re all outside fifty kilometres.”
“They’re all outside fifty kilometres.”
Dave turns and looks at her. His expression is the one he had at the first Event Storm – patient, slightly amused, waiting for the city person to catch up.
“Maya. The fifty-kilometre promise was about trust. It was about knowing where the food comes from. About the farm having a name and a face. Mrs Patterson trusts the box because she knows Dave grew the zucchini. She doesn’t know it’s fifty kilometres. She knows it’s Dave’s.”
Maya opens her mouth to argue and stops. Dave has just described something she’s been treating as a rule – fifty kilometres – that is actually a proxy for something else: trust. Traceability. Relationship. The distance was never the point. The distance was the mechanism that ensured the point.
“Are you saying the fifty kilometres doesn’t matter?”
“I’m saying I don’t know if it matters. I know that when a woman opens her box on Thursday and sees the farm’s name on the card and knows that someone she trusts picked those tomatoes, she doesn’t get out a measuring tape.”
The three-tier model
Maya, Dave, Charlotte, and Diane spend two days in a Sydney hotel meeting room working through the options. It’s the kind of session that used to happen around a kitchen table with sticky notes. Now it’s four senior people in a room that costs $400 a day, with a whiteboard, a laptop, and a projection screen.
Charlotte leads the structure. She draws a table on the whiteboard – the same format she’s used since the decision table work in Series 4, because decision tables are the way GreenBox thinks about rules now.
| City type | Farm radius | Box tier | Price | Promise |
|---|---|---|---|---|
| Dense farming (Perth) | <50km | Local | $25 | All produce from named local farms |
| Moderate farming (Melbourne, Brisbane, Adelaide) | <100km | Local | $25 | All produce from named regional farms |
| Sparse farming (Sydney, future) | <150km | Regional | $22 | Produce from named farms within the region |
Maya stares at the table. “That’s three tiers.”
“It’s two,” Charlotte says. “Local and Regional. The difference is the radius. And you need a third option.”
She adds a row:
| Any city | Any radius | Mixed | $20 | Curated produce, sourced for quality |
“Four tiers?”
“Three meaningful ones. Local ($25) where you have the farm density. Regional ($22) where you don’t. Mixed ($20) everywhere, same as now.”
Diane leans forward. “The question isn’t the tiers. The question is: what do you call it? How do you explain to a Sydney subscriber that their box is different from a Perth subscriber’s box?”
Jas, on a video call from Perth, has been listening. “You don’t explain the difference. You explain the similarity. Every box comes from farms we know. Every box has the farm’s name on the card. Every box is curated for the subscriber. The radius changes. The relationship doesn’t.”
Maya looks at the screen. “Say that again.”
“The radius changes. The relationship doesn’t. That’s the brand. GreenBox isn’t about kilometres. It’s about trust. Perth subscribers trust their box because the farm is twenty minutes away. Sydney subscribers can trust their box because the farm is ninety minutes away and we still know the farmer’s name.”
Charlotte writes on the whiteboard: Trust is the product. Distance is a feature.
The substitution engine
The three-tier model has technical implications. Tom joins the call from Perth to walk through them.
“The substitution engine is parameterised by city,” he says. “That was the architecture decision from the DDD work. Each city has its own decision table rules. Perth’s rules assume year-round tomatoes. Adelaide’s rules handle the three-week gap. That part works.”
“So what’s the problem?”
“The problem is that the current parameterisation assumes one tier per city. Perth is local-only. Melbourne is local and mixed. Brisbane is local and mixed. But Sydney needs three tiers in one city: local for the three nearby farms, regional for the Central Coast and Hawkesbury, and mixed for everything else.”
He pauses. “The substitution rules are different for each tier. A local box can only substitute with other local produce. A regional box can substitute with anything from the regional radius. A mixed box can substitute with anything. Right now, the engine doesn’t know about tiers. It knows about cities and produce availability. We need to add a tier dimension.”
“How long?”
“Two weeks for the engine changes. Another week for the decision table updates. Plus testing – the Example Mapping for the new rules will take a few sessions because the combinations are more complex.”
Charlotte raises the architectural point that nobody has said out loud yet. “This is the first time the substitution engine needs to handle a variable that’s not about produce or preference. It’s about service level. You’re adding a business model concept to what was a pure logistics engine.”
Tom nods. “I know. It’s the right decision architecturally – the tier is a property of the subscription, not the supply chain. But it means the subscription context and the supply matching context need to talk to each other in a way they haven’t before. The bounded contexts are still correct – subscription and supply matching should be separate. But the interface between them needs to carry more information.”
“Is this a redesign?”
“No. It’s an evolution. The boundaries are right. The contract between the boundaries needs to grow. That’s healthy. That’s what bounded contexts are for – they evolve independently, and the contracts between them get richer as the business gets more complex.”
Priya, also on the call, adds: “I’ll write the contract tests. This is exactly the kind of change that broke things during Two Squads. If we define the new contract between subscription and supply matching as a set of tests, we’ll catch it before it reaches production.”
Tom: “Agreed. Contract tests first. Then the engine changes. Then the new decision tables.”
Charlotte looks at the whiteboard. The technical discussion is over in ten minutes. In Series 1, a change like this would have taken two weeks of debate. The team has a shared language now – bounded contexts, contract tests, decision tables – that makes architectural decisions faster because everyone understands the concepts and the constraints.
Mrs Patterson’s opinion
Maya flies back to Perth on Friday evening. She’s exhausted – the kind of tired that comes from making decisions that feel like compromises. The three-tier model is the right answer. She knows it intellectually. But it feels like another step away from the founding vision. First the two-tier model. Now three tiers. The radius expanding from fifty kilometres to a hundred and fifty. The local promise stretching.
Sunday morning. Maya is at the South Fremantle farmers’ market – the one where she first met Lee, where the GreenBox story started. She’s buying bread from a stall she’s been coming to for years. And there, at the next stall, browsing the organic honey, is Mrs Patterson.
Mrs Patterson – the Stirling Highway subscriber. Loyal since week two of the pilot. Hates beetroot. Has Maya’s personal mobile number. “I’ve never met any of you,” she said during the continuous discovery interviews, “but I feel like you know me.”
“Mrs Patterson,” Maya says. “It’s Maya. From GreenBox.”
Mrs Patterson’s face lights up. “Oh! The box girl! I didn’t recognise you without the – well, I’ve only seen you in emails, haven’t I?”
They talk for ten minutes at the market. Maya asks how the boxes have been. Mrs Patterson says they’ve been lovely. “That new recipe card for the roasted cauliflower was excellent. Nadia’s recipe?”
“Jas found it, actually.”
Maya hesitates. Then she asks the question she came to the market hoping not to ask but needing to ask.
“Mrs Patterson, do you know where the produce in your box comes from?”
“From the farms, dear.”
“Do you know how far away the farms are?”
Mrs Patterson considers this. “Somewhere nearby, I assumed. The card says the farm name. Last week it was Morrison Farm. The week before it was – oh, I don’t remember. I trust what’s in the box, dear. I don’t check the labels.”
Maya stands in the Saturday morning sun at the farmers’ market and hears the thing she needed to hear. Not from a JTBD interview. Not from a decision table. From the subscriber who has been the emotional anchor of GreenBox since week two.
I trust what’s in the box. I don’t check the labels.
The fifty-kilometre promise was Maya’s promise to herself. A way of honouring her parents’ farm, the local-first ethic, the belief that food should come from nearby. Mrs Patterson didn’t buy a radius. She bought trust.
The Fremantle waterfront
That evening. The bench near the fishing boat harbour. The same bench where Maya and Ren have sat through every crisis – the JTBD gut punch, the near-shutdown, the Series B question, the partnership dilemma.
The sun is going down. The harbour smells the way it always smells: salt, diesel, and the ghost of fish and chips from the takeaway on the corner. Ren is already there. Thermos of tea. The same bench.
Maya sits down. She doesn’t say anything for a while. The light is doing that Fremantle thing – turning everything gold and long-shadowed.
“I promised local,” she says. “Within fifty kilometres. That was the whole point.”
Ren pours tea. She doesn’t answer immediately. She never does.
“The point was that a woman on Stirling Highway doesn’t have to think about dinner on Tuesday,” Ren says. “Does she care if the zucchini came from fifty kilometres or two hundred?”
“Mrs Patterson would care.”
“Have you asked her?”
Maya takes a sip of tea. She looks at the harbour. A pelican is sitting on the same pylon as last time – or a different pelican, doing the same pelican thing.
“I asked her today. At the market.”
“And?”
“She said she trusts what’s in the box. She doesn’t check the labels.”
Ren lets that sit.
“So the question you need to answer,” Ren says, “is: are you keeping the fifty-kilometre promise for Mrs Patterson? Or for yourself?”
Maya knows the answer. She’s known it since the car park in Penrith when Dave said the same thing in different words. The fifty-kilometre radius was her way of staying connected to her parents’ farm. A concrete rule that encoded an emotional truth: food should come from nearby, from people you trust, from land you can see.
But the emotional truth doesn’t require the concrete rule. Trust doesn’t have a radius. Relationship doesn’t have a kilometre limit. Dave’s farm is two hundred and seventy kilometres from Perth and his zucchini is in Mrs Patterson’s box every week. She trusts it because it’s Dave’s, not because it’s close.
“For myself,” Maya says.
“Is that wrong?”
“No. But it can’t be the reason the company doesn’t grow.”
Ren smiles – the barely-there one. “That might be the most grown-up thing you’ve ever said.”
“Don’t tell Nadia. She’ll think I’m getting boring.”
They sit until the light is gone and the harbour lamps come on. The tea is finished. The pelican has flown somewhere. The fishing boats are in.
The decision
Monday morning. Maya walks into the leadership meeting and puts the three-tier decision table on the screen.
“Sydney launches with three tiers. Local for the farms we have within fifty kilometres – that’s two farms, and I want to build that number over time. Regional for the Central Coast, Hawkesbury, and Southern Highlands farms – that’s the main supply. Mixed for everything else.”
She pauses. “The fifty-kilometre promise stays for cities where we can keep it. Perth, Melbourne, Brisbane. For new cities, the promise evolves. We promise named farms, traceable produce, and a relationship between the grower and the subscriber. The distance is whatever the local geography requires.”
Tom nods. Priya nods. Sam nods. Charlotte makes a note. Diane says nothing – she said everything she needed to say in the hotel room.
Dave, on the phone from Margaret River, says: “Good. Now let me tell you about the Central Coast farms. I visited three last week. Two are real.”
“You went to Sydney without telling me?”
“Helen and I were visiting Claire. I took a detour. The farm in Ourimbah is excellent. The one in Kulnura is smaller but the soil is good and the farmer’s young. He’ll grow with you. The third one – forget it. Too much marketing, not enough product.”
Maya laughs. “You’re writing the playbook without knowing it.”
“I don’t know what that means. But I know good soil when I see it.”
What the supply chain looks like now
By June, the national supply chain architecture is taking shape. Charlotte and Sam design the operating model with input from Dave, Diane, and the city leads.
Hub-and-spoke model (all cities except Perth). Each city has a logistics hub – a rented cold storage facility where farms deliver produce and where boxes are packed. Perth keeps the direct-collection model because the farm density supports it. Every other city uses hubs. The hub adds cost (rent, staff, cold chain equipment) but enables the wider farm radius.
Tiered supply by city. Each city is classified based on farm density within fifty kilometres:
- Dense (Perth): two tiers, local and mixed
- Moderate (Melbourne, Brisbane): two tiers, local and mixed
- Sparse (Adelaide, Sydney, future cities): three tiers, local, regional, and mixed
- The classification is in the playbook. Phase 1 market assessment determines the classification.
