PLAYBOOK

1. The one thing brands that keep selling have in common

Selling for a moment in Japan is easy. Selling for years is not. When we put nine brands' sales on the same timeline as their social posts, the variable that tracked sales was not ad spend and not total views. It was the volume of unpaid third-party posts: reviews and videos by people nobody paid. Call it organic UGC.

It works for three reasons. Trust: a stranger's words are believed many times over a brand's. Inventory: a review posted this month brings in new customers next year. Reproduction: posts breed posts. Ads create the first awareness; UGC keeps a brand selling. And UGC is not luck. It follows rules, which the rest of this playbook lays out.

2. The old way: Shiseido invented it in the 1920s and perfected it in 2006

The Japanese cosmetics business model was finished a century ago, by Shiseido, in three inventions. In 1897 it launched Eudermine and borrowed the authority of pharmacology to justify a high price. After the 1923 earthquake it built a chain-store system of roughly 3,000 shops with fixed retail prices, then a training school for them in 1935. In 1934 it selected nine "Miss Shiseido" from more than 240 applicants, trained them for seven months and put them behind counters; in 1937 it started the Hanatsubaki-kai customer club. In today's words: research-backed pricing, price-controlled distribution, ambassadors and community.

The perfected form was TSUBAKI, the shampoo Shiseido launched in March 2006 with more than ¥5 billion a year in advertising and six of Japan's best-known actresses in one campaign. It sold about ¥4 billion in its first month, closed the year at 1.8 times plan, and took Shiseido from fourth to first in shampoo share. Ad budget and shelf count converted straight into sales. That was the era.

3. Why it stopped working: four assumptions that collapsed

The barrier to entry was deregulated away

In April 2001 Japan abolished per-product approval for cosmetics in favor of an ingredient list and full-ingredient labeling: anything within the list can launch without prior approval. In April 2005 the law split "making" from "marketing", creating the Marketing Authorization Holder and making fabless brands a legal category. By March 2024 there were 4,324 licensed cosmetics marketing businesses and 4,357 licensed manufacturers. Japan's haircare shelf alone carries more than 60 makers and 150 brands.

There is no longer one yardstick for reach

Television had GRPs. Social, search, review sites, video, live commerce and AI answers have no shared measure of "spend this, reach that". Budget size stopped being an advantage by itself.

Consumers changed who they believe

Counter staff and beauty magazines were controllable. @cosme, LIPS, X reviews and TikTok "honest reviews" by people with the same skin problem are not. The moment you hire the reviewer, the post becomes an ad and the trust is gone.

Products bypass the shelf

Amazon, Rakuten, Qoo10 and TikTok Shop mean a shelf is no longer a moat. "Own the shelf and win" became "be findable and buyable on every path".

4. The new way: how eight brands sold 1–100 million units

We measured each brand from launch to today across YouTube (long and short), Instagram Reels, TikTok and X, daily, against Amazon sales. The winning moves split by company size.

Large companies: translate the research into words consumers can repeat

THE ANSWER (Kao, November 2024). Entered a shelf of 150 haircare brands with the line "the answer from 100 years of research", and reframed the category as science versus sentiment. Launched first in variety stores (LOFT, Hands) to let high-engagement shoppers build UGC, then widened to drugstores. Thirteen of its top fifteen videos by views were sponsored, all shot as UGC-style clips. Seasonal line extensions and, in 2026, a subscription program kept it in view. One million units in seven months; 2.3 million in about a year.

ELIXIR (Shiseido, 1983). Forty years of collagen research condensed into one word, tsuyadama, the "glow bead" on the cheek anyone can check in a mirror. The word gives the product a way to explain itself in self-service aisles and gives customers a vocabulary for reporting results, which is what UGC is. Scheduled reformulations every few years give buyers a reason to repurchase, media a reason to cover, and retailers a reason to reset the shelf. Its 2017 Wrinkle Cream was Japan's first quasi-drug allowed to claim wrinkle improvement. Number one in skincare sales for nineteen consecutive years.

