Selling small batch food online

You roasted six kilos on Sunday. It's Thursday, two kilos are still sitting there, and they're not the same two kilos they were on Monday.

Small batch food works when you stop making stock and start making cycles. Pick one production day a week, close orders the night before, make to the orders plus a small buffer, and ship the next morning. Print the date you made it on the bag, not just a best-before. Then build the business on people ordering again rather than on finding new people, because a bag of coffee has a three-week window and a customer has years.

That's the shape. Below is what each part costs.

Put the date on the bag, and make it the make date

A best-before date tells someone when the food becomes questionable. A roast date, a bake date, a batch date tells them when it was at its best, which for small batch food is the thing they're paying you for.

Print it. On the bag, on the label, in the order confirmation. Roasted 4 August. Not "roasted fresh weekly", which is what everyone writes and which means nothing.

Two things happen when you do. Customers who know the category will trust you immediately, because printing a make date is a thing you only do when you're not hiding anything. And customers who don't know the category will learn from you, which makes them yours.

The uncomfortable half of that is that you can no longer sell a bag from three weeks ago at full price to someone who won't notice. Correct. That's the discipline the date imposes, and it's the point of it.

The weekly cycle is the whole operating system

Every small roaster and small-batch maker who runs a calm operation runs the same rhythm, and it turns a spoilage problem into a scheduling problem.

Day	What happens
Thursday to Sunday night	Orders open. People buy for next week.
Sunday 11pm	Orders close. You now know you need 34 bags.
Monday	Production day. Roast, bake, bottle, make.
Monday evening	Rest, cool, degas, label.
Tuesday morning	Pack and ship. Everything leaves on the same day.
Tuesday afternoon	Local pickups and deliveries.

The costs of not doing this are invisible until you count them. Making to order daily means five short production runs a week instead of one long one, five setups, five clean-downs, and five sets of packaging materials broken open. Making ahead means you're guessing, and the guess sits on a shelf getting older.

Publish the cycle on your shop, in one sentence: "Orders close Sunday night. Everything is roasted Monday and ships Tuesday." It sets expectations, it stops the "where is my order" message on a Wednesday, and it quietly tells the customer this is a real operation.

The buffer is the judgement call. Make roughly 10 to 15% over your order count on the things that keep, and nothing extra on the things that don't. The overage covers a spilled batch, a walk-in customer and the person who orders at 6am Monday having missed the cut-off.

Roast loss, batch size and the money that evaporates

Coffee has an arithmetic that catches people out, and it's worth spelling out because the same shape shows up in other products.

Green coffee loses weight in the roast. Water goes off as steam and the chaff blows away, and depending on how far you take it you lose somewhere in the region of 15 to 18% of the mass you paid for. So a kilo of green becomes roughly 830 to 850 grams in the bag.

That means the green price is not your cost. If green costs you $9 a pound, your roasted coffee costs about $10.70 a pound before you've bought a single bag, and if you built your pricing off the green price you're already 16% behind.

Then the rest of it:

• Bag with a one-way valve: roughly $0.35 to $0.60 each in small quantities.
• Label: around $0.10 to $0.15 if you print your own, more if they're done properly.
• Shipping box and filler: $1.00 to $1.50.
• Postage: whatever your carrier actually charges, which is more than you think for a two-bag order.

Batch time is the other constraint. A 1kg sample roaster runs a batch in twelve to sixteen minutes plus cooling, so you're getting three to four batches an hour realistically. Six kilos of green is about two hours at the machine, which produces roughly fifteen 12oz bags. That's your ceiling until you buy a bigger drum, and knowing it stops you promising forty bags on a Tuesday.

The same discipline works for anything else you make in batches. Weigh what goes in, weigh what comes out, and price off the second number. Yield loss on jam, on granola, on chocolate, on rendered anything, is real and it's always higher than the recipe suggests. There's a fuller method for building a price from the actual cost of a made thing in how to price what you make.

Freshness windows, by product

Write your own window down before you decide anything about delivery, because the window sets the shipping options rather than the other way round.

Product	Realistic peak window	Ships well
Filter coffee, valved bag	Day 4 to about day 21 off roast	Yes
Espresso, valved bag	Day 7 to about day 21	Yes
Granola, dry mixes	Four to six weeks	Yes, easily
Cookies, biscotti, brittle	Two to four weeks	Yes
Chocolate	Months, but heat is the enemy	Yes, outside summer
Shelf-stable jars, sauces	Months once the process is right	Yes, once the rules are sorted
Fresh pasta	Three to four days chilled	Only overnight, only with cold packs
Bread	Best the day it's made	Local only
Unpasteurised juice, fresh dairy	Days	Local only, and check what you're allowed to sell

Coffee is worth one extra note, because the window has a front edge as well as a back one. Freshly roasted coffee releases carbon dioxide for several days, which is what the little valve on the bag is for, and coffee brewed the day after roasting often tastes flat or sharp rather than good. Espresso in particular is usually better a week in than a day in. Ship on the day after you roast, tell the customer the roast date, and let them decide when to open it. Do not ship on day five and call it fresh.

Shipping perishables without a cold chain you can't afford

You don't have refrigerated logistics and you don't need it for most of this. What you need is a schedule that doesn't strand food in a warehouse.

Never ship perishable food on a Thursday. It sits in a depot from Friday night until Monday, and you'll be refunding it on Tuesday.

Monday, Tuesday and sometimes Wednesday. That's the shipping window for anything with a short life, and it should be published rather than explained after the fact.