Farm partnership network. Dave’s assessment criteria are documented – finally, reluctantly, with Dave grumbling that “you can’t put soil quality in a spreadsheet.” The criteria aren’t a spreadsheet. They’re a checklist:
- Commercial scale (minimum 100 boxes of produce per week)
- Reliability (can commit to a delivery schedule)
- Quality (Dave’s visit or, in future, a trained farm partnership manager)
- Communication (willing to use the farm portal for order management)
- Values alignment (interested in the GreenBox model, not just the revenue)
“That last one is the most important,” Dave says. “And it’s the one you can’t measure. Either they care about what you’re doing or they’re selling to whoever pays. I’ve seen both.”
Substitution engine. Tom and Priya ship the tier-aware substitution engine in two weeks, as promised. The contract tests catch three edge cases before they reach production: a local box that was receiving regional substitutions, a regional box that wasn’t considering seasonal availability from coastal farms, and a mixed box that was being over-constrained by local-only rules that didn’t apply.
The decision tables are now multi-dimensional. Each rule considers: city, tier, season, produce category, subscriber preferences, and farm availability. It’s more complex than the original Perth rules. But the decision table framework handles the complexity because each dimension is independent. You don’t need to understand all the dimensions to update one. Adelaide’s seasonal adjustments can change without affecting Perth’s preference matching.
Charlotte, reviewing the updated architecture, says to Tom: “This is what the DDD work was for. You designed the system to handle complexity you hadn’t seen yet.”
Tom, in a rare moment of unguarded satisfaction: “I know.”
The hook
Adelaide is working. Two hundred and forty subscribers, climbing. The hub-and-spoke model is efficient. The farms are delivering. The Friday delivery schedule has become part of the Adelaide team’s identity – they call themselves “the Friday crew,” and Kath brings pastries from a bakery near her farm for the packing team.
Sydney is next. The three-tier model is ready. Dave has identified ten farms across the Central Coast, Hawkesbury, and Southern Highlands. The playbook’s Phase 1 is complete: the market assessment confirmed the job, the demographics support the pricing, and the competitive landscape is – complicated.
Because Sydney already has a produce box company.
Not a big one. Eight people. Three thousand subscribers. A different tech stack (Shopify and spreadsheets), different culture (scrappy, founder-led, no process to speak of), different sourcing model (wholesale markets, not farm-direct). They’ve been running for two years. Their subscribers are loyal in the way that early subscribers always are – they remember when the boxes were hand-packed in the founder’s garage.
They’re called Sprout Box. And they’re exactly the kind of company GreenBox was three years ago.
Diane sees an opportunity. A company with three thousand Sydney subscribers, established delivery routes, local brand recognition, and a team that knows the market. An acquisition could cut Sydney’s Phase 2 and Phase 3 from eight weeks to four. Instead of building from scratch, buy the beachhead.
Charlotte sees a risk. A different tech stack means integration work. A different culture means management work. A founder-led company with no process means everything that GreenBox has spent three years building – decision tables, bounded contexts, structured hiring, the planning cadence – is absent. Acquiring Sprout Box is faster. Integrating Sprout Box is harder.
“Buying a company is easy,” Charlotte says. “Making it part of yours is the hardest thing you’ll do.”
She’s speaking from experience. The meal kit company – the one that failed, the one that haunts her. It was acquired by a larger company six months before it folded. The integration destroyed what was left. Charlotte watched it happen. She knows what bad integration looks like.
“And not buying them,” Diane says, “means competing with them. In their home market. While they have three thousand subscribers and you have zero.”
Maya looks at them both. Charlotte’s caution. Diane’s velocity. The tension that has defined this entire series, compressed into a single decision.
“Tell me about Sprout Box,” Maya says.
That’s the next conversation (coming 19 January). And it’s the first time GreenBox considers growing by buying, not building.