🎯 What we are actually testing

Not revenue. The question is whether households order more than once, and whether an order left in a box is still excellent when they open it.

One household ordering four times tells us more than four households ordering once.

1 The offer — what neighbors hear

We are not competing with Grab / LINE MAN. They refuse trips this short — too little money for the time. We are filling the gap they won't serve, for neighbors who don't want the walk.

2 The kit

• 1 × large "mother" cooler in the vehicle, packed with ice. Branded and visible — this is our advertising.
• 10 × small coolers that live permanently at the customer's gate, like a mailbox. ฿100–150 refundable deposit.
• Ice packs (2 per order), sealed bottles, separate topping cups, branded stickers.

Why stationed boxes: we never carry empty coolers out or back, never make a collection trip, and never wash between customers. We carry one big cold box of orders and drop into the box already at each house.

3 What can be delivered

4 The run

1. Orders close at 2:00 PM.
2. Blend and freeze the smoothies. Pack bottles + ice packs into the mother cooler.
3. 4:00 PM — one lap of the village.
4. Regulars: open their box, drop it in, close it. New customers: hand delivery + offer them a box.
5. No return trip. Ever. Anything to collect rides the next day's run.

5 "If you're not home" — say it up front

One required question on the order form, then repeated back in the confirmation so there is never a surprise:

Being late is normal on a route. The box is what makes a set time survivable — it turns "4:00 PM sharp" into "sometime this afternoon" without anyone being disappointed.

6 How we confirm someone really is in the village

Prepaid removes the reason to lie. Nobody prepays ฿150 for a smoothie to an address they can't receive at. So we are not defending against liars — we're defending against mistakes: wrong village picked, house number typed wrong, someone living just outside the gate.

• The village LINE group is our best check, and it's free. Those groups are admin-vetted — you're in it because you live there. Anyone who comes from the group post is almost certainly a resident.
• Village from a dropdown, never free text. Kills typos and gives us clean data on which villages people are asking for.
• House number as its own field — in a moo baan that's the real address, and it's what we navigate by.
• "Use my location" checks GPS against the village boundary — but only flags, never blocks. Hard-gating on GPS would reject real customers ordering from work, or with location turned off. Green tick when it agrees, silence when it doesn't.
• The first delivery is the real verification. We drive there once. If it's wrong we find out immediately, fix it, and it never happens again.
• Then it's permanent. We already capture their LINE ID on every order, so once an address is confirmed it stays "known good" for that customer forever.

The edge case to decide before we launch — someone 200m outside the gate asking "will you come to me?" That's a route decision, not a software one, and we'll be asked in week one.

📱 What the order form would look like

Rough mock-ups — not built yet, just to see the shape of it.

Mock-up 4 is the expansion engine — it collects village + preferred time, which tells us exactly where to open next and at what hour.

7 Pilot week — before any announcement

Run it for 3–4 neighbors we already know, for one week. Zero reputational risk while we get it wrong.

We are measuring: how long a lap really takes, how cold it still is after 60 minutes in a closed box in the afternoon sun, whether the boxes stay put, and what the packaging needs to be.

9 What we measure (first month)

★ Repeat households

Ordered twice or more — this is the one that matters

Orders per run

5–12 would be a healthy settled number in a 200-door village

Boxes placed

How many households took a stationed cooler

Quality complaints

Target: zero

✅ Already decided

❓ Questions to decide together

💡 Ideas & things not settled