Story

The Full Story

Figures converted from CNY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

From a 2020 IPO sold as a "digitally-native, KOL-powered, Perfect Diary masstige" platform to a 2025 pitch built around skincare, R&D, and "science-backed" premium portfolios, Yatsen has changed almost every load-bearing element of its narrative without ever calling it a failure. The skincare pivot management announced at the start of 2022 has, on the second derivative, worked: gross margin moved from 64% to 78%, skincare reached 53% of mix, and FY2025 closed with non-GAAP profitability. But the story arrived three years late, two Eve Lom goodwill writedowns short, and only after every Gen-Z digital-native KPI from the IPO deck was quietly retired. Near-term revenue guidance has been credible (9 of 10 quarters met or beat); long-dated promises have been only partially honored.

1. The Narrative Arc

Six distinct chapters. Color = where management's emphasis sat each year.

The shