How I’m Sourcing Physical Products in the AI Era (Without a Corporate Budget)?
Painless AI Lab – A Hong Kong Practitioner’s Real-World AI Journey (Part 29)
If you’ve ever read a traditional business book about product sourcing, you know the standard advice: define your target audience, research market size, study the competition, position the product carefully, and build a detailed revenue forecast.
That’s approach worked, but it also required a lot of time, money, and professional research. For a big company with resources, it’s fine. For a regular guy in a small Hong Kong apartment trying to build a side hustle? It’s a fast way to burn out before selling a single item. So how do you actually find a product worth selling when you don’t have a team or a big budget?
The Trap of Pure AI Brainstorming:
When AI tools became popular, my first instinct was to treat them like a shortcut. I would ask things like: “Analyze current trends and give me the top 10 products to sell online right now.” The answers were usually generic lists of trendy gadgets that thousands of other people were already looking at. By the time you find a supplier, the trend has often peaked, and you end up in another price war. AI is useful, but it shouldn’t be the starting point. You still need a real human observation first.
My Real Approach: Start with a Daily Frustration
For me, product sourcing doesn’t start with a spreadsheet. It starts with irritation. What do I complain about regularly? What do my friends complain about over dinner? If something is annoying enough to frustrate people in real life, there’s a good chance many others feel the same way.
Take the Ai-Powered Hair Dryer as an example. Mine was loud, heavy, and required constant button adjustments just to get a decent temperature. It was a small daily annoyance, but enough to make me look for something better. Once I have a real problem, that’s when I use AI. I ask it to check whether better options already exist, what the current market looks like, and who the main competitors are. AI is fast at filling in those research gaps.
The Offline Reality Check:
Looking online is not enough. Many of the better factories, especially those working on newer designs, don’t have strong websites or don’t list their latest prototypes publicly. That’s why I still believe in going to trade exhibitions and industrial expos when possible. Walking the aisles, touching the products, and talking directly with manufacturers gives you information you simply can’t get from a screen. Sometimes one conversation at an exhibition is more useful than hours of online research.
How I Actually Find Suppliers?
Once I have a product direction, the next challenge is finding real suppliers. Here are the methods I use:
- Trade Exhibitions
Events like the Hong Kong Electronics Fair and Global Sources are very useful for electronics, tech products, and small home appliances. You can meet manufacturers face-to-face, see real samples, and discuss OEM options directly. Even if you cannot attend in person, you can still browse the official exhibition website to find exhibitors and identify potential suppliers. - Wholesale Markets
Local wholesale markets allow you to buy in small quantities at near-factory prices. This is helpful when you want to test a product without ordering large volumes. - Customs Data
Some platforms provide open exporter and importer records. You can search by product and see which companies are actually shipping certain items. This helps identify real suppliers instead of pure trading companies. - Online Platforms (LinkedIn & Others)
LinkedIn can be useful for reaching factory representatives or sales managers directly, especially when you already know the type of product you want.
Test Fast and Keep the Risk Low
Even with good research, nobody can fully predict whether a product will sell. Luck still plays a role. My current approach is to keep the risk small. Instead of ordering large quantities, I start with samples, use AI tools to create simple promotional videos, and test them on social media. If people show interest, ask questions, or leave comments, that’s a useful signal. If the response is weak, I move on without losing much money. Building something sustainable is less about getting it perfect the first time and more about staying in the process long enough to find what works.
Have you ever tried sourcing a product to sell online?
What was the biggest difficulty you faced?