Do you start a product from customer need, or from seeing others make money?
Originally written in Chinese. Translated with AI, reviewed by Gary. Read the Chinese original →
Three weeks ago, in a mall in the Klang Valley, a promoter tried to sell me mooncakes: buy four get two free, even buy four get four. Mid-Autumn was still a while away, and I was a bit surprised: promotions starting this early? Discounting early doesn't necessarily mean they aren't selling, but recent reports on China's mooncake market got me thinking about this business.
It looks like easy money
Mooncakes have always seemed like a very profitable business. You sell in a concentrated month or two each year, prices are high, and there's gifting demand. In 2025, Guangzhou Restaurant sold about RMB1.652 billion of mooncakes alone, more than 30% of its annual revenue, at a gross margin above 53%. That's before other costs, but it's hard not to be tempted by those numbers: so a mooncake can become that big a business.
I've heard this kind of story: the economy is bad, a few friends have a bit of money and don't know what business to do, they get talking and decide to start a mooncake brand. Find a manufacturer, design packaging, film videos, run ads, hoping to make a quick profit over Mid-Autumn. In the end they made nothing and lost money instead. Wanting to get in because you see others making money isn't wrong in itself. The problem is that the market already has so many choices. Why would customers need one more of us?
This optimism can also travel upstream. Brands place orders one after another, the factory feels capacity is short, so it buys machines, expands production lines and hires more people. But a brand ordering from a factory doesn't mean consumers actually bought the goods in the end. This year's factory orders may become inventory sitting with brands. Next year, if brands can't sell and quit, the orders don't come back, but the extra equipment, plant and staff are already there. Consumer demand hasn't been validated, and both sides have already put their money in.
It also reminds me of poon choi sold over Chinese New Year, when this influencer did it and that influencer did it too. Those who only made it after receiving orders at least had a basis for their stock. The ones under real pressure were those who made a big batch first, froze it in cold storage, then slowly looked for customers. Look at one pot's price minus ingredients and it seems very profitable. But if many pots don't sell, the maths is completely different.
Starting from customers, or from someone else's profit
If we start from customer needs, we ask: are mooncakes too sweet? Too many in a box? Is there a gift option at a reasonable price?
If we only see others making money, it's easy to ask first: what's the cost? Which OEM? How should the packaging look? How much to spend on ads? All of that needs doing, but the most important question still hasn't been answered: why should customers buy mine?
A family that only needs two boxes of mooncakes won't suddenly need twenty because the market went from 20 brands to 200. You hire influencers, I add ads; you do buy four get two, I do buy four get four. Everyone works very hard, but we may just be spending more money fighting over the same two boxes.
There's a joke online: on the 15th a mooncake is 16 yuan, on the 16th it's 1 yuan; same taste, same sweetness, so why waste my money? It's exaggerated, but it captures the reality of festive businesses: the cake is the same cake, but the demand for celebrating and gifting has gone. The mooncakes haven't expired yet, but the window for making money is almost closed. At the start you calculate how much each box makes; at the end you only think about how to clear the stock.
We can make a box of mooncakes that looks about the same, but that doesn't mean we've copied someone else's brand, customers, channels and supply chain. Someone else's gross margin doesn't automatically become our profit just because we sell the same product.
Stock against confirmed orders
My own view is: find customers first, take pre-orders, use confirmed orders to decide how much to stock, then top up based on sales. Factories expanding should do the same: besides looking at how much brands order, find out how much actually sold in the end. What you really need to calculate is whether this round still makes money after ads, commission, discounts and leftover stock.
Honestly, when I see others making money selling durians, opening cafés, doing livestreams, even doing AI, I get tempted too. Sometimes I've thought a lot about the product, costs and marketing, only to realise I still haven't worked out why customers would need me.
If you spend far more time studying how much others make than why customers buy, it's easy to keep chasing products, trends and ads, and end up chasing your inventory.
So next time you feel "this looks easy to do", ask yourself first: have you really seen that customers need it, or have you only seen that others seem to be making a lot of money?