KATE Lip Monster (Kanebo, May 2021). Launched into a lipstick market that masks had flattened, on the insight that people wanted lipstick precisely because masks made it pointless. The "doesn't transfer" claim invited a test anyone could film: put the mask on, take it off. Named shades ("Heat Haze", "Dark Fig", "Final Boss") turned one-lipstick-per-person into a collection. Chronic stock-outs, including a published apology, became news and restock-tracking became content. 350 million views at a sponsored share of 0–15%. 28 million units to date, on almost no ad spend.

Large companies' greatest asset is R&D, and R&D is not repeatable by consumers. "Super lamellar technology" is not a sentence anyone posts. "The answer", "glow bead" and "doesn't come off" are. The second lesson: do not leave the timing of attention to chance. ELIXIR schedules it with reformulations, THE ANSWER with seasonal drops. And a shrinking market, like lipstick under masks, is the best place to enter: competitors have pulled investment and the unmet need is still there.

Mid-size companies: don't compete on the existing axis, invent one

YOLU (I-ne, September 2021). Into a saturated shampoo shelf, YOLU invented "night beauty": friction, dryness and morning frizz reframed as a care occasion nobody had named. The top twenty TikTok videos (4.2–13.5 million views each) were all sponsored and talked not about the concept but about the feel and the store ("doesn't weigh hair down", "you can get it at Don Quijote"). The concept is what the brand itself keeps talking about, through Sleep Day collaborations and candle-night events; "moisturizing shampoo" would have had nothing to say in months without a new SKU. I-ne had already proven the design-plus-drugstore-plus-social formula with BOTANIST. 10 million units in the first year, 100 million in four and a half, number one drugstore haircare brand by sales.

Perfect One Focus (Shinnihonseiyaku, September 2021). Pores are the top skin concern under thirty, and under-thirties in Japan now open their wallets on the Qoo10 Mega Sale calendar, four times a year. Instead of rejuvenating the parent brand (a 2006 all-in-one gel for over-forties sold on TV shopping), the company launched a separate brand and moved only the invisible assets: formulation and direct-marketing know-how. It made Qoo10 the main channel and concentrated TikTok ads in the days before each Mega Sale, inserting the product into "what to buy in the Mega Sale" lists. POSBUZZ shows the view peaks landing on Mega Sale months and growing each cycle: 21.45 million views in May 2024, 37.36 million in November 2025. Year-one sales above ¥1 billion with 60% e-commerce, then those numbers bought the drugstore shelf. For a mid-size brand, the e-commerce number is the ticket to the shelf.

LuLuLun (July 2011). Sheet masks were a several-hundred-yen "special occasion" treat. LuLuLun rewrote them as the everyday replacement for toner, and backed the rewrite with a bulk box at tens of yen per sheet, a "365" on the pack, and a genuinely easier routine. It redefined its market from "reward-mask share" to "every woman who does skincare daily". Nine reformulations against boredom; blocks of faceless packs stacked on shelves as their own advertising; and "Traveling LuLuLun", regional ingredients in regional packs sold only locally and never online, which turns souvenirs into a distribution network of people handing the product to people. Two billion sheets; number one in face-mask share.

Mid-size brands do not aim for first place on the existing axis. They add one new axis to the market and stand at its far end: "night beauty", "daily mask". The by-product is a reason to talk. A post about a new habit is not a product plug. And because they cannot spend across the board, they concentrate money on the moment of purchase, which in Japan now means the Mega Sale window.

New entrants: spend everything on being believed

Capsule Serum (MD Inc., September 2023). Rode the ingredient-led buying wave without competing on concentration or price. The concept, "a fresh serum that bursts as you use it", resolved an anxiety everyone half-felt: that active ingredients degrade the moment you open the bottle. But POSBUZZ's read of the top 100 posts shows what actually brought people in: 40 posts mention the cleansing oil, 43 mention pores and blackheads, 8 mention "bursting" or "fresh serum". The entry point was a concrete problem, not the hero concept. Formats were "honest review", "one month later", "asked my friend with perfect skin", and the videos landed on "you can get it at Don Quijote" and "found it at Matsukiyo": about 6,000 stores stocked from launch, so what was talked about tonight could be bought tomorrow. 200,000 units in six months; 2.5 million to date.

CHPT.9 (Smart Cosme, October 2025). Pore care is the most contested claim space in Japan and the one where the Yakki law is strictest. It is also the one you can show. CHPT.9 built everything on the before-and-after of pore gunk coming out, which creates certainty without promising a single efficacy claim in words. Dozens of creators showed the same moment from different angles (over 90% of Reels and Shorts sponsored via gifting); "the generic of department-store brands" appeared organically as a price justification; and the landing point was new: TikTok Shop Japan, buy where you watched. Estimated Amazon monthly sales went from ¥480,000 to ¥41.7 million in eleven months, about 85 times, and a second product followed eight months later so that one hit could become "CHPT.9, the pore brand".

New entrants have no track record and no authority, so they stop trying to persuade with words and make people certain with their eyes. Enter through a specific problem, not a worldview; nobody sits through an unknown brand's story, but everyone stops for their own pores. Build exposure and the place to buy at the same time. And draw the second product into the plan from day one: one hit means a product sold, two means a brand exists.

	Large	Mid-size	New entrant
Weapon	Research assets, distribution	Inventing an axis, speed	Visual proof, agility
Message	Translate technology into consumer words	Coin a term that begs explanation	Before-and-after that hits "want" directly
Channel	Staged rollout (specialty first, then mass)	Own a channel; sync to the sale calendar	Launch exposure and point of sale together
Staying power	Scheduled reformulation	Own the context and the channel	Serial launches into a brand

All three design "a reason for third parties to talk" first. Research translation, a coined word, a serum that bursts: every one is a device for being talked about. Pushing with ads alone, without that reason, does not last at any size.

5. Why your view counts are inflated: the four contaminations

Large Japanese companies do this analysis by hand, over months: which creators a competitor used, how spend was allocated, where share jumped, with distribution, ad, influencer and brand-tracking data reviewed weekly. One brand director told us the variable with the highest correlation to sales was not ad spend but the number of organic posts. Everything else is a means to that.

But the total-views number in your agency report is contaminated in four ways, and we found all four in the nine launches:

1. Name collisions. A brand name that is also a common word pulls in unrelated posts; in one case more than half the hits.
2. Official ads counted as buzz. The brand's own commercials added to the "conversation" total.
3. All sponsored. Sponsored videos are often deleted within weeks; of ten we captured at one launch, five were gone in a month. A later analyst sees a brand that "sold organically". It did not.
4. Fake organic. Undisclosed ads. Tell: nineteen near-identical videos posted in the same minute.

Same "hundreds of millions of views", opposite situations. THE ANSWER's launch Reels ran at 80% sponsored: a paid-awareness launch that must now build organic. Lip Monster's 350 million views ran at 0–15%: a brand that should feed the next wave with supply and new shades rather than ads. Look only at totals and you never make that call. It also breaks creator pricing: a 45,000-follower creator delivered 7.15 million views (158× followers); a 738,000-follower creator delivered 7.24 million (9.8×). Follower-based fees ignore the difference. This is why we built POSBUZZ: sales and posts on one timeline, the post behind each sales spike identified, contamination stripped, and the analysis re-run every week rather than once.

6. What is next: red ocean, three channels, and the Mega Sale rhythm

The barrier to entry is gone for good; haircare's 150 brands is where skincare and makeup are heading. The most dangerous position in that market is single-channel dependence: TikTok only, Qoo10 only, one drugstore chain only. One algorithm change, one fee revision, one shelf reset by someone you do not control, and the revenue is gone. We have watched "TikTok traffic halved overnight" and "dropped from the core assortment" happen.

The data also shows something brands miss when they look weekly: posts and sales often do not correlate. Look daily and the big sales spikes fall on distribution events: a new chain, a marketplace change, a pop-up. How big the spike is depends on how much UGC has accumulated beforehand. No UGC, and a sale window passes with the product sitting on the shelf; with UGC, the same sale lifts the baseline a step. Point of sale and conversation are both required.

Channel	Role	Tactic
Online	Where UGC is built	Sync the buzz to marketplace sale windows, as Perfect One Focus did with Qoo10 Mega Sale.
Offline	Where the "sell a lot at once" moment happens	Staged rollout; owned channels like Traveling LuLuLun. If you run pop-ups, OOH or TV, design them to be filmed.
Overseas	The third growth room	Pick the market by population × purchasing power. For a Japanese brand that means Taiwan and the US; for you, it is Japan, and the same logic applies in reverse.

Three channels triple the complexity on the making side: small lots for pop-ups, limited packs, and a different ingredient and labeling regime per country. Which is the point of the next chapter.

7. Who you build with matters more than what you build

Since the 2005 split, "making" can be outsourced; LuLuLun sold two billion sheets without a factory and Capsule Serum was in 6,000 stores in year one. Almost every winning move above rests on that. So choosing a manufacturing partner is not a production decision, it is whether the strategy can be executed. For a fabless brand, the OEM is the business infrastructure: formulation, quality, lead time and regulatory all depend on one company, and yet most brands pick on unit cost. There are 4,357 licensed manufacturers; anyone can make the product. The difference is in what happens after: will they sit in the buyer meeting and vouch for the formulation, help with fixtures and pack design, run the production-to-logistics calendar with you, and follow you overseas with export paperwork and regulatory files? The tell in a first meeting: ask how their best-selling client brand sold. A maker talks about formulation; a partner talks about shelves and promotions.

The paradox of the deregulated era: production can be bought and ads can be bought. Only third-party conversation cannot. So the brand's job narrows to designing the reasons UGC happens, and everything else should sit with a partner who can carry it end to end.

8. What a foreign brand should do first

The Japanese edition of this playbook ends with advice by company stage. For a brand entering Japan from abroad, it condenses to five moves.

1. Measure the category before you plan the launch. Not total views: organic UGC volume, contamination stripped, for your three closest competitors in Japan. If two or three brands already own unpaid conversation on your axis, do not enter on that axis. Invent one, as YOLU did.
2. Enter through a problem, not your worldview. Your brand story is for repeat purchase. Pores, frizz, "doesn't transfer" bring people in. Capsule Serum's hero concept drew 8 of the top 100 posts; cleansing and pores drew 83.
3. Make the claim legal and filmable. The Yakki law allows 56 efficacy statements; most of your US or EU claims are not among them. The workaround is not weaker words, it is a visible result and the words Japanese reviewers already use. Where the law is strictest, video persuades most.
4. Sync to the Mega Sale and build UGC before the window, not during. Qoo10's quarterly Mega Sale is when under-thirties buy skincare. Seed creators and reviews in the weeks before, concentrate any ad spend in the days before, and use the e-commerce number as your ticket to Japanese retail.
5. Set year-one KPIs on the causes, not the sales. Do branded search and branded posts survive when ads stop? How many undisclosed third-party posts appear per month? Until those exist, adding ad budget is pouring water into a bucket with a hole in it. Most of the brands that vanished did exactly that.

A hundred years ago Shiseido organized people who would talk about it, by hand: Miss Shiseido at the counter, the Hanatsubaki-kai in the community. UGC management is not a new social-media theory; it is the industry's original principle, finally measurable. The fifty years of mass advertising were the anomaly. What has changed is only that every conversation now stays on the record and can be counted, and that the cause of a sales spike can be checked every week instead of guessed.