A pre-order system has four moving parts: an order window, a batch menu with caps, payment collected at the time of ordering, and a pickup handoff. Get those four right and a home bakery runs itself on a rhythm. Get any one of them wrong and you spend Friday night doing archaeology on a message thread trying to work out who ordered what.
Most home bakers assemble this out of whatever is nearby: DMs for orders, Venmo for payment, a notes app for the list. That works, genuinely, for a while. Then it stops, and it stops in a way that costs you customers rather than just time.
The short version: Post a menu the same day every week with a hard quantity cap on each item. Keep the window open two to four days and close it 48 to 72 hours before bake day. Take payment when the order is placed, not at pickup, because an unpaid reservation is a wish. Confirm automatically and remind the day before. Hand off in one window at one place. The whole thing depends on quantities decrementing in real time, which is the single thing a spreadsheet cannot do while you're away from your phone.
Pick a day and never move it. Menu goes up Monday morning, or Sunday night, or whatever fits your week. The specific day matters far less than it never changing. Customers who know when to look will look. Customers who have to guess will forget, and forgetting looks exactly like churn in your numbers.
Two to four days open. Under two days and people miss it. Over four and orders trail off anyway, because it falls outside how far ahead most people plan food.
Close 48 to 72 hours before bake day. You need that gap to shop, do the math, and start any long ferments. Closing the night before means shopping in a panic on the morning you're supposed to be baking.
Say the deadline in plain words, every time. "Orders close Wednesday at 9pm" beats "orders close soon." People act on specific deadlines and ignore vague ones.
A worked example of the full cycle:
| Day | What happens |
|---|---|
| Monday 8am | Menu posts with quantities |
| Mon–Wed | Orders come in, items sell out as caps fill |
| Wednesday 9pm | Window closes, numbers are final |
| Thursday | Shop to an exact list |
| Friday | Mix, bake, package, label |
| Saturday 9–11am | Pickup window |
The rigidity is the point. All the coordination lives in known blocks instead of running continuously across seven days, which is what makes this survivable alongside a job.
Three to six items. Fewer than instinct suggests. Every added product multiplies your ingredient inventory, packaging types, and timing conflicts on bake day, and a single person cannot batch-produce a dozen products in one session without the day falling apart.
Choose products that share ingredients and oven temperatures. Two doughs and four products beats four doughs and four products by an enormous margin.
Cap every item, below your true capacity. If you can bake 24 loaves, cap at 20. Selling out early is a good outcome: it's a clean demand signal, it trains people to order early, and it leaves you room when something goes wrong. Under-delivering on an order someone already paid for is the one failure that permanently costs you a customer.
Rotate one slot, keep the rest fixed. Regulars want the sourdough to be there every week. A single rotating seasonal or experimental slot gives people a reason to check the menu without destabilizing your production. It's also how you test new products before they earn a permanent place, covered in testing new products in small batches.
Let quantities decrement live. This is the mechanical heart of the whole system. When a customer takes the last cinnamon roll four-pack, the item has to show sold out immediately, to everyone, without you touching anything. A manually-updated count works exactly as long as you're holding your phone, and it fails the first time two people order during dinner. Overselling is the failure customers remember.
This is the part people skip, and it's the part that determines whether the model holds.
An order without payment is a stated intention. Intentions no-show at a rate that will make you abandon pre-ordering for the wrong reason, concluding the model doesn't work when what actually didn't work was letting people reserve for free.
Payment at order time changes several things at once:
No-shows mostly disappear. People collect things they've paid for.
You're never financing inventory. Money arrives days before you buy ingredients, so a slow week can't put you underwater. For a business with no credit line, this matters more than the waste reduction.
Your bake list is exact. Paid orders are facts. Reservations are estimates, and you'll bake a cushion against them, which reintroduces the waste you were trying to eliminate.
Your reconciliation disappears. No chasing, no "did Sarah pay," no cross-referencing a payment app against a spreadsheet on Friday morning.
The objection is that asking for money up front feels pushy. In practice, customers have been paying up front for online orders for twenty years and find it completely unremarkable. What feels unusual is a business that lets you order and pay later. How to create a pre-order system for your food business walks the mechanics.
Confirm immediately, automatically. A customer who orders and hears nothing wonders whether it worked, and then messages you to ask, which is exactly the manual work you were trying to remove. The confirmation should state what they ordered, what they paid, and when and where to collect it.
Remind the day before. This single message is the highest-return thing in the entire system. It converts most would-be no-shows into pickups, because the usual cause of a no-show is not indifference but a Saturday that got away from someone.
One window, one place. Two to three hours, one location. Individually negotiated pickup times will consume your entire weekend, and they're the most common way a functioning pre-order model degrades back into chaos.
Package by name and stage in order. Bags labeled with the customer's name, arranged so you can hand them over without searching. A pickup window should be handing over pre-sorted bags, not assembling orders while somebody waits.
Have a stated late policy and say it in advance. "Orders not collected by 11am can be picked up Sunday morning" or "uncollected orders can't be held" — either is fine, but decide before it happens rather than improvising while someone's texting you at 2pm. See handling a customer no-show on a large prepaid order.
Almost every home baker builds this out of a form, a payment app, and a spreadsheet first. It's the obvious move and it works at low volume. Here's specifically where it stops, so you can see it coming.
Overselling. A form doesn't know how many you have left. Two people order the last item and both get confirmations. You find out Friday.
Split inboxes. Instagram DMs, Facebook Messenger, texts, and email are four separate places, and message requests from people who don't follow you land in a folder you don't check. An order that sits unread for three days is worse than an order you declined.
Payment reconciliation. Payments arrive in a different system than orders, with names that don't always match. Matching them by hand is a weekly chore that grows linearly with your success.
The Friday morning rebuild. You open your spreadsheet and realize you haven't reconciled it against your messages since Tuesday. Now you're reconstructing the week before you can start baking.
No audit trail. A customer says they ordered two loaves. You have one written down. Neither of you is lying, and there's no record to settle it.
The threshold where this reliably breaks is somewhere between ten and twenty orders a week, and it arrives faster than expected once referrals start. None of it reflects badly on you. It's four tools that don't talk to each other being asked to do a job that requires them to.
When the pieces are joined, the same week works differently. You post a link. Customers see the menu with live quantities and order and pay in one step. Items sell out on their own. Confirmations and reminders send themselves. Thursday you open a list of exactly what to make, and it's already correct because it was built by the orders rather than assembled from them.
The work that disappears is coordination, not baking. And coordination is usually the largest block of unpaid time in a home bakery, which is why removing it shows up in your hourly rate rather than just your mood. How to price baked goods covers why that unpaid time is the number most bakers get wrong.
A Homegrown storefront is built for exactly this shape of business: one link for your bio, a batch menu with quantity caps that go sold-out automatically, checkout and payment at order time, confirmations and pickup management, and a listing in a local marketplace where nearby customers can find you. It's $10 a month billed annually, or $12.50 month to month, with no commission on your sales at any volume. Setup runs about fifteen minutes.
Once the orders are final, bake day stops being a planning exercise and becomes a sequence. Working backward from the pickup window is what keeps it from sprawling.
Thursday: the math and the shop. Convert orders into ingredient quantities, not into a shopping vibe. Twenty-two loaves at 500g flour each is 11kg of flour, which is two and a bit bags rather than "some flour." Doing this from the order list rather than from memory is what stops the mid-bake grocery run that costs an hour.
Thursday night: anything with a long lead. Levain builds, soakers, doughs that want a cold overnight ferment. This is the step that makes Friday possible.
Friday morning: mix in production order. Sequence by what needs the oven first and what tolerates waiting. A schedule written on paper beats one held in your head, particularly around hour four when you're tired and there are three timers running.
Friday afternoon: bake in oven-load groups. Group products by temperature so you're not swinging the oven up and down. Every temperature change costs you fifteen minutes of preheat you didn't budget.
