22 Years, 2.21 Million Followers, 3 Products Left: SOFINA Closes Its China Flagships

Imagine you sell skincare in China.

It's Double 11. Biggest shopping season of the year. Every brand in the country is shouting at once.

And you post a goodbye note.

That's what SOFINA did.

What happened

On November 1, Japanese skincare brand SOFINA, owned by Kao (花王), posted closure notices on both its Taobao/Tmall and Douyin flagship stores.

The stated reason: 因品牌战略升级与业务布局调整, brand strategy upgrade and business layout adjustment. The notice called it a decision made 经过慎重考虑, after careful consideration, and described it as regrettable.

Both stores stop selling on December 1, 2026. Orders paid before 5pm on November 30 still ship, and customer service runs to December 31.

Now look at what was left to close:

And the brand's Xiaohongshu and WeChat accounts? Both stopped updating at the end of December 2025.

So nothing happened overnight here. The lights had been going off one at a time for nearly a year.

When beauty trade outlet Jumeili (聚美丽) asked store customer service where shoppers should buy the products after December 1, it got no clear answer.

Shoppers were blindsided

SOFINA is not a minor label. It landed in China in March 2004 as the first high-end skincare brand Kao brought into the market, and it's been sold here for 22 years. Founded in 1982, the name comes from the Greek sophia, wisdom. Chinese coverage notes it has been grouped with Shiseido and Kanebo as one of Japan's big three.

Plenty of Chinese women know it for primers and sunscreens, and oily-skin shoppers treated its oil-control primer as a staple.

So the response online was straightforward dismay. One recurring tag Chinese netizens attached to the news: 时代的眼泪, tears of an era. Headlines quoted shoppers who had just ordered for Double 11, and one fan who had used it for ten years.

Twenty-two years buys that kind of goodwill, and it does not come back.

"We're not leaving China"

On November 4, Kao answered.

A Kao representative told 21st Century Business Herald the closures are an operations optimisation and do not mean SOFINA is exiting China. Going forward the brand will run mainly through cross-border ecommerce, and the JD overseas self-operated flagship will keep trading normally.

Which checks out. Beauty outlet Qingyan (青眼) reported SOFINA still has an official Vipshop store, and the JD overseas store still carries a few primers and a cushion.

The mechanics matter here. Cross-border means stock sits in bonded warehouses and ships after the order lands. Inventory cost drops. So does the brand's grip on the channel.

So SOFINA is shrinking rather than dying, trading full local operations for a lighter cross-border footprint that holds onto existing fans without paying for the apparatus.

The framing most coverage gets wrong

Here's where it would be easy to write the lazy version of this story: Japanese beauty is collapsing in China, and SOFINA is the latest domino.

The numbers in Jumeili's own reporting say otherwise, and this is the part worth getting right.

In the first half of 2026, in their relevant markets including China, Shiseido grew 10.0%, Kao grew 10.5%, Kosé grew 24.0% and I-ne grew 74.6%. Jumeili's read is that for several Japanese beauty companies the China-related market is a genuine growth opportunity.

And the closure wave is slowing, not speeding up. From 2025 to now there have been ten-plus cases of closures, counter withdrawals and market exits. But compared with 2025's contraction, Jumeili reckons this year has improved, with cases in the single digits including SOFINA.

Different brands are also making different trade-offs rather than all fleeing. TATCHA shut online channels. DECORTÉ (黛珂) and Sekkisei (雪肌精) pulled offline counters but still trade online in China.

The risk is real at the corporate level too. In July 2026, PREMIER ANTIAGING resolved to dissolve and liquidate its Shanghai subsidiary, and the sensitive-skin brand sitrana (施萃娜) no longer has a findable Tmall flagship. Across two years there have been six dissolutions, liquidations or deregistrations among Japanese cosmetics companies in China, including at Shiseido Group, POLA ORBIS and I-ne.

So the honest version is narrower and more useful than "Japanese beauty is finished." Kao grew 10.5%. What failed was one brand running one operating model.

Though the Double 11 scoreboard is still a warning. In Tmall's early Double 11 beauty ranking, the Japanese brands in the top 20 were SK-II, CPB and Shiseido, under 20% of the list. In the personal care top 20, no Japanese brands at all. Proya, Winona and Mao Geping were on the beauty list.

The part every foreign brand should read twice

In October 2021, Kao's China cosmetics division signed an agreement with Lily & Beauty (丽人丽妆), one of China's large ecommerce service partners, to run SOFINA's Tmall flagship, Douyin and Xiaohongshu channels. All three.

Kao's cosmetics chief said at the signing that SOFINA was a brand with real growth potential that he hoped would keep growing in a fast-changing online retail market.

Five years later both of those stores posted closure notices on the same day.

Jumeili's diagnosis is blunt and it is the reason to care about this story: under China's traffic dynamics, the pace of new product iteration and pressure from local competitors, a model that rests on a flagship store plus an outsourced operator can no longer make the growth maths work.

Sit with that for a second, because it's a structural point rather than a Japanese one.

A Tmall flagship is not a vending machine. You don't plug it in and watch it print money. It needs new products, content, livestreams, Xiaohongshu seeding, price discipline and somebody whose job depends on the brand rather than on this quarter's GMV.

SOFINA's social accounts went quiet in December 2025. The store kept trading for eleven more months. A store with no content behind it is just a checkout waiting to be switched off.

One more signal, while you're reading tea leaves. In June 2026 Kao announced a cheaper SOFINA skincare line, BASIC+, at roughly 1,000 to 2,000 yen, around 42 to 84 yuan. The brand's previous core pricing sat at 3,000 to 5,000 yen, roughly 126 to 211 yuan. When a brand halves its price positioning globally and closes its China flagships in the same year, the China closure is not really a China decision.

Closing isn't the failure

Cross-border can be the right call. Lower cost to serve, no local inventory, and loyal fans can still buy. For a brand with 3 live links and a dormant Xiaohongshu account, that beats paying rent on a shop nobody visits.

What went wrong was upstream of the closure. A store with 2.21 million followers was allowed to fade for a year before anyone made a decision, and when the decision came, the brand couldn't tell its own customers where to go next.

What a brand owner should do about it

1. Take your China store's pulse this week.

Three questions, and you can answer them in an hour. When did your Tmall, Douyin and Xiaohongshu accounts last post? How many live product links do you have? What's your repeat purchase rate?

SOFINA's social channels went dark eleven months before the store closed. That gap is where the warning lived, and it was visible to anyone who looked.

2. Count your China new-product launches against your Chinese competitors'.

This is the diagnosis Jumeili lands on, and the one most foreign brands never measure.

Pull your China-market SKU launches over the last twelve months, then pull Proya's or Winona's for the same period. If they are launching several times a quarter and you are launching the global line on a delay, you are not in a slow market, you are in a fast market at the wrong speed.

That number is also the honest test of whether your China business is a business or a distribution arrangement.

3. Audit your operating partner, and never let one partner hold every touchpoint.

Ask your TP for monthly new product launches, content output, livestream hours and repeat purchase rate. If the answer is a GMV chart and nothing else, they are running a shop, not building a brand.

And look hard at the Kao structure: one operator handed Tmall, Douyin and Xiaohongshu together. When a single GMV-compensated partner owns every consumer touchpoint you have, nobody in the arrangement is paid to do the slow work. Split the brand-building mandate from the store-running mandate, even if both sit with the same firm.

4. Choose your model deliberately, with the trade-off written down.

Full local flagship, cross-border store, or distributor. Each suits a different stage and each costs something specific.

Cross-border lowers your inventory risk and weakens your control of the channel, your access to platform traffic and your ability to run local livestreams. Reasonable trade for a brand defending an existing fanbase, bad one for a brand still building awareness. Decide which you are, write down the numbers that would make you switch, and review them quarterly.

5. Write the exit plan while things are fine.

Nobody drafts this and it costs almost nothing. If you ever wind a channel down, where do fans buy, how are they told, and who holds the customer list?

SOFINA's customer service could not answer where to buy next. Twenty-two years of goodwill and 2.21 million followers, and the last impression was a shrug.