Ninety days is not a deadline — it is a measurement window. Long enough to
generate real market data, short enough to keep optionality. Everything in
this checklist serves one purpose: arriving at Day 90 with data you can act on.
Days 1–30 · Set the ground
The first month is administrative, and it decides whether month two runs or
stalls. Do these in order; each one blocks the next.
- Confirm your import and labeling compliance. Category-specific
rules (food, cosmetics, health products) differ sharply; a labeling fix after
goods arrive costs weeks at customs.
- Check your trademark position. Register what you can, and know
where you are exposed, before any public campaign makes the name worth copying.
- Pick one or two channels — not all of them. China is not one
market; it is at least six ecosystems with different economics. Entering all of
them at once means learning none of them.
- Set a price anchor from local evidence. Compare against
comparable brands already selling, not against your home-market multiple.
Days 31–60 · Let content lead, commerce follow
In China, discovery runs ahead of distribution. A brand that arrives with
content already circulating converts measurably better than one that opens a
store and then looks for traffic.
- Start where search-and-social overlap. RED (Xiaohongshu) is the
default first platform for consumer brands entering from abroad — it is where
Chinese consumers check what a foreign brand actually is.
- Begin with many small voices, not one big one. A cluster of
mid-tier creators and KOCs outperforms a single celebrity post for entry-stage
brands, at a fraction of the cost and with far better signal.
- Localize the story, don't translate it. Occasions and gifting
frames travel well; your home-market copy usually does not.
Days 61–90 · Read the data, then decide
The point of a 90-day window is the gate at the end of it. Three reads matter
more than gross sales:
- Organic share of discovery. Are people finding you, or are you
buying every visit? Rising organic search and save-rates are the strongest
early signal.
- Repeat behavior where it applies. First-purchase promos flatter
everyone; the second purchase tells the truth.
- Content-to-commerce linkage. Which content actually moves
product — not impressions, conversions.
Then make one of three calls: expand (add a channel, add cities),
adjust (fix price, assortment or message and re-test), or
exit cleanly. All three are legitimate outcomes of a well-run test.
The only failure is drifting past the gate without deciding.
Three mistakes we see most
- Launching everywhere at once. Six storefronts, six sets of
rules, no depth anywhere. Start narrow, learn fast, expand on evidence.
- Treating China as one market. A Shanghai boutique shopper and a
value-channel shopper in a tier-3 city are different businesses. Choose whose
problem you solve first.
- Shipping global assets untranslated in either sense. The words
may translate; the relevance usually does not.
If a shared, professionally run store fits your category better than building
your own from zero, that is one of the models we run — but the checklist above
applies either way.