Last July we held 23 orders in our studio for 6 days because the forecast was over 30°C across Ontario. The phone rang on day two — a customer in Ottawa who said she'd been waiting since Tuesday, and asked if we could "just put more ice packs in." She was polite, a little exasperated, completely reasonable. We refunded her within an hour and she ordered again the next week when the temperature dropped. Here's the math, the customer service, and why we don't have a "ship anyway" button.
The problem in numbers
Chocolate melts at 34°C. Above 30°C, even in an insulated box with an ice pack, the temperature inside a shipping parcel climbs past 34°C within about 90 minutes. Bloom — the white, chalky streaks that appear on the surface of chocolate that has been exposed to heat and then cooled — sets in within a few hours of the chocolate being above its melting point. Bloomed chocolate is safe to eat, but it looks terrible. A customer opening a $54 box of bonbons that have white streaks all over them is a customer who will not order again.
Last summer's numbers from the 23 held orders:
- Average delay: 4.2 business days from order to ship
- Refund rate of held orders: 11% (most people waited)
- Repeat-order rate within 60 days of the heat wave: 78% — almost identical to the cool-month rate of 81%
- Complaint rate: 0% from customers we proactively emailed; 100% from the two we didn't reach in time
The 78% repeat-order rate is the key number. Customers who we held their order for, who we proactively told what was happening, came back at almost the same rate as customers who got their boxes on time. The cost of holding was 4 days of customer-service time, two weeks of email back-and-forth, and roughly $400 in unfulfilled product that we had to either donate or eat ourselves.
The policy
From May through September, we do not ship when the forecast high at the destination is above 30°C. We email every affected customer within 24 hours of the temperature crossing the threshold and offer them the choice of: (a) waiting until the temperature drops, (b) upgrading to express overnight with extra insulation for an additional $18, or (c) cancelling the order with a full refund.
Express overnight is the option that surprises people the most. The math: in a 5°C warehouse, packed with four ice packs and an insulated liner, an overnight parcel stays under 18°C for about 6 hours. Even in a 35°C truck dock. We have shipped 14 overnight orders in summers past and exactly one of them arrived bloomed — and that was an ice-pack supplier issue, not a temperature issue. Express overnight is the safety valve for birthdays, anniversaries, and the 2% of cases where waiting is genuinely not an option.
The customer service scripts
There are three emails we send, in this order. Subject lines are deliberate — they tell the customer what the email is about before they open it, so they don't think it's a marketing blast.
Day 0 (heat hits the forecast): subject "your order is safe with us, but we need to talk about timing." Body: confirms the order, explains the heat policy in 2 sentences, asks which of the three options they want, gives a specific date range when shipping will be possible again, and signs off with Aline's name. No automated reply-to address — the customer can reply and it goes to a real person.
Day 3 (if no response): subject "still holding your order — quick question." Body: 2 sentences reminding them of the choices, no pressure, no hard sell. If they want to keep waiting, they can reply "wait" and we keep their slot. If they want to cancel, they reply "cancel" and we refund same-day.
Day 7 (final check): subject "your order is going back to the kitchen — let us know if you want it." Body: explains that we're going to make a final decision on their order at end of day — refund, hold, or send (depending on forecast). The "send" option is reserved for cases where the customer has specifically said "ship it anyway, I understand the risk" — which has happened maybe 6 times in 3 years, all for wedding favours where the date was immovable.
Why not just ship anyway with more ice packs?
Because it doesn't work past about 4 hours. The ice packs are good, the insulated liner is good, but the parcel sits in a delivery truck for 6-12 hours, often in direct sun, before the customer opens it. You can get a parcel from Toronto to Ottawa in 4 hours by express, but you cannot keep a parcel below 18°C in a non-refrigerated delivery truck for that long in a Canadian July. The physics is against you.
We've considered refrigerated trucking. It's 8x the cost of standard shipping and would add $40 to every summer order. For most customers, the "wait 4 days" option is fine and costs them nothing. Refrigerated trucking would be solving a problem for the 2% of cases where waiting is impossible, at the expense of every customer in the queue. Not worth it.
What we'd do differently if we were a bigger shop
If we were shipping 500 boxes a day instead of 12, the answer might be different. At volume you can negotiate refrigerated rates, you can run your own delivery route in the GTA, you can justify a cold-chain fulfillment partner. At our size — 12 boxes a day average, 35 on a busy December day — the math doesn't work. The hold-and-email approach is what a small atelier can do, and customers seem to genuinely appreciate the honesty.
The 78% repeat-order rate is what I'd point to if someone asked whether the policy is worth it. It's not just a number — it's the difference between a customer who feels looked-after and a customer who feels ignored. We chose the first one. We'd choose it again.