Friday evening: package and label to the order list. Bag by customer, not by product. A bag with a name on it is a finished order; a tray of loose loaves is a task you've deferred to Saturday morning when you'll be busier.
Saturday morning: stage and hand off. Bags out in a findable order, ideally alphabetical. If you're searching a table while a customer waits, the window will run long, and running long is what makes people stop coming to the early slots.
Two habits that pay for themselves quickly. Keep a simple log of what you baked, how long each stage took, and anything that went wrong, because after four weeks you'll know your real capacity instead of guessing at it. And keep your kitchen reset between stages rather than at the end, which sounds fussy until you're packaging on a counter still covered in flour. Keeping a home kitchen organized between batches covers the practical side.
Food safety runs alongside all of this and your cottage food exemption doesn't excuse you from it. Cross-contact control matters most in a shared home kitchen, especially where you produce both allergen-containing and allergen-free items. The FDA's allergen labeling guidance covers what has to be disclosed, and the cottage food label checklist covers getting it onto the package correctly.
A system with no orders in it is just a page. The first round comes from people you already know.
Go back to whoever already asked. Somebody has offered to pay you for something. Message them with a date and a price.
Announce it once, plainly, in your existing channel. "Menu posts Mondays, orders close Wednesday, pickup Saturday 9 to 11. Here's the link." No apology for the change and no long explanation.
Post the menu even in weeks when it feels pointless. The habit is the asset. Consistency for four weeks builds an audience that checks; irregular posting trains people to stop.
Ask your first customers for one referral each. After a good handoff, "if you know anyone who'd want a loaf, send them my way" converts better than any paid channel, because the product is doing the selling.
Put the link everywhere. Instagram bio, Facebook page, your email signature, the bottom of your labels. The most common reason someone doesn't order is that they couldn't find where to.
More on the first-customer problem in how to get your first 100 customers and turning Instagram followers into repeat customers.
Post a menu on a fixed day with a quantity cap on each item, keep the order window open two to four days, take payment at the time of ordering, close the window 48 to 72 hours before bake day, and hand off in a single pickup window. The critical mechanical requirement is that quantities decrement in real time so you cannot oversell.
Two to four days. Shorter and people miss it; longer and orders trail off anyway because it falls outside how far ahead most people plan. Closing 48 to 72 hours before bake day gives you time to shop and start any long ferments without a panic.
Up front, at the moment of ordering. Unpaid reservations no-show at a rate high enough to make people give up on pre-ordering entirely, and payment up front also means you're never buying ingredients with your own money before a sale. Customers find paying online completely normal; it's letting them pay later that's unusual.
Three to six. A single person cannot batch-produce more than that in one session without the day collapsing. Products that share doughs, ingredients, and oven temperatures are worth far more than products that don't. Keep most slots fixed for your regulars and rotate one.
That's a good problem and a clear signal. Raise your caps modestly next week, or raise your prices, or both. Selling out consistently within a few hours usually means you're underpriced rather than undersupplied. Don't respond by raising caps beyond what you can comfortably produce, because under-delivering on a paid order costs more than a missed sale.
You can, and most bakers start there. It works up to roughly ten to twenty orders a week, then breaks in specific ways: overselling because DMs don't track quantity, missed orders in message-request folders, and payments that have to be manually matched to orders. See how to sell food on Instagram without taking orders in DMs.
Keep them separate and cap them. Run the batch menu as your backbone, accept a strictly limited number of custom orders per week, price them higher to cover consultation time, and take a deposit at booking. Unlimited custom work is the most common way a promising home bakery becomes an exhausting one. See why custom orders keep food businesses overcommitted.
If pre-ordering is new to you, don't convert everything at once. Take your best seller, the one you know sells out, and offer it by pre-order only for four weeks. Keep everything else as it is. You'll learn the window timing, the cap sizing, and the pickup rhythm on a product where demand is already proven, and the mistakes will be small.
Once that feels automatic, add the second product. The whole model, including how it compares to baking speculatively, is in bake to order vs. bake to sell, and if you're building the bakery from scratch, how to start a home bakery covers everything upstream of the ordering system.