The rest of it:

• Insulate what needs it. A foil liner and a gel pack costs somewhere around $3 to $5 and is the difference between chocolate and a puddle in July. Charge for it as a summer surcharge rather than absorbing it silently.
• Overnight or two-day for anything genuinely perishable. Which usually means the shipping costs more than the product, which usually means the product should be sold locally instead. That's a real conclusion and plenty of makers reach it.
• Bags with valves need to breathe, boxes need to not crush. A 12oz bag of coffee survives almost anything. A jar of sauce with a loose lid is a claim waiting to happen.
• Photograph the parcel before it goes, once, for your own records, if you're shipping anything glass.

The general packing principles that survive a sorting facility, including what actually protects a corner and what only looks like it does, are in packaging that survives the courier.

Local delivery and pickup deserve more attention than they usually get. If your product's window is four days, a farmers market stall plus a pickup window at your unit will outperform national shipping on both margin and quality, and it costs you a Saturday instead of a courier account.

The rules question, and the one category that's different

Food rules in the US are layered and local, and this is the part where getting it wrong is expensive rather than embarrassing.

The broad shape, which you should verify rather than trust: most states have some form of cottage food framework covering food made in a home kitchen, and it usually covers shelf-stable things like baked goods, dry mixes, granola and confectionery, and usually excludes anything that needs refrigeration to stay safe. Whether you can ship across state lines under those rules varies, and it's one of the first things to ask about, because a lot of people build a shipping business on a permit that only allows in-state sales.

Roasted coffee and dry goods generally sit at the lighter end. The heavier end is anything you seal in a jar and describe as shelf-stable. Acidified and low-acid canned foods are the most closely regulated corner of the whole field, for good reason, and the process commonly involves having your specific recipe and process reviewed by someone qualified before you can sell it at all. That review is not a formality and it is not something to do after you've made two hundred jars.

Labelling is its own layer. Expect to need the product name, an ingredient list in descending order by weight, allergen information, net weight, and your business name and address. The details differ, so get them from the source.

Where to find the source: your state's department of agriculture or health department, whichever handles food businesses where you are. Call them and describe exactly what you want to make and where you want to sell it. It's usually a short conversation with someone who has had it a hundred times. Do not substitute a Facebook group, and do not substitute this article. The person on the phone knows; nobody on the internet knows your state.

The same principle applies to any home food business, and the food-safety line between things that need refrigeration and things that don't is worked through from the baking side in how to start selling baked goods.

Subscriptions, and the honest version

Repeat orders are what make small batch work. A customer who buys one bag is a nice afternoon. A customer who buys two bags a month for three years is the business.

But be careful about the word "subscription", because most small sellers mean two different things by it.

The first is genuine recurring billing: a card charged automatically every four weeks without anyone doing anything. That's a specific piece of machinery, and if it's the centre of your model you should check that whatever you're using actually does it before you build around it.

The second is what most small makers actually run, and it works fine: a prepaid bundle sold once and shipped several times. Three bags for $54, shipped one a month for three months. Six bags for $102, one a fortnight. It's a single product, a single payment, and a note in your calendar. The customer has committed, you have the cash up front, and there's no monthly chase.

Here's the honest limitation on Sailo specifically, and it points the same way. On the manual rails, bank transfer and cash on delivery, Sailo never touches the money and takes nothing from it. Which also means nothing can be charged automatically on those rails, by Sailo or anyone else, because the payment happens between the buyer's bank and yours. A "subscription" on a manual rail is a reminder you send and a payment you check, twelve times a year, by hand. Card payments need a Stripe account cleared for charges, and Sailo takes 1–3% of the goods on top of Stripe's own fee.

So if recurring billing is genuinely the core of what you're building, price a tool that does it and budget for it. If it isn't, sell the prepaid bundle and skip the whole problem. Most makers turning out fifteen kilos a week are better served by the second answer, and the method for getting people to come back at all, which is the harder problem, is in how to get repeat buyers.

Worked example: Marcus, Sacramento

Marcus roasts on a 1kg machine in a shared commercial space, four hours every Monday.

His week: orders close Sunday at 11pm. A typical week is 41 bags of 12oz, split across three coffees. That's about 14kg of green, roasted in roughly twelve batches over four hours including cooling and cleaning.

The money on a single bag: green at roughly $8.60 a pound, which after 16% roast loss is about $10.25 a pound of roasted coffee, so about $7.60 of coffee in a 12oz bag. Bag and valve $0.45. Label $0.12. He sells at $19.

That's about $10.80 a bag before shipping, packaging and his own time. Across 41 bags, roughly $440 a week gross margin. Take off the roasting space, boxes, postage he doesn't fully recover, and the four hours plus the two hours of packing, and it's a real but modest number.

The two things that changed it. He introduced a three-bag prepaid bundle at $52, shipped monthly, and about a third of his regulars moved onto it. That smoothed his weekly order count from a range of 22 to 58 down to a range of 34 to 48, which meant he stopped over-roasting on quiet weeks and under-delivering on busy ones.

The second was smaller and stranger. He started printing the roast date in large type on the front of the bag instead of small type on the back. His repeat rate went up. He can't prove why, but his theory is that customers noticed it, mentioned it to people, and it became the thing they said about him.

He also stopped shipping on Thursdays after a summer weekend where four orders sat in a facility until Monday. Photographing the bags well enough that people order in the first place is its own skill, and the one-window method for food is in what to photograph when you sell food.

What to do next

Pick your production day and your cut-off, and write them into one sentence on your shop page today.

Then weigh a batch in and weigh it out, and recalculate your price from the second number. If the gap surprises you, that gap has been coming out of your wages.

Then print the make date on the next batch, in large type, on the front.