AliExpress's New AI Search Helps Shoppers Find the Right Product
Online shopping searches could get smarter, driving more sales and fewer dead ends.
Have you ever searched for something online, used the wrong word, and gotten nothing useful? That's called vocabulary mismatch. A product may be titled “sporty running top,” but you type “gym tee” and find nothing. AliExpress built a smarter search helper called SAM-D2Q to fix that.
Old search helpers only read product descriptions and invent extra search phrases. But they ignore the product's photo and what shoppers actually buy. SAM-D2Q goes further. It learns from both the product image and text, then creates extra search phrases that match what real customers type. It also learns which phrases lead to actual purchases, not just clicks. Think of it like a clever store clerk who knows you'll like the red dress even though you asked for a “party outfit.”
AliExpress tested SAM-D2Q in its live store. The results were impressive: sales rose 3.38% and the number of paid orders increased 2.27%. That's huge for an online marketplace with millions of shoppers. The system doesn't change prices or products — it simply helps the right products meet the right customers.
The catch? This technology works behind the scenes at AliExpress for now. Smaller shops may not get it soon. And while it's smart, it still depends on good photos and honest product titles. But for shoppers, it means less frustration. For sellers, it means more sales from the same listings. And for AliExpress, it's proof that AI that understands both words and images pays off.
- SAM-D2Q helps product search understand shoppers even when they use different words than the seller.
- It looks at product photos and text together, plus which searches actually lead to buying.
- In live tests at AliExpress, it increased sales by 3.38% and paid orders by 2.27%.
Why It Matters
Smarter online search means less time hunting, better product matches, and more sales for businesses.