For years, the standard pitch about China’s baby sector was that it was booming. That pitch is dead. China recorded 7.92 million births in 2025, the lowest number since 1949, after 9.54 million in 2024. The 2024 figure was a Dragon-year bump, not a recovery. So the volume story is over. The value story is not. Chinese families now spend around RMB 48,000 a year on a child aged 0 to 3, and the whole maternity and baby market was worth about RMB 4.63 trillion in 2025, still growing near 7% a year. Fewer babies, far more money per baby. That single sentence should drive your entire China plan.
I am Olivier Verot, founder of GMA. I have been in Shanghai since 2012 and my team has taken diaper, baby food, bath care and juvenile equipment brands into Tmall, JD, Douyin and the mother-and-baby store networks. I have also seen several of them burn a year’s budget on the wrong channel, which is most of what this article is about.
Why the demographics matter less than you think
The one-child policy ended in 2016. The reproduction rate did not explode. Anyone who expected it to misread the country. Marriages happen later, first pregnancies happen later, and the cost of raising a child in a tier-1 city has become the main contraceptive. Add pollution, stress, smoking and falling sperm counts, and you get a fertility problem that no policy fixed. Beijing added a national childcare subsidy in July 2025, RMB 3,600 per child per year for ages 0 to 3. It helps household budgets. It will not reverse the curve.
What did survive the one-child era is the spending pattern. Two parents and four grandparents concentrate their money on one child. Chinese families call it the six-pocket effect and it is very real at the till. A grandmother in Chengdu will argue about the price of her own rice and then pay RMB 3,000 for a stroller. Having a child in China carries family pressure, and that pressure converts into purchases. So your addressable population shrinks every year while your addressable basket grows. A brand selling a RMB 39 wipe pack is in trouble. A brand selling a RMB 249 sensitive-skin cream with a clean lab report is not.
Category by category: where the money actually moved in 2025
The China Toy and Juvenile Products Association published its 2025 retail figures in March 2026. They are the cleanest snapshot of the split between categories that are still growing and categories that are being squeezed. Look at strollers against safety seats: same shopper, opposite direction.
| Category | 2025 domestic retail | Year on year | What it tells you |
|---|---|---|---|
| Toys (excluding collectibles) | RMB 103.5 billion | +5.8% | Educational and play value still sells |
| Collectible and trendy toys | RMB 67.7 billion | +45.4% | Older kids and adults, not babies |
| Child safety seats | RMB 6.0 billion | +10.1% | Safety regulation drives purchase |
| Strollers | RMB 16.4 billion | -3.6% | Fewer babies hit hardware first |
| Baby bottles | RMB 7.25 billion | -1.5% | Volume down, unit price up |
| Baby diapers | Above RMB 65 billion | Flat volume, premium up | Price war at the bottom, growth at the top |
Diapers: a price war at the bottom, an opening at the top
Diapers are the category where foreign brands lost the most ground, and fastest. Chinese domestic brands now take roughly 68% of retail volume and 73% of online volume, a swing of about 38 points in five years. The bottom of the market is a bloodbath: livestream hosts sell RMB 0.6 per piece packs, margins are gone, and no European brand should even try to compete there.
The top is a different market. Diapers priced above RMB 2 per piece grew about 18.5% year on year and now account for around 35% of value. The trigger word is 低敏, low-allergen. Search volume for low-allergen diapers on Chinese platforms roughly doubled in 2025, and more than 60% of parents say preventing nappy rash is their first selection criterion. Two new national diaper standards landed in 2025, adding microbiological testing and tightening limits on formaldehyde and fluorescent agents. If your product already passes stricter EU or Japanese specs, that is not a compliance cost. That is your entire marketing argument, and most brands fail to say it out loud in Chinese.
Baby food beyond milk powder
Everyone obsesses over infant formula. The interesting money has moved to complementary food, 辅食, which covers purees, rice cereals, noodles, and the fast-growing baby snack aisle. Chinese industry estimates put the category around RMB 71.6 billion in 2025, on track to pass RMB 100 billion by 2030, with a compound growth rate near 9%. Foreign brands still hold roughly 42% of value, so the door is not closed the way it is in diapers.
