🍞 Selling, Growing and Running the Bakery

Most small bakeries do not fail because the bread is bad. They fail because the baker never found out who was buying, what they were buying it for, and whether that particular product actually made money. Both of those are things you can find out this month, with a notebook and a scale.

Start with the notebook. For two weeks, write down every sale: what it was, roughly what time it went, and anything a customer asked for that you did not have. That last column is the valuable one. It is a free list of products your market wants and nobody is selling them.

At the end of two weeks you will be able to name your customer groups in plain words rather than in guesses. A bakery usually has three or four: the daily bread buyer who wants the same loaf at the same time for the lowest sensible price; the passing trade who buys one thing because it is there and smells good; the occasion buyer who wants a cake for a birthday, a funeral or a wedding; and the trade buyer, a shop or a kiosk, who buys quantity and wants a price. Each of these buys something different. The daily bread buyer is buying reliability. The occasion buyer is buying an event and will pay far more for it. Selling both from one counter is normal; pricing both the same way is a mistake.

Now the scale, because knowing who buys is only half of it. You also have to know what each product costs you. The costing method is arithmetic and the arithmetic is the transferable skill. Work in seven steps:

1. Compute your batch ingredient weights from your formula and your target dough weight.
2. Price each ingredient per kilogram using the delivered price, meaning the purchase price plus transport plus any loss on handling, not the shelf price.
3. Multiply each ingredient weight by its price per kilogram and add them up to get the batch ingredient cost.
4. Divide by the number of units the batch produced to get ingredient cost per unit.
5. Add energy per bake, packaging per unit, labour per unit, and a share of fixed overhead such as rent, licences, depreciation and loan repayments.
6. Add a wastage allowance, either as a percentage uplift on cost or by dividing by the fraction of production that is actually saleable.
7. What you have now is cost per unit. Your selling price is that plus the margin the business needs.

Step two catches more people than any other. If flour arrives at your door after a taxi fare and a spilled bag, the delivered price is higher than the price written on the sack, and every loaf you cost from the sack price is under-costed.

Step six is the one that is most often skipped and most often fatal. Fermentation and scaling losses alone can be up to 5 percent and more depending on the work process. On top of that sit rejects, meaning misshapen, burnt or under-scaled product, unsold returns, and damage in storage from pests, damp and spillage. For a bakery selling into a route or a shop, unsold returns are usually the largest and most variable line of all. No published wastage benchmark for small bakeries was available for this course, and that is honest rather than unhelpful, because a guessed allowance is worse than a measured one. Measure your own for one month. It is measurable in a week.

One set of indicative figures may help you see the shape of the arithmetic, and you must treat it as a shape and nothing more. A commercial guide from South Africa, dated June 2026, suggested a bread ingredient cost of around R7 to R10 per loaf, a recommended retail price of R18 to R25, a gross margin on bread of 40 to 55 percent and on custom cakes of 60 to 75 percent. Those are one commercial source's estimates for one country in one month. They are not a benchmark, they are unverified, and the ratios will be wrong for your town. Compute your own.

One last point that decides your product range. Water activity, not total moisture, governs spoilage. Fresh pan bread sits at around 0.93, which is well above the mould threshold and well inside the range where bacteria grow. Batter cakes are around 0.81 and soda crackers around 0.30. Bread is not microbiologically stable; its safety comes from being eaten quickly. Biscuits are progressively safer, which is why biscuits are the natural product for a bakery without cold chain or reliable daily distribution. If your customers are two hours away on a bad road, the honest answer may be that you should be selling them biscuits rather than bread.

There are three ways to sell what you bake, and they are three different businesses. Most bakers drift between them without noticing that the money works differently in each one.

Retail means selling to the person who eats it, over your own counter or at a market. It gives you the highest price per unit and the fastest cash, because the money is in your hand before the customer leaves. It also gives you the highest cost per unit sold, because you carry the shop, the hours, the staff standing about at quiet times and the unsold product at the end of the day. Retail is a good business for high-margin occasion products. That indicative South African guide put gross margin on custom cakes at 60 to 75 percent against 40 to 55 percent on bread, which is one source's unverified figures for one country, but it points at something true: the birthday cake customer is not comparing your price to anything.

Wholesale means selling to a shop, a kiosk, a school or a canteen who then sells it on. You accept a lower price per unit in exchange for volume and predictability. That trade is usually worth making, because a wholesale order lets you bake full batches, and a full oven is where your energy cost per loaf falls. But wholesale brings three problems that retail does not have.

The first is returns. You must agree, in writing, before the first delivery, who owns the loaf that does not sell. Firm sale means the shop buys it and the loss is theirs. Sale or return means you take back what did not go and the loss is yours. These are completely different prices, and a shopkeeper who negotiates a firm-sale price and then quietly returns stock is taking money out of your business. Write it down. Both of you sign it.

The second is that returned bread is a food safety problem, not just a money problem. Rope spoilage is caused by Bacillus spores that arrive in the flour and survive baking, because they withstand temperatures up to about 130 degrees C while the loaf core only reaches about 90 to 97 degrees C. They germinate in the cooling loaf and digest the crumb, first producing a sweetish, fruity or rotten smell detectable at 12 to 24 hours, then a sticky discoloured crumb that draws thread-like strands like spider webbing when the loaf is torn. Published control guidance says plainly to eliminate contaminated leftover bread from the production cycle, because returned bread reseeds the bakery. Returns go out. They do not go back into the mix, and they do not sit on a rack next to tomorrow's cooling loaves.

The third is payment. Retail pays cash. Wholesale pays later, and later is a decision you have to manage rather than accept. That is the subject of Lesson 5.

Contract supply is the third form: a written agreement to deliver an agreed quantity to an agreed specification at agreed times for an agreed period. Schools, hospitals, mines, hotels and government feeding schemes buy this way. It is the most stable income a small bakery can get and it is the hardest to win, because the buyer will ask you for things a counter customer never asks for.

Expect to be asked for traceability. The minimum practical system for a small bakery is to record which flour lot, meaning miller, batch number and delivery date, went into which production date, and which production date went to which customer or outlet. Without that link a recall is impossible, and a serious buyer knows it. Keep it in a hardback book, one line per bake. It costs you two minutes a day.

Expect also to be asked about labelling, and about allergens. Codex requires a date of minimum durability to be declared, in day and month format where the durability is under three months and month and year otherwise, and requires special storage conditions to be declared where the date depends on them. Fresh bakery items are among the products Codex exempts from date marking, but that exemption is a Codex provision and many countries do not grant it, or grant it only for unwrapped product sold on the day of baking. Wrapped, dated bread almost always needs a date mark. Never treat the exemption as a right; ask your national food authority.

On allergens, the Codex declared list is cereals containing gluten, crustacea, eggs, fish, peanuts and soybeans, milk and milk products including lactose, tree nuts, and sulphite at 10 mg per kg or more. Note what is not on it: sesame. Many jurisdictions do mandate sesame, and several add mustard, celery and lupin. Treat the Codex eight as the floor and obtain your own country's list before you print a label.

Finally, price each channel separately. Do the full seven-step costing from Lesson 1 once for retail and once for wholesale, because the labour, packaging, delivery and wastage lines are genuinely different. A single price list applied to both channels always means one of them is being subsidised by the other, and usually it is the profitable one paying for the loss-maker.

Packaging in a bakery does three jobs, in this order: it keeps the product safe, it keeps the product good, and only then does it carry your name. Bakers who do the third job first end up with beautiful bags full of mouldy bread.

Start with the rule that matters most, because it is free and almost everyone breaks it. Do not wrap bread above 95 degrees F, that is 35 degrees C. Wrapping warm bread condenses moisture inside the bag, and that is a mould incubator. Improper wrapping temperature is listed among the direct causes of poor keeping quality. Typical cooling to that point takes 2 to 3 hours. The same 35 degrees C limit applies to slicing, because bread sliced hotter than that caves in.

While it cools, do not stack loaves. Stacking traps moisture and warmth, which is exactly what rope and mould want. Cool in a single layer on clean racks in clean air.

That word clean is not decoration either. Baking destroys vegetative mould, so every mould spore on a finished loaf arrived after the oven. The listed sources of mouldy bread are unsanitary equipment, contaminated packaging and wrappers, mouldy racks and tools, and dust exposure. Read that list again and notice that your packaging is on it. Bags stored open in a dusty corner, or under a leaking roof, or on the floor, are a contamination route straight onto a product you just sterilised in the oven. Keep wrappers covered, clean and off the floor. Clean the slicer daily. Remove and destroy mouldy product immediately and never let it back into the production area.

Now choose the packaging itself, and choose it for the job. Ask three questions.

First, what does this product need protecting from? Bread needs protection from moisture loss and from handling damage, and it needs to breathe until it is cool. Biscuits need protection from moisture gain, because their long life comes from being dry at a water activity of around 0.30, and a bag that lets in humidity throws that away. Cakes need protection from crushing.

Second, how far does it travel and how long does it sit? A loaf sold across a counter within two hours needs almost nothing. The same loaf going out on a route in a hot box needs a wrapper and a box that will not sweat.

Third, what must the label carry? A date of minimum durability, expressed as day and month where the durability is under three months, plus any special storage conditions the date depends on. Codex exempts fresh bakery items from date marking, but many countries do not grant that exemption and wrapped bread almost always needs a date. Allergens must be declared, and the Codex eight are the floor, not the complete list, so obtain your national list. Note also that precautionary wording such as may contain is not a substitute for control. If you can eliminate cross-contact by scheduling your allergen-free products before your allergen-containing ones, and cleaning between, do that instead of labelling around the problem.

Only now think about branding, and keep it cheap and consistent. A brand for a small bakery is three things.

A name people can say and remember. Say it out loud to five people who do not know your bakery and see whether they can repeat it back.

A mark that survives a rubber stamp. The cheapest branding in the world is a hand stamp and an ink pad on a plain paper bag. It works on any bag size, it costs almost nothing per unit, and it never runs out of stock in the wrong colour. Design your mark so it still reads when it is stamped crookedly in one colour, because that is how it will actually appear.

A promise you keep every single day. Fresh at six. Always the same weight. Always the same price. Consistency is the entire brand for a daily bread customer, and it is the one thing a big competitor cannot copy in your street.

There is a legal edge to that promise about weight. If your country regulates the declared weight of a loaf, then the loaf must meet its declared weight after cooling, which makes baking loss a legal calculation and not just a costing one. Baking loss is the weight lost by evaporation during baking and cooling, and no published percentages by product type were available for this course, because the loss is not a set value and fluctuates with several parameters. Close that gap yourself in one afternoon. Weigh the scaled dough, weigh the cooled loaf, and calculate: baking loss percent equals scaled dough weight minus cooled loaf weight, divided by scaled dough weight, multiplied by 100. Then work backwards to the scaling weight you need to hit your declared weight after cooling. Do this before you print a weight on a bag.

Delivery is where a lot of good bread goes bad. The loaf that leaves your bakery correct can arrive squashed, sweating, warm, or already smelling faintly sweet and fruity because rope has started. Distribution is a production step, and it needs planning like one.

Build the route around cooling, not around the driver. Bread must not be wrapped above 35 degrees C, and typical cooling to that point is 2 to 3 hours. That single fact sets your whole timetable. If your customer needs bread at seven in the morning, and the loaf needs two to three hours out of the oven before it can be wrapped, then it must come out of the oven around four, which means it goes into the oven earlier than that, which means the dough was mixed the night before or very early. Work backwards from the customer's hour every time. A route planned forwards from when the driver wakes up always ends with somebody wrapping warm bread to make the van.

Do not stack loaves while they cool, because stacking traps moisture and warmth and favours rope and mould. That applies in the crate as much as on the rack. A crate packed tight with barely-cooled loaves is a stack with a lid on it.

Think about what happens inside a closed vehicle in the sun. Rope is markedly more frequent in hot seasons, and its conditions are a warm loaf with crumb moisture around 35 to 42 percent and crumb water activity around 0.90 to 0.95, which is exactly a normal fresh loaf. The first sign is a sweetish, fruity or rotten smell of the crumb detectable at 12 to 24 hours, followed by a sticky, discoloured crumb that pulls thread-like strands like spider webbing. If your customers start describing bread as smelling sweet or off the day after delivery, you have a rope problem, and no amount of cleaning the van alone will fix it.

The controls for rope are chemical as well as physical. Published control measures are: formulate to a pH of 5.4 or less, or below 4.6 through fermentation; use sourdough or acidifying agents containing acetic acid; acetic acid salts such as sodium diacetate or calcium acetate are protective; an acidity regulator dose of 0.2 to 0.3 percent on total flour in hot months; propionic acid or monocalcium phosphate at 0.1 to 0.5 percent; cool rapidly and do not stack during cooling; disinfect slicing equipment with a 2 percent vinegar solution; and eliminate contaminated leftover bread from the production cycle. Calcium propionate, the standard mould inhibitor, is typically used at 0.1 to 0.3 percent on flour weight, and it works best below pH 5.5, so a fast, lightly fermented dough may leave it underperforming while the baker blames the supplier. Whether you may legally use it, at what maximum, and whether you must declare it on the label, is set by your national food authority. Find that number before you sell.

