The offer was to fund four more shops on the same model, and it was declined without much deliberation. This is normally the point at which a profile becomes sentimental: the artisan who values craft over money, the family business that resists the corporate tide. The owner, when interviewed, found that framing irritating and said so. Her objection was not romantic. It was that the model did not survive duplication, and she could demonstrate why. The shop's margin came from a single fact: she knew what her customers would buy before they did. Ordering was done by intuition built over nineteen years of standing behind the counter, and it produced a wastage rate of under two per cent in a trade where six is respectable. Nothing about this was written down. It could not be, because it was not a procedure but a very large number of small observations, most of which she could not have articulated on request. A second shop would be run by somebody else. That person would order from a system, and a system built from her data would encode the preferences of her customers, not theirs. The wastage would rise, the margin would fall, and the resulting business would be an ordinary shop with an unusual amount of debt. She had done the arithmetic and it was unflattering. At six per cent wastage the model did not merely earn less; it stopped working. The margin that funded her staffing, which was generous and was itself part of why the intuition worked, would have to be cut, and cutting it would remove the continuity that produced the observations in the first place. None of this is an argument against growth in general. She was careful to say so. Plenty of businesses scale because the thing that makes them good is a process, and processes copy. Hers was not a process. It was a person, and persons do not copy, and pretending otherwise is the most common way that a good small business becomes a bad medium-sized one. The interviewer pressed her on succession, which is the obvious weakness. She agreed that it was. Her plan, such as it is, involves a colleague who has worked alongside her for eleven years and who now does roughly half the ordering. Whether nineteen years of observation can be transferred in this way is, she said, an experiment with one subject and no control, and she will not know the answer until she is not there to see it. She has since turned down two further offers. The shop has one branch, eleven staff, and a wastage rate that was 1.8 per cent last year. She regards this as neither a triumph nor a sacrifice, but simply as the size the thing happens to be.