Masa Madre – A Factory and a Shop in One

A bakery is a factory and a shop in one. Before you read the four Masa Madre cases: who they’re for, what I learned about running a bakery in the Netherlands, and what surprised me while writing them.

Who is this for and why read this

A bakery is nothing more than a factory and shop in one. Bread, in this case a coiled sourdough pastry called the Volta, goes into an oven (after an involved fermentation process called proofing) and comes out, hopefully, just in time to be sold. If you add customers and competition to the equation, and fuel it with healthy funding and a good team, the venture comes to life. But the best-laid plans are just that, baking bread involves a lot of complexity: from the right environment to proof and bake, to bakers who have to learn the intricacies of the craft, and customers who voice their opinion with their wallets. Add to this a founder pair with vastly different personalities and goals, and you can see why I chose a bakery.

You don't need to be planning a bakery to read this. I was actually inspired by an ice cream shop that opened several locations in my city over a period of 2 to 3 years, and by the five or six bakeries that closed in the 10 years I've lived here. What makes one business scale and another fail? When a business like this closes, people often blame a lack of foot traffic. As you'll read in these cases, there are a lot of factors that decide whether a business succeeds. Luck is the smallest part of it, and planning takes you part of the way, but it's the people inside, who execute and learn every day, who make any venture a success.

The name Volta, which here is the coiled pastry the whole bakery is built on, comes from poetry. Emma Adler explains in Public Books that it "denotes a shift that sonnets exhibit between their opening octave and concluding sestet or between their first 12 lines and final couplet," and that "rather than a shift in perspective, the volta in a novel is a shift to visceral enjoyment." I sincerely hope that you find that enjoyment here too.

The cases

Masa Madre is a bakery in Utrecht built on one product, the Volta, and on two founders who bring different things to it. André Salazar makes the bread, and Wouter Mulder runs the business around it. The four cases follow them from two weeks before opening to a second shop in its first year, and each one stops at a decision that you take before you see what they did. I've grouped them under three themes, because I think anyone running a business that makes what it sells will recognise all three.

Production challenges against business realities

The Rent Runs Anyway opens two weeks before the doors do. The money left covers less than 4 weeks of paying for everything and selling nothing. The product is priced at three supermarket loaves and isn't yet as good as that price needs it to be, because the dough behaves differently in the new building than it did in André's flat. The question is whether to open before the product is ready. Most founders meet the principle underneath it early: a first impression you only get once, weighed against money that runs out on a known date, when nobody can tell you how long the dough will take to come right.

Professionalising operations on knowledge that is hard to share

One Day a Week is set seven months in. A trained baker follows the written method exactly and the bread still comes out wrong, while trade is down on the autumn and Wouter has costed a €28,000 machine. The evidence fits three diagnoses: the baker needs more time, the written method needs numbers, or the knowledge only moves by working next to someone for years. Each comes with a different fix at a different price. What I wanted you to practise is naming the kind of problem before you pay for a fix, because the same evidence supports all three and every fix is a bet on one of them.

Scaling

Two Doors is set two years in. A second site comes up and a family loan is called in, so whatever they open needs a bank loan. Once the bank applies its own sales assumption, every way of opening loses money in its first year, and the question becomes which loss to carry. Two Rooms follows the second shop from the weeks before it opens, and then takes André out of the bakery for three weeks, his first time away since the doors opened, to find out what still works without him. The question there is what the business gives up so it no longer depends on one person, and who pays for that. Across both cases the principle is the same. A plan should be tested against its main assumption failing, not only against the version where everything goes to plan, and growing usually means deciding what you're willing to give up rather than what you add.

The domain

I wrote every case against a set of constraints I researched first and kept in a document I call the Domain Clock, which records what any bakery in the Netherlands can and can't do, and how fast. Where I could, I checked a rule against the statute itself or a lender's own pages, and otherwise against a craft text or the trade body. The research is strongest on labour rules and money, and weakest on location and customers, where most of what you'll read came out of writing the cases rather than out of research beforehand. Some of the production detail came out of the cases too.

Product. Unpackaged sourdough is sold the day it's baked and whatever is left at closing is waste, which makes waste per 100 loaves the number a bakery like this watches every day. When the bread is baked matters as much as the recipe, because a loaf out of the oven at 7 in the morning and eaten at 9 at night is, for the customer, a different product. Part-baked bread finished in a shop oven is standard supermarket practice in the Netherlands, and Lidl has done it in store since 2010. To a Dutch customer it therefore reads as supermarket bread, which limits how a craft bakery can supply a second location.

Production. Fermentation sets the working day. A sourdough's first rise takes about 2.5 hours at a dough temperature of 24°C, and longer when the dough is cooler. Shaped dough can then be held cold for up to about 18 hours at around 6°C. A production space below about 18°C slows the dough and pushes waste up, and a Dutch winter morning starts below that. The difference between two machines mattered more than I expected, because a retarder only holds shaped dough cold, while a retarder-proofer also brings it up to rise on a programme. Neither touches the first rise, which is where most of a baker's judgement sits. Lead times add to this, with a commercial deck oven taking around 11 weeks from order to install, and food business registration and a written hygiene plan both needed before the first sale.

Labour. The bakers' collective agreement sets the full-time week at 38 hours, so one full-time baker covers five production days and a sixth needs a second person or overtime. Hours before 06:00 carry a surcharge, and night-work rules cap an adult night shift at 10 hours, with 12 allowed at most 5 times in 2 weeks. By default the same rules limit someone to 36 night shifts in 16 weeks, which a baker starting at 04:30 five days a week would exceed by more than double (I haven't yet verified how the bakers' agreement handles this). An apprentice takes 3 to 4 years to become a competent solo baker.

Finances. A shop lease in the Netherlands runs five years and renews for five more, with a year's notice, and can only be ended at the end of a term. A tenant who wants out in year four is still committed to year five. Banks lend to a business that has traded for at least 12 months and want last year's figures. For a small BV the annual accounts are normally ready by the end of May, which decides when a loan application can realistically go in. Owners of a small BV are asked to guarantee its loans personally, and a guarantor's spouse or registered partner has to consent in writing.

Location. The lease terms make location the least reversible decision a bakery takes, and beyond that, what I have comes mostly from the cases. People walking past a shop is a different number from people who come back, and only the second one tells you whether the street works. Seasonality, with summer holiday weeks and January trading below the year's average, is in the research as an assumption I haven't sourced.

Customers. The research has least to say here, and most of what follows comes from the cases. The cases sort a bakery's customers by why they come rather than who they are: neighbours who come whatever the bread is like, people who come for the product and leave when it slips, and people who came for something new. Competition arrives in the later cases as a chain with more shops and more money behind it, open longer hours and selling a broader range.

Production and learning notes

What I think will surprise most people reading this is how frustrated I felt at the characters' decisions. People think writing is control, but in fact it's a process where decisions come out on paper that, when you read them back, don't always make complete sense. When Wouter and André first met, I knew their personalities and backgrounds well, but how they would work together, and how those personalities would turn into strengths and weaknesses, only came out during the writing. So I really struggled when André couldn't put his knowledge into words, or refused the retarder-proofer, which I, taking Wouter's side, thought would solve all of these problems.

As mentioned, bakeries are incredibly complex businesses to understand. They also have a natural scaling ceiling, as André explains in one of the cases: the more you do to preserve the product before it reaches the customer, the more it resembles a supermarket product, which follows a different business model. A bakery produces less, so it has to charge more, which works because its product is inherently better than what you find in a fast-moving consumer setting like a supermarket. The main challenge was describing those bottlenecks while still letting the business grow in a healthy way.

An early mistake I made, and tried to correct, was not taking the customer's perspective. I was looking at operational challenges, personality clashes and building the team, rather than at why this bakery deserves to exist. That meant describing people like me, who walk into a bakery and judge it, rarely out loud and more often by not coming back. Part of learning, I feel, is also to listen.

Lastly, why stop at four cases? As you may have noticed, my previous case group, Racine, isn't finished yet. I'm waiting for its founder, and for myself, to make up our minds about the next step. This case group is the antidote to leaving something open-ended, I wanted a clear beginning, middle and end, while leaving open what Masa Madre should do next. Much like in a real venture, the entrepreneur decides in the moment, pays for what happened in the past, and anticipates what's next. That's how a case should end.

Do you have an idea of what you'd want to build next? Would you like a case written about it, anonymised and as different from your own idea as possible? Even then, seeing the principles laid out can be valuable. 

Send me a message

The individuals, companies, and events in this case are fictional composites, created for illustrative and educational purposes. Any resemblance to actual persons, businesses, or events is coincidental.