What Chinese mothers read on a baby food label in 2026: sodium first, then added sugar, then the additive list, then the origin of the grain. They photograph the back of the pack and post it on Xiaohongshu for other mothers to judge. One of our food clients redesigned its Chinese label to put sodium per 100g on the front. Conversion on its Tmall product page went up by a quarter with no change in ad spend. Channel note: Douyin’s share of online sales in the baby snack and complementary food category rose from about 28% in 2023 to 42% in 2025 according to Chinese e-commerce trend reports. That shift is faster than almost any other food category.
Juvenile products: strollers, seats, bottles
Strollers fell 3.6% in 2025. This is the category where fewer births hit hardest, because a family buys one stroller and passes it down. But average price keeps climbing. The Chinese stroller market is fragmented, with domestic groups like Goodbaby at the top end alongside imported names, and a long tail of low-cost makers fighting on price. New domestic premium brands built on design and social media have taken share from mid-range European brands in three years.
Child safety seats went the other way, up 10.1%, because regulation and public safety campaigns pushed adoption. Baby bottles slipped 1.5% in value while unit prices rose, which is the same pattern seen across hardware. If you sell bottles, teats or sterilisers, read our detailed breakdown of the baby bottles market in China before setting a price ladder. Anything that touches a child in a car or on a bike falls under CCC certification. Get that done before you build a store, not after.
Baby bath and skincare
Baby wash and care was worth around RMB 32.9 billion in 2024 and an estimated RMB 35.9 billion in 2025 according to Chinese research house Zhiyan. Children aged 0 to 3 account for more than half of it. The move in 2025 and 2026 is away from the single all-purpose bottle toward products split by age, by skin type and by season. A brand that sells one baby lotion for everyone reads as cheap. A brand that sells a 0 to 6 month formula, a 6 to 36 month formula and a winter barrier cream reads as expert, and charges 40% more. Our guide on how to market baby care products in China goes through the positioning work in detail.
Safety is not a feature in China, it is the whole product
Chinese parents grew up with the 2008 melamine milk scandal. That memory did not fade, it hardened into a habit. Before buying, a mother will check the registration number, look for the test report, search the brand name plus 翻车 or 检测 to see if anyone found something wrong, and read the negative reviews first. This is the single biggest gap I see with European founders. They treat certification as paperwork for customs. In China it is the sales argument you put on the first image of your product page.
Practical version. Publish your lab reports in Chinese as content, not as a PDF buried in a footer. Put the certification mark in the product card image. Name the testing body. If you have a European allergen certification or a Japanese skin patch test, translate the actual document. And answer safety questions in the review section within a day, in Chinese, with a named person. Every mother who reads that thread is a prospect.
Domestic brands caught up, and that changed the pitch
Ten years ago, imported meant safe and Chinese meant risky. That argument no longer works. Domestic players like Babycare, Yiying and a dozen others invested in materials, design and testing, and Chinese parents now rate them as good as imports in several categories. They also move faster: a Chinese diaper brand can go from consumer complaint to reformulated SKU in five months. A European brand takes eighteen.
So being foreign is no longer a reason to buy. It is a starting credential you must convert into something specific: a raw material nobody else uses, a manufacturing origin you can prove, a clinical study, a heritage of fifty years in maternity hospitals. Vague European premium positioning gets you compared on price, and on price you lose.
Where Chinese parents actually buy
Online reached 46.1% of maternity and baby sales in 2025, up from 35.1% in 2023, which means offline still holds the majority. That surprises most brands. Four channels matter and they do different jobs.
- Tmall and JD. Still close to half of online baby product sales. This is where the purchase is confirmed after the decision was made elsewhere. Treat your flagship store as a conversion page, not a discovery engine.
- Douyin. Interest e-commerce. Nobody searches for your brand, the algorithm pushes a video of a mother using the product to people whose behaviour signals a baby at home. It works for low-consideration, visual, demonstrable products: bath toys, snacks, wipes, feeding sets. It works badly for anything that needs a long safety explanation, unless you run a store livestream where a nurse answers questions for two hours.
- Mother-and-baby specialty stores. This is the channel foreign brands underestimate. Chains like Kidswant and thousands of independent stores still drive a large share of sales, especially in tier-3 and tier-4 cities, because store staff act as advisors and the mother trusts a face. These stores also run WeChat groups for their customers, so an offline distributor is also a private domain asset.
- WeChat parent groups. The real recommendation engine. Groups formed around a hospital, a due-date month, a neighbourhood or a store. A product mentioned favourably in one 300-person due-date group sells more than a mid-tier influencer post. You cannot buy your way in. You get in through the store owner, the group admin or a paediatric nurse.
The platform mechanics, store types and cross-border options deserve their own read: see the maternity and baby e-commerce market in China. For the influencer side, from top KOL to mother KOC seeding, we covered it in how parenting influencers help you sell to Chinese parents.
Katarina, a Croatian baby care brand that stopped buying traffic
Katarina runs a family baby bath and skincare brand from Zagreb, organic, sold in pharmacies across Croatia, Slovenia and Austria. She opened a Tmall Global cross-border store in 2024. After eighteen months she had done about RMB 2.1 million in revenue. The problem was underneath the number: 62% of orders came from paid traffic, her acquisition cost sat around RMB 340 while her average basket was RMB 220. She was paying to lose money on every second order.
What she had already tried, and what failed. Three Douyin livestreams with a mid-tier host, RMB 480,000 spent, roughly RMB 190,000 in sales, and no repeat purchase afterwards. The host read her ingredient list out loud and the audience did not care, because nobody in that audience knew the brand or trusted it. Livestream converts existing trust into a sale. It does not create trust.
What worked. First, the boring part: full NMPA filing for children’s cosmetics under Chinese rules, plus a Chinese translation of her Austrian dermatological test report, published as a proper article on her Chinese site and pinned on her Xiaohongshu account. Second, KOC seeding instead of KOL buying: 140 real mother accounts, each with 2,000 to 20,000 followers, sent product and asked to post their honest six-week routine, no script. Cost per post was under RMB 400. Third, one distributor covering 90 mother-and-baby stores in Zhejiang, chosen because the owner ran active WeChat customer groups. Fourth, a WeChat service account that sent a skin-care schedule matched to the baby’s age, which pulled reorders without paid media.
Why it worked: in baby care the purchase decision is made in a conversation between mothers, not in an ad. The KOC posts and the store groups put the brand into that conversation, and the test report gave the mothers something safe to repeat. Eleven months later, revenue was 2.4 times the previous year, the paid traffic share had dropped from 62% to 29%, and acquisition cost was around RMB 150. Not a miracle, and not fast. It took roughly five months before the curve moved at all.
What a foreign baby brand should actually do in 2026
1. Build a Chinese website, and make it answer questions
Chinese parents want the best for their child and will spend to get it, but they check first. A .cn site hosted in China or in Hong Kong, with your certifications, your factory, your ingredient explanations and your product comparisons, is the base layer. New in 2026: this site is also what feeds the AI answers. When a mother asks DeepSeek or Doubao whether your cream suits a six-month-old, the model builds its answer from Chinese-language pages it has indexed. If nothing exists in Chinese about your brand, the model either says nothing or repeats a competitor. Writing structured Chinese content that directly answers parent questions, with the question as the heading and the answer in the first two lines, is how you get quoted. That is generative engine optimisation, and in baby care it is already influencing purchases.
2. Forums moved to Xiaohongshu, but the behaviour did not change
The old mother-and-baby forums lost their audience to Xiaohongshu. The behaviour is identical: a mother asks a question, other mothers answer, and the brand mentioned three times wins. Xiaohongshu functions as a search engine for this category. Type a symptom, a baby age or a product problem and you get hundreds of posts. Your job is to be present in those results with real user posts, not with brand advertising, which Chinese mothers spot instantly.
3. Reputation decides everything
Every mother will look up your product before buying, and she reads the bad reviews first. A handful of unanswered complaints about a rash will cost you more than a year of advertising will earn. Monitor Chinese platforms weekly, respond in Chinese, and fix the product issue when the complaint is right. This is dull, ongoing work, and it is the difference between a brand that compounds and one that resets every quarter.
4. Social media: WeChat for retention, Douyin for reach
Baby products have a built-in advantage: a predictable customer lifecycle. You know what a mother needs at three months, at nine months, at two years. A WeChat official account plus a customer service account can push the right product at the right week, automatically. That is private domain marketing and it is the cheapest revenue in the category, because reaching an existing customer costs nothing. Weibo still has reach but is now a PR and crisis channel more than a sales one. Kidswear and seasonal ranges behave differently again, we detailed that in our piece on the China kidswear market.
5. KOC beats KOL in this category
Top influencers are expensive and, for baby products, less effective than a hundred ordinary mothers. Chinese parents trust a mother with 5,000 followers who used the cream for six weeks more than a celebrity paid to hold the tube. Budget accordingly: one big KOL campaign buys around 200 KOC posts, and the KOC posts keep producing search results for years.
6. Baidu still matters, and now so do the AI assistants
Google is blocked. Baidu remains where Chinese parents verify a foreign brand name, and where your distributors and store buyers check whether you are real. Baidu SEO plus a controlled Baike page is basic hygiene. On top of that, treat DeepSeek and Doubao as new discovery surfaces and make sure your Chinese content is structured for them. Both feed on Chinese-language sources. Ignore them and in two years you will be invisible in the place where parents ask their questions.
FAQ
Is it still worth entering China’s baby market with the birth rate falling?
Yes, if you sell premium and can prove safety. Volume categories are shrinking, and if your model needs unit growth you will suffer. But spend per child aged 0 to 3 is around RMB 48,000 a year and Chinese projections put it near RMB 72,000 by 2030. A brand with a specific benefit, a documented formula and a price above the domestic average has more room now than five years ago, because the mass segment has been taken by local players and the premium segment is where the growth is.
Which product categories are still growing?
Complementary baby food and baby snacks, around RMB 71.6 billion in 2025 with growth near 9% a year. Baby bath and skincare, close to RMB 36 billion and moving to age-specific and skin-type-specific products. Child safety seats, up 10.1% in 2025. Premium diapers above RMB 2 a piece, up 18.5%. Strollers and bottles are declining in value, so enter those only with a clear high-end position.
Do I need a Chinese entity, or can I sell cross-border?
Cross-border on Tmall Global or JD Worldwide is the usual first step for baby care, food and accessories, because it avoids full local registration and lets you test demand. The limit shows up quickly: no offline distribution, no mother-and-baby store listings, slower delivery and no domestic advertising for some categories. Most brands that pass roughly RMB 10 million in annual China revenue switch to a general trade setup. Cosmetics for children require NMPA filing regardless of route.
What budget should I plan for a first year?
For a serious entry in baby products, plan on registration and testing, Chinese content and website, store setup, and twelve months of KOC seeding plus a controlled paid budget. Most of our clients in this category land between EUR 80,000 and EUR 200,000 for year one depending on how many SKUs need filing. The money that gets wasted is almost always the paid media spent before the trust content exists. Build the proof first, buy traffic second.
Working with GMA on baby products
GMA is a digital marketing agency based in Shanghai, and baby and maternity is one of the categories we handle most often, from diapers to bath care to feeding equipment. We do the unglamorous part first: Chinese product filing support, safety content in Chinese, then Tmall or Douyin store setup, KOC seeding with real mother accounts, and WeChat private domain for reorders. If you want a frank read on whether your product has a place in China in 2026, talk to us and we will tell you if it does not.
Sources worth reading: China Briefing on China’s maternity and baby market demographics, and the 2025 toy and juvenile products retail figures published by Xinhua.
Read also:
