Moat

Moat: not proven — a multi-brand operator competing on twelve factors at once

The cleanest summary of Yatsen's competitive position is written by Yatsen itself. The Competition section of the FY2025 20-F lists the basis on which the company believes it competes: "perceived value, including pricing and innovation, product efficacy, service to the customer, promotional activities, advertising, special events, new product introductions, e-commerce initiatives, direct sales, KOL collaborations, and other activities" [1]. That is twelve factors. When a moat list is twelve factors long, none of them is a moat — the company is telling you it competes on execution across a wide front, not on a structural advantage that protects price, share, or returns from rivals. The same paragraph adds that Yatsen "compete[s] with both established multinational and domestic brands, as well as small targeted niche brands that continue to enter the Chinese and global beauty markets" — a low-barrier arena where the company's risk factor framing concedes that many competitors "have greater financial, technical or marketing resources, longer operating histories, greater brand recognition or larger customer bases" than Yatsen does [2].

The verdict, on the primary record across six fiscal years: Moat not proven. A wide moat is not on the table. A narrow moat is possible in the owned premium-skincare equity (Galénic, DR.WU, Eve Lom), but the financial fingerprints a moat should leave — pricing power, segment-level operating profit, durable revenue compounding, brand resilience under stress — are either absent or contradicted in the record. Two facts make the "not proven" read the responsible one rather than "no moat": gross margin has expanded every year through a 36% revenue drawdown, and the skincare mix shift has demonstrably moved that gross margin higher. Both are consistent with the building of a narrow moat from a credible brand portfolio. Neither is yet evidence of a built one.

Moat rating

Moat not proven

Evidence strength (0-100)

Durability (0-100)

Weakest link

Platform traffic costs

One-sentence verdict. Yatsen has a credible portfolio of pre-existing skincare brand equity it did not build and does not yet earn a moat premium on — its gross margin is real, its mix shift is real, but both core segments still lost money at the operating line in FY2025 [3], one of the prestige brands (Eve Lom) has been goodwill-impaired twice for ¥757.1 million of cumulative write-downs [4], and Q1 FY2026 demonstrated that a single platform's traffic costs can erase a year of margin progress in a single quarter [5].

How a moat would show up — and what is missing

A moat earns returns above the cost of capital that competitors cannot compete away. In a brand-driven consumer business, that shows up as some combination of (1) sustained pricing power, (2) sticky customer behavior — repeat-purchase, retention, low elasticity, (3) operating margins that converge upward over a multi-year window because variable customer-acquisition cost falls as the brand compounds, (4) segment-level operating profit that is positive and rising, and (5) margin and share that survive a stress event — a recession, a price war, a channel shift, a regulatory change.

No Results

Five tests, zero clean passes. The honest read is that the company sits in a structurally low-moat arena — "vigorous competition from both domestic and international players" and the explicit listing of multinationals with "greater financial, technical or marketing resources, longer operating histories, greater brand recognition or larger customer bases" [2] — and earns no measurable structural advantage over those rivals at the operating line today. That can change. It has not yet.

Naming the candidate sources of advantage

A moat verdict has to walk through specific mechanisms rather than label-shop. The candidate moats for a digital-first multi-brand China beauty house are limited, and each can be tested against the primary record.

Intangible assets: pre-existing premium brand equity (Galénic, DR.WU, Eve Lom)

This is the strongest candidate moat Yatsen has, and the only one that comes close to a real economic mechanism. None of the three was built by Yatsen — Galénic dates to 1978 under Pierre Fabre, DR.WU to 2003 under dermatologist Ying-Chin Wu, Eve Lom to 1985 in the UK [6] [7] [8] — and Yatsen acquired them between October 2020 and March 2021 with the explicit strategic intent of climbing the value chain into prestige and clinical [9]. The mechanism: skincare is a daily-use, regimen-driven category where brand trust correlates with efficacy claims and clinical credibility, repeat-purchase elasticity is lower than in color cosmetics, and gross margin is structurally higher. Management says it directly: skincare "typically feature[s] higher gross margin and greater customer loyalty" than color cosmetics [10], and the CFO has framed the entire margin path through this lever.

The evidence the candidate is real:

Skincare grew 63.5% YoY in FY2025 to RMB2.28 billion (US$325.7 million) and crossed 53% of revenue for the first time, against an NBS national beauty-retail growth rate of 5.1% [11] — Galénic and DR.WU are clearly taking share in their tiers. Q1 FY2026 extended that gap: premium and clinical skincare (Galénic + DR.WU + Eve Lom combined) grew 61.4% [12].
Galénic was launched in China only in 2021, and within four years its "No. 1 Brightening Radiance Energy Concentrated Care" serum had become "a key player in the premium brightening serum category" [6] — anchor-product economics consistent with a brand that retains buyers across a regimen.
DR.WU was "recognized by Euromonitor in 2023 as Asia's Leading Mandelic Acid Skincare Brand" — a third-party category-leadership citation [7].

The evidence the candidate is not yet a built moat:

Eve Lom — the third leg of the prestige strategy — has been goodwill-impaired in two of the last three fiscal years: RMB354.0 million in FY2023 plus RMB397.8 million against the Eve Lom reporting unit specifically in FY2024 [4]. Pre-existing prestige brand equity, paid for and capitalized, was written down because the operating economics did not materialize after acquisition. That is the cleanest disproof you can ask for of the "owned premium brand equity is a moat" thesis: when the equity belonged to a brand that did not connect with the Chinese consumer, the acquirer ate the difference. The Galénic and DR.WU thesis remains intact for now precisely because the impairment tested only Eve Lom — but the structural lesson is that owning a prestige label does not, on its own, deliver moat economics.
The skincare segment itself lost RMB31 million at the operating line in FY2025, after RMB449 million of loss in FY2024 [3]. A real brand moat earns operating profit at the segment level. Yatsen's does not yet.
Brand resilience cuts both ways. The "hero brand" Perfect Diary, organically built in 2017 and the only organically-built brand of the six, was cut in half between FY2021 (RMB4.87 billion) and FY2023 (RMB1.97 billion) and has flatlined for three years; its color-cosmetics segment grew just 1.9% in FY2025 and reversed to -5.0% in Q1 FY2026 [13]. The organically-built brand was the one that broke under stress. That outcome is consistent with brand equity that lived in a marketing playbook rather than in the customer relationship — i.e., not a moat.

Read of this candidate. Real but narrow, concentrated in Galénic and DR.WU, and not yet earning measurable economic surplus. Not a wide moat. Probably a narrow moat in the making, contingent on the two unimpaired prestige brands continuing to compound without Yatsen needing to re-invest the brand equity through ever-higher traffic spend.

Cost / scale advantage via the Cosmax JV

Yatsen and Cosmax — one of the largest contract cosmetics manufacturers in the world — co-built a 66,462-square-meter manufacturing and R&D hub in Guangzhou that began operating in August 2023 [14] [15]. Yatsen sits as a minority shareholder. The candidate moat: scale economics, formulation insider access, and supply-chain agility unavailable to smaller competitors.

The candidate fails the exclusivity test. The same ODM ecosystem — "Cosmax, Intercos, Shanghai Zhenchen, HCP, Axilone and Qiaxing" — supplies every China beauty incumbent and entrant [15]. The Yatsen-Cosmax JV does plausibly improve cost-of-goods, lead time, and formulation insight at the margin, and it is one reason gross margin has held up through the revenue trough. But Cosmax is not Yatsen-exclusive in any commercial sense disclosed in the record; Proya, Botanee, Florasis and every other domestic challenger has access to the same supply base. A shared supply chain is a competitive enabler in this industry, not a moat.

R&D as moat — credibility-building, not yet protective

Yatsen positions R&D as "a fundamental pillar of our competitive advantage" [16]. The hard inputs are real: R&D expense was 3.3% / 3.2% / 3.2% of net revenue in FY2023-FY2025, with absolute spend rising to RMB137.3 million (US$19.6 million) in FY2025 [17]; the patent portfolio reached 269 items as of December 31, 2025 (10 utility model, 158 design, 72 invention, 29 pending) [17]; R&D infrastructure spans a 1,849-sqm Guangzhou center, a Shanghai global R&D center (May 2024), joint laboratories with Ruijin Hospital and Sun Yat-sen University, a 920-sqm Galénic facility in France, and the Cosmax JV in Guangzhou [17]; and named technology platforms include Biotec™, Smartlock™, and ActiveAnchor™ [18]. On Q1 FY2026 R&D ran at 3.9% of revenue, a step-up from 2.7% a year earlier [19].

No Results

What the candidate is not. R&D intensity is a credibility input for the premium and clinical positioning the company wants to defend, not — yet — a moat output. There is no economic test it passes:

No price premium on the back of R&D output is demonstrated in the filings, and the company concedes it competes against "heavy price competition in the market place" even while trying to "gradually improve our gross margin by introducing higher-margin new products and stricter discounts" [24] — that is mix and discipline, not pricing power.
Most of the patent portfolio (158 design, 10 utility model) is form-factor IP that does not block competitors from competing in the same category; only the 72 invention patents — and some are still in transfer to Yatsen — represent the kind of defensible formulation IP that could support an efficacy claim against a private-label or competitor copy [17].
The 1,849-sqm Guangzhou center and the 920-sqm Galénic facility are modest by global standards. L'Oréal and Estée Lauder operate multi-hundred-thousand-sqm global R&D footprints with thousands of scientists — the candidate has to be that the China-relevant edge from a smaller, more agile platform is real, not that the absolute R&D base is larger.

Read of this candidate. Credibility, not moat. R&D is necessary to play in clinical / dermo-cosmetics and to defend the Galénic and DR.WU positions over time, but it is not in itself protective and it has not produced a single named technology platform that has visibly compressed S&M intensity. The S&M line is what would compress if R&D delivered a brand-strength surplus, and S&M intensity has gone in the wrong direction in the most recent quarter.

Distribution / channel — capability without an exclusive

Yatsen describes a "comprehensive and robust omni-channel ecosystem" across Tmall, JD.com, Douyin, Pinduoduo, Vipshop, RedNote, Bilibili, Weixin Official Accounts / Mini Programs / Video Channels, plus 77 owned offline experience stores, plus Sephora and Sam's Club for the prestige skincare brands and The Colorist for color cosmetics, plus international distribution for Eve Lom [20] [21]. The candidate moat is the operating capability — the software — of running the same brand across five platforms with coordinated price points, promotional calendars, and KOL alignment.

This capability is real and built over seven years. It is also the table-stakes operating model in this arena; the industry tab calls it directly — "in this industry, 'distribution' is software, not real estate. … A brand that can't operate fluently across all five of Tmall, Douyin, RedNote, Weixin, and Pinduoduo will eventually stop growing." Proya, Botanee, Florasis, Maogeping and others all do the same. Operating fluency is necessary to play and not sufficient to win. Worse: the channel is structurally pricing-power-negative — the platforms set the cost of one customer impression, and when they raise it (as Douyin did in Q1 FY2026 [22]), the cost flows straight through to operating margin in the same period.

Switching costs / network effects — quantify or omit

Yatsen's filings discuss "increasing customer retention, and encouraging repeat purchases" as a goal of the broad portfolio strategy, not as a metric the company has achieved [24]. No net revenue retention, cohort retention, average revenue per user, repeat-purchase ratio, or membership-based KPI is disclosed in the annual report or the quarterly releases. A switching-cost moat requires a quantified cost or workflow disruption on leaving — none exists for a Galénic serum customer who can re-buy a competitor's serum the next day on the same Tmall storefront. There is no network-effect moat: the value to one Perfect Diary customer of another Perfect Diary customer existing is zero. These categories of moat are not present and not claimed.

Regulatory / structural — table-stakes

The PRC NMPA registration regime (Regulations on the Supervision and Administration of Cosmetics, effective January 1, 2021) raises the lead-time and cost of new product launches industry-wide, and penalties for non-compliance reach 30 times the value of the affected product [23]. This is a floor under the industry, not a moat for Yatsen — every domestic and multinational competitor faces the same regime and most have larger regulatory teams.

The Eve Lom case study: a paid-for moat that wasn't

The cleanest empirical refutation of the "owned premium-brand-equity is a moat" thesis is Yatsen's own track record on Eve Lom. The brand was acquired in March 2021 from Manzanita Capital, an iconic UK prestige skincare label founded in 1985 with a globally-distributed cleanser balm and recognized brand authority [8]. It was as close to a textbook "buy a moat" deal as a Chinese acquirer could make.

No Results

Management's own language on the impairments — "due to weaker operating results than expected" — is the disclosure-friendly version of: pre-existing brand equity, even at a beloved global label, did not translate into share, pricing, or repeat in the Chinese market under Yatsen's stewardship [25]. For an analyst pricing in the Galénic + DR.WU thesis, Eve Lom is the on-record disproof that pre-existing prestige equity automatically delivers economic surplus. It can; it didn't, for one of three.

The Q1 FY2026 stress test: a single-quarter platform shock

The most informative single page in the moat record is the Q1 FY2026 release. Revenue grew 22.5% to RMB1.02 billion, gross margin reached an all-time high of 80.2%, and skincare grew 58.5% — every "brand pulling weight" KPI lit up [26]. Yet operating expenses jumped to 89.9% of revenue from 83.2%, S&M alone climbed to 72.2% of revenue from 66.4% — and the operating loss margin widened from -4.1% to -9.7% in a single quarter. Management's named cause: "higher traffic acquisition costs on the Douyin platform" [22].

No Results

The single-quarter shape of this is diagnostic of who owns the surplus. If Yatsen had a moat, the 580bps of gross-margin lift since FY2023 should be slowly compressing the S&M ratio (because brand strength compounds and acquisition cost per customer falls). Instead the company gave back 580bps of S&M intensity inside one quarter on a single platform's commercial decision. A wide-moat brand does not have its operating margin set quarterly by a third-party platform's traffic costs. A narrow-moat brand might absorb the shock at the gross-margin line; this one absorbed it at operating income. Operating margin in this business is the residual between two things — gross margin and traffic cost — and Yatsen owns neither price-setting power against rivals nor cost-setting power against platforms.

The Q1 print also tested which brands actually have pull. Skincare grew 58.5%; color cosmetics fell 5.0% [26]. The organically-built brand that carried Yatsen through the IPO contracted; the acquired prestige labels grew. That is consistent with the narrow-moat candidate residing in the acquired skincare set, not in the operating capability Yatsen claims to have built around digital-first DTC.

How Yatsen stacks against operators where the model does earn moat economics

The peer set established by the industry and business tabs gives a useful upper bound. Proya runs a digitally-led China multi-brand model at 73.3% gross margin and 17.6% operating margin [27]. Botanee runs a dermo-cosmetics specialty model at 74.5% gross margin and 10.7% operating margin. Both are GAAP-profitable, both compete in the same regulatory regime against the same multinationals, both use the same ODM/OEM base, both run the same Tmall-Douyin-RedNote channel stack. They are evidence that the China beauty arena can support a narrow moat — and they are the operating yardstick for what an earned skincare-mix shift looks like.

No Results

The Proya comparison matters most. Same arena, broadly similar gross-margin level, 22 points of operating margin gap. That gap is the visible distance between an earned skincare-mix moat and a developing one. The Estée Lauder data point is the opposite caution: a 74.0% gross margin with -5.5% operating margin is a reminder that high gross margin on its own is not a moat — Estée Lauder has a wide moat structurally but is paying for a travel-retail destocking cycle. Both observations cut against any reading that Yatsen's 78.2% gross margin alone proves anything about durability.

What the FY2025 turn actually proved — and didn't

The FY2025 numbers are the most moat-favorable in the company's history and deserve to be read precisely. Revenue grew 26.7% to RMB4.30 billion, gross margin reached 78.2%, and the company posted its first non-GAAP profit since 2021 at RMB8.4 million [28].

What this proves about durability:

The skincare mix lever works mechanically — Galénic and DR.WU can grow above the China beauty industry trend, and growth at that mix lifts gross margin even when no underlying pricing-power dynamic is present.
The company can be operationally disciplined — G&A intensity fell from 14.7% of revenue (FY2023) to 7.1% (FY2025) under the headcount reduction from 3,497 (end-2021) to 1,623 (end-2025) [29]. This is an execution moat in cost, not a structural one in pricing.

What this does not prove about durability:

Both reportable segments still lost money at the operating line in FY2025: Color Cosmetics -RMB59 million, Skincare -RMB31 million [3]. The consolidated non-GAAP profit is the residual of segment gross profit minus unallocated corporate, not the brands earning surplus on their own.
The S&M ratio (66.3% of revenue) barely changed despite the gross-margin lift [30]. A brand moat would have started compressing it.
The FY2025 turn was promptly reversed in Q1 FY2026 — first quarter after the milestone — on a single platform's traffic-cost decision, which is the opposite of moat-protected economics [5].

What would change the verdict

Three signals would move the verdict from "Moat not proven" toward "Narrow moat":

No Results

The first of these — skincare reaching segment-level operating profit — is the highest-information watchpoint. It is the line that the entire bull case rests on, and on FY2025 numbers (RMB2.28B revenue, RMB31M segment operating loss) it is one good year of disciplined growth away. Whether it arrives, and whether it survives the Q1-FY2026-style platform shocks, is the substance of the moat question for the next four to eight quarters.

What an intelligent investor should hold in their head

This report's view, in plain English:

The arena is structurally low-moat. Twelve named competitive factors, vigorous rivalry, multinationals with greater resources, platforms with pricing power over the marketing line, an ODM ecosystem shared with every competitor. No domestic peer has built a wide moat here; the best (Proya) has built a narrow one.
Yatsen has candidates for a narrow moat — owned premium-skincare equity in Galénic and DR.WU. They are growing, they are recognized in their tiers, and they are the right mix to compound over time. They have not yet earned the financial fingerprints of a built moat.
Yatsen has disproofs of moat-like behavior already in the record. Eve Lom's two impairments, Perfect Diary's failure to defend share organically, segment-level operating losses in both segments in the best year of the cycle, and a single-quarter Douyin shock that erased the year's margin progress.
The verdict is "Moat not proven" — narrowly, with the arrow pointing in the right direction. Calling it a narrow moat today over-credits intent for output; calling it "no moat" under-credits the genuine premium-skincare equity Yatsen has assembled. A reader should underwrite this name as a cycle-and-execution story with optionality on a narrow moat developing, not as a moat-protected compounder.

The single most important number in this whole tab is the FY2025 Skincare segment operating loss of RMB31 million on RMB2.28 billion of revenue [3]. That number, on next year's print, is the moat verdict.

References

Generated Jun 19, 2026 10:16 UTC