Now the commercial question that decides your route. What can you actually deliver at this distance? Fresh pan bread has a water activity of about 0.93, which is above the mould threshold and inside the bacterial range. It is not microbiologically stable and its safety comes from being eaten quickly. Batter cakes sit at about 0.81 and soda crackers at about 0.30. Nothing grows below 0.60. So the further and less frequently you deliver, the more your range should shift toward biscuits and dry goods. A bakery without cold chain or reliable daily distribution that insists on sending fresh bread down a four-hour road is fighting biology and will lose. Sending biscuits down the same road is a business.

Be careful about one category in particular. Water activity values for buns, scones, pastry and cream or custard fillings were not retrieved for this course. Filled products are the highest-risk line a small bakery can run, and this course will not guess at their figures. If you want to put custard slices on a delivery route, get proper guidance from your national food authority first.

Three practical rules for the vehicle. Load in delivery order so the first drop is at the door and nobody unpacks the whole van on the roadside. Use crates that stack on each other rather than on the bread, so weight never rests on the product. And carry a written delivery note with the production date on it, signed by the person receiving, because that piece of paper is simultaneously your proof of delivery, your traceability link from production date to outlet, and your evidence in an argument about returns.

Finally, count what comes back. Unsold returns are usually the largest and most variable wastage line for a bakery selling into a route or a shop. Record them per outlet, per day, for a month. Then cut the drop size at the outlets that return the most. That is the fastest cost saving available to a delivering bakery and it costs nothing but the counting.

A bakery can be profitable and still close. Profit is what is left after costs. Cash is what is in the tin on the day the flour merchant wants paying. They are not the same thing, and the gap between them is where small bakeries die.

Here is how the gap opens. You buy flour today, with cash. You bake tomorrow, paying wages and fuel, with cash. You deliver on Wednesday to a shop that pays at the end of the month. The shop sells your bread on Thursday and holds your money for three weeks. During those three weeks you must buy flour again, and again, and pay wages twice more. Every wholesale order you win makes this worse, not better, because growth eats cash before it produces any. That is why bakeries most often fail just after they get busy.

So manage credit deliberately. Six rules.

1. Decide who gets credit at all, and write it down. Retail customers pay cash, always. Trade customers may have terms. New trade customers pay cash for the first month, then get terms once they have proved they pay.

2. Put the terms in writing before the first delivery, on the same page as the price and the returns policy from Lesson 2. State the payment period, the day of the month you invoice, and what happens if payment is late. A verbal arrangement is a gift, not a contract.

3. Take deposits on occasion work. A custom cake is made for one person on one date and cannot be resold if they do not collect it. Occasion products carry the highest margin in a bakery, indicatively 60 to 75 percent against 40 to 55 percent on bread in one commercial guide's unverified figures, but the risk sits entirely with you until the customer pays. Take a deposit that at least covers your ingredient and labour cost, and take the balance before the cake leaves the premises.

4. Invoice on the day you deliver, not at the end of the month. Every day between delivery and invoice is a day added to the wait, and it is a day you gave away for nothing.

5. Keep a debtors page. One line per customer: what they owe, since when. Look at it every week. Debt that is chased in week two is usually paid. Debt that is remembered in month four usually is not.

6. Set a credit ceiling per customer and stop delivering when it is reached. This is the rule bakers find hardest and the one that saves them. A customer who owes you three weeks of bread and wants a fourth week is not a customer, they are a lender, and you are the bank.

Records are what make all of this enforceable. You already need a delivery note with the production date, signed by whoever receives the goods, because it is your traceability link from production date to outlet. That same signed note is your proof in a payment dispute. Keep the copies in date order in a hardback book. When a shopkeeper says they never received Tuesday's delivery, the argument lasts ten seconds if you have their signature and forever if you do not.

Now the cash side of the counter. Handle takings the same way every day: count at open, count at close, write both figures down, and bank or store the difference separately from your float. If more than one person serves, they should each count in and out. This is not about distrust. It is about being able to find a mistake on the day it happens rather than a month later when nobody remembers.

Keep your costing honest as prices move. Your cost per unit was built from delivered ingredient prices, energy per bake, packaging, labour, a share of overhead and a wastage uplift. Every one of those moves. Flour price moves most, and fuel moves fastest. Re-run the seven-step costing at least every quarter, and immediately after any big input price change. A price list that has not been reviewed in a year is a guess about a world that no longer exists.

And count the wastage as a cash item, because it is one. Fermentation and scaling losses alone can be up to 5 percent and more depending on the work process, and on top of that sit rejects, returns and storage damage. Every unit you throw away must be paid for by the units you sell. There is no published wastage benchmark for small bakeries to compare yourself against, so measure your own for a month and treat the number as a cost line with a name, not as bad luck.

One last discipline. Pay yourself a wage, and put it in the costing as labour. A baker who takes money from the till when they need it and calls the rest profit has no idea whether the business works. Put the wage in, and whatever is left after that is the true profit of the bakery.

Growth in a bakery is not a mood. It is a sequence, and the sequence has an order that protects you. Buy capacity you cannot fill and the machine eats you. Take orders you cannot bake and you lose the customer permanently. The plan below runs for twelve months and moves in the order that keeps cash and quality intact.

First, understand the tiers. One commercial South African guide, dated June 2026, put total indicative startup cost at R8,000 to R25,000 for a home bakery, R60,000 to R150,000 for a micro bakery, and R400,000 to R800,000 or more for a retail shopfront. Those are figures from a single source for a single country and they are indicative and unverified. Prices for the same machine differ by an order of magnitude between a new import and a locally fabricated or second-hand unit, and import duty, VAT and freight dominate the delivered cost in most African markets. Treat those numbers as a shape, not a price. What is useful is the shape itself: each tier is roughly an order of magnitude above the last, and skipping one means carrying fixed costs your sales cannot yet feed.

Months one to three: measure. Do not buy anything. Run the two-week customer notebook from Lesson 1, cost your top three products fully with the seven-step method, measure your own wastage for one month, and measure your own baking loss so you know your true scaling weights. You will find money here before you spend any. A bakery wasting product does not need a bigger oven; it needs to stop wasting product, because every unit thrown away has to be paid for by the units sold.

Months two to four: fix the cheapest constraints. The one piece of equipment this course insists on at every tier is a probe thermometer, because dough temperature control, doneness testing and wrapping temperature are all unusable without one, and it is the cheapest quality control instrument in a bakery. An accurate scale is the other non-negotiable, because baker's percentage is unusable without one. Neither is a growth purchase. They are the price of entry.

Months three to six: fix your fuel, because it is often the largest saving available. Published figures for wood-fired baking say a traditional wood oven uses more than 0.5 to 1 kg of wood per kg of baked wheat flour, so a bake using 10 kg of flour consumes 5 to 10 kg of wood. Improved ovens are reported to cut fuel by 50 to 80 percent against traditional ovens. A Ugandan rocket baking oven programme reported preheat time cut by at least two thirds, firewood reduced to one tenth of previous use and an efficiency increase of 70 percent, with capacities from 12 to 24 loaves on small units up to 760 loaves on large ones and indicative capital of USD 1,616 to 7,474. An Ethiopian Mirt stove was reported at USD 11 for a single unit and USD 126 for a clustered four-unit installation, with a 50 percent fuel reduction. Baking in these improved ovens is done at about 250 degrees C.

Here is the arithmetic that turns those figures into a decision. In one documented Ethiopian case, a bakery's annual fuel cost fell from USD 1,900 to USD 970, a saving of USD 930 a year, while income rose by USD 730 a year, giving a payback of about 15 months. Payback is simply the capital cost divided by the annual benefit. Do that sum for every purchase you are considering, using your own quoted price and your own measured fuel use, and buy the shortest payback first. Electricity and gas consumption figures for deck and rotary ovens were not available for this course, but you do not need them from a book: the nameplate on the machine gives the kilowatt rating, and rating multiplied by hours multiplied by your local tariff gives you the cost.

Months six to nine: add the sales channel, not the building. Use Lesson 2 to take on one wholesale or contract customer, priced separately, with returns and payment terms in writing. Full batches lower your energy cost per loaf, so filling the oven you already own is cheaper growth than buying a second one.

Months nine to twelve: make the premises legal and permanent. Licensing costs and authorities differ by country and often by municipality, so no course can give you the figure. What transfers is the checklist of questions. Answer these six in writing, with the name of the office that answered:

1. Who inspects food premises here?
2. What certificate must the premises hold?
3. Do food handlers need a medical certificate?
4. Is a trading licence separate from that certificate?
5. Who certifies gas installations and fire compliance?
6. What must a label carry?

A learner who can answer those six for their own town has done this module correctly. As an illustration of the categories only, one South African source lists a Certificate of Acceptability from the local municipality environmental health department, food safety training per person, a municipal trading licence, a fire compliance certificate and a gas installation certificate from a registered installer. The categories travel. The costs and the offices do not.

Then review and start again. Re-run your costing, re-count your wastage, and check whether the customer notebook says something different from last year. Growth that is measured is growth you can repeat.