Douyin interest e-commerce (xingqu dianshang): selling to people who weren't searching for you
Short version: Douyin interest e-commerce (xingqu dianshang) is "product looking for people," not the other way round. On Taobao a shopper types what they already want and the platform matches it. On Douyin the algorithm decides someone might want your product before they've thought of it, and serves the content that creates the itch — then closes the sale in the same feed. That's the whole model. Foreign brands keep treating it like a marketplace with a shelf, and then wonder why their beautifully optimized listing sells nothing.
Douyin coined the term "interest e-commerce" back in April 2021, then broadened it to "omnichannel interest e-commerce" in 2022 once it added a searchable, shelf-based side. The label sounds like marketing jargon, but the mechanic underneath is the single most important thing to understand about selling on the platform, and it's the thing most import brands never internalize. So let me be concrete about what it actually is, why the numbers have shifted, and where foreign brands throw money away.
"Product looking for people" is not a slogan — it's the engine
Traditional Chinese e-commerce is people looking for product (人找货). A shopper on Taobao or JD has already decided they want running shoes; they search, they compare, they buy. Demand exists before they arrive. Your job is to be findable and to win the comparison.
Douyin flipped that. It's product looking for people (货找人). Nobody opens Douyin to shop. They open it to be entertained, and the recommendation engine — reading watch time, likes, dwell, what they bought last week — decides which of them is likely to want your product right now, and drops the content that lights the fuse into their feed. Interest gets manufactured on the spot, and the "buy" button sits in the same video or livestream. The purchase happens on an impulse the platform created, not one the shopper walked in with.
That one difference rewrites your whole playbook. On a search platform, content is a cost that drives people to your listing. On Douyin, content is the store. If the content doesn't earn watch time, the algorithm never shows it, and no amount of listing optimization saves you — because there was never a shopper searching for you in the first place.
The FACT+S map, minus the buzzwords
Douyin's own framework for this is called FACT+S. It's a genuinely useful map of where GMV comes from, once you strip the acronym down to plain English:
| Lever | What it means | What it's good for |
|---|---|---|
| F — Field (自播) | Your own always-on merchant livestream and account content | The base load — steady, ownable, your best margin |
| A — Alliance (达人) | A matrix of creators and KOCs selling your product for commission | Reach and credibility you don't have to staff |
| C — Campaign (营销活动) | Platform moments and big pushes — 618, Double 11, product launches | Spikes and new-audience acquisition |
| T — Top KOL (头部大 V) | The marquee livestreamers with millions of viewers | Volume and awareness — expensive, rarely profitable alone |
| +S — Search / Shop / Shopping mall | The shelf side: in-app search, your storefront, the mall tab | Capturing the demand the content already created |
The mistake I see constantly: brands blow the entire budget on the T — one big-name livestreamer, one huge night, a screenshot of the GMV for the head-office deck — and build nothing on F. Top-KOL sales are rented demand. The stream ends, the traffic evaporates, and you're back to zero with a thinner wallet. The brands that compound are the ones grinding out their own daily livestream (F) and seeding a creator matrix (A), using campaigns and top KOLs as accelerant, not as the engine.
Why the "+S" shelf side is quietly winning
Here's the shift that matters right now. For years Douyin was pure interest e-commerce — all feed, all impulse. But over the year to mid-2025 its shelf-based e-commerce (search, store, mall) grew about 49% year-on-year, outpacing the platform's overall e-commerce growth of roughly 34%, and shelf now accounts for close to half of e-commerce GMV. The current mix is roughly shelf 40%, merchant-led livestream 30%, influencer-led livestream 30%.
Read that carefully, because it's easy to misread. It does not mean interest e-commerce is dead and you should treat Douyin like Taobao. It means the funnel closed a loop: content creates the interest, and a growing share of people then search your brand on Douyin or return to your store to buy — sometimes days later. The content still does the manufacturing of demand; the shelf just catches more of it now instead of letting it leak to Tmall or Taobao. If your Douyin storefront and in-app search presence are empty, you're creating demand and handing it to a competitor who bothered to be there when the shopper came back.
What foreign brands get wrong
- Porting hero content. Your gorgeous global brand film is watch-time poison here. Native Douyin content is fast, face-forward, spoken in the vernacular, and useful or entertaining in the first two seconds. Beauty that doesn't hold attention gets buried by the algorithm before anyone sees the logo.
- Treating the listing as the work. Optimizing your product page is a Taobao instinct. On Douyin, if the content upstream isn't earning distribution, the perfect page is a shop with no street.
- Confusing a livestream spike with a business. One big KOL night is not traction. Ownable daily self-broadcast is. Rented demand looks like progress and behaves like a sugar high.
- Ignoring the shelf until it's too late. Brands pour into feed content, create real search demand for their name, and never stand up the store or optimize in-app search — so the demand they paid to create converts on someone else's platform.
How I'd actually sequence it
Start with F: a modest, consistent, genuinely native self-broadcast — presence over polish, five days a week beats one big event. Layer A: a small matrix of mid-tier creators and KOCs on commission, so demand generation doesn't live or die on your own account. Only then use C and T — campaigns and the occasional top KOL — as accelerant against content you've already proven converts. And from day one, keep the +S shelf populated: a clean storefront, brand-name search covered, so the interest you manufacture has somewhere to land. The order matters more than the size of the cheque.
Bottom line
Douyin interest e-commerce isn't a marketplace you list on — it's a demand-manufacturing machine that happens to have a checkout attached. "Product looking for people" means your content does the selling and the algorithm does the targeting, so the brands that win invest in ownable content and a creator matrix, not one rented livestream. And with the shelf side now near half of GMV, you close the loop by being there — store and search — when the interest you created comes back to buy. Get the sequence right and the platform builds demand you never had to chase.
If you're trying to work out whether Douyin belongs in your China plan — and what it would actually take to run it well — that's the work I do. Reach out. For the related picture on where that demand converts, read Douyin and Xiaohongshu in-app stores vs Tmall, and on getting found once people do search you, Douyin as a search engine.