Running a business isn't about finding one cause
Originally written in Chinese. Translated with AI, reviewed by Gary. Read the Chinese original →
Today I came across a book with an essay on "cause and effect" that I found really interesting.
The author is W. Daniel Hillis, an American computer scientist and co-founder of Applied Minds. He wrote a book called The Pattern on the Stone, which explains, in very simple terms, the logic behind computers, systems and the complex world.
In this essay he makes a bold claim:
The idea of "cause and effect" should be retired.
Of course, he isn't saying causality is completely useless.
In engineering, computers and machines, cause and effect works very well. What you put in determines what the system puts out; you press a button and the machine performs an action. That kind of world is easier to design, control and predict.
The problem is that humans love to squeeze everything into cause and effect.
When business is good, we look for a reason. When business is bad, we look for a reason too. When we feel unwell and see a doctor, we're also looking for the one cause behind the problem. When a company runs into trouble, we naturally ask: which step did we get wrong?
If only I had...
I used to make this mistake a lot in my own post-mortems.
I'd keep thinking: if I hadn't made that decision, would things be different today? If we hadn't spent so much on ads back then, would cash flow be better? If we hadn't tied up so much in inventory, would the company be lighter? If we'd changed direction earlier, would everything be different now?
After running a business for a while, I found this kind of thinking traps you very easily.
Because most things aren't simply A causing B.
Especially companies, markets, cash flow, platforms and teams. None of these move in a straight line.
When sales were bad, I used to wonder if it was the ads. When cash was tight, I'd wonder if it was inventory. When fundraising stalled, I'd wonder if it was the market.
Only later did I realise that often it isn't a single problem. It's a system.
A company is a system
Ads affect sales, sales affect cash flow, cash flow affects inventory, inventory affects product choices, and product choices affect live-stream conversion. When conversion is weak, ads become less efficient; when ads become less efficient, cash gets even tighter.
So you think you're solving one problem, when actually you're facing a system where everything affects everything else.
That's also why you can't build a business just by listening to a guru's one trick. A trick often only fixes one part, but what's really hard about a company usually isn't one part. It's the flow of the whole system going wrong.
When the system enters a negative loop, every problem amplifies another. Sales slow down, so cash gets tight; cash is tight, so you don't dare restock; you don't restock, so the live stream has no good products; no good products, so conversion gets worse; conversion gets worse, so ads get more expensive. In the end it looks like an ad problem, but behind it inventory, cash flow, product, platform and team may all be stuck together.
My biggest lesson of the past few years: running a business can't just be about finding the answer to "where did it go wrong" every day.
Reviews matter, of course, but a review isn't meant to torture yourself or to ask "what if I'd done things differently" every day. A truly valuable review doesn't keep asking where the problem started. It's about seeing the pattern behind it. Seeing which variables are affecting each other, and whether the company is in a positive or a negative loop. Often a problem doesn't appear suddenly; the system amplifies it step by step.
At the same time, bad business isn't necessarily one person's fault; falling sales aren't necessarily just an ad problem; slow fundraising isn't necessarily just a market problem.
In the end, running a business isn't about finding one cause.
It's about adjusting variables inside a complex system, over and over, until the whole system runs again.