The best test of a business strategy I know comes from two people who write a cartoon about a town in Colorado. Trey Parker and Matt Stone were asked how they write an episode of South Park, and the answer was a rule they said took them years to work out. Take the beats of your outline and look at the words that sit between them. “If the words ‘and then’ belong between those beats, you’re f****d.” What should sit there instead, they said, “is the word ‘therefore’.”

The rule works because “and then” means the second thing does not depend on the first. You could cut it, swap it, or move it, and nothing else would notice. “Therefore” means the second thing exists because of the first, and “but” means the first thing created a problem the second has to answer. Either way the beats are holding each other up. A story built that way cannot be shortened without losing something, which is another way of saying every part is pulling weight.

Most business strategies are written in “and then”. We will build the product, and then hire a sales team, and then expand into Europe, and then add a second product line. Each item is sensible on its own and none of them needs the others. That is a plan, or more accurately a list, and a list is easy to copy because a competitor can pick any three items from it and do the same.

Compare that with how Jeff Bezos picked what Amazon would sell first. He wanted to sell online, therefore he needed a category where the internet had an advantage a shop could not match, which meant a catalogue too big for any one building. But a catalogue is only useful if you can ship what people order. Books had around three million titles in print, far more than the largest bookstore in the world could stock, and they are small, roughly the same shape, hard to damage, and cheap to post. Books were therefore the category. Each choice follows from the last. You could not swap books for furniture or groceries without the whole chain falling apart, and that is exactly the point. The strategy was internally symbiotic: every piece made the other pieces work better.

Roger Martin’s definition of strategy is the formal version of the same idea. “A strategy is an integrated set of choices that positions you on a playing field of your choice in a way that you win.” The word doing the work is integrated. A set of choices that do not depend on each other is a list of initiatives, and an integrated set is one where “therefore” and “but” sit between every pair.

There is a catch, and it is the part I find hardest. The Amazon chain reads cleanly because it is being read backwards. At the time, nobody wrote “therefore” between the beats in advance, and I do not think you can. The world moves too quickly to pre-ordain a strategy, and a chain written on day one is a guess dressed up as a plan. What actually happens is closer to discovering the strategy as you launch the plane, except that analogy has you already in the air with no way to change the wings.

A better picture is a bridge, built from both sides at once. On the left bank you build infrastructure: the things you can do, which means systems, data, people and capital. On the right bank you build insights: the things you have learned about the market that other people have not, or have not acted on. Neither side knows exactly where the span will meet when it starts. You build a bit on the left because the right has shown you something, and a bit on the right because the left now lets you see further. The place where the two halves meet is your competitive advantage, and you did not pick it, you found it.

The two ideas fit together like this. Parker and Stone’s rule is not about writing the chain in advance, it is about how the beats relate once they exist. Within each half of the bridge every block should still be a “therefore”: we learned X, therefore we built Y, therefore we could see Z. What you give up is knowing the middle. What you pre-ordain is only which two banks you are building from, and the discipline that nothing goes on the bridge that is not held up by the block before it. Bezos did not know in 1994 that the catalogue would turn into a marketplace or that the shipping problem would turn into a logistics business, but each of those is a “therefore” from the block before it, which is why they stuck and why the rest of retail could not haphazardly bolt them on afterwards.

The test is simple and slightly brutal, which is what makes it useful. Write your strategy as a sequence of beats, then read the words between them. If a competitor could lift any one beat out and it would still make sense on its own, that beat is an “and then” and you have not decided anything yet. If removing one beat breaks the ones either side of it, you have a strategy, and someone copying it would have to copy all of it at once. And if the beats only exist on one bank, you have a capability or an opinion, not a bridge.

P.S. the same test works on a pitch deck. Most of them are “and then” from the problem slide to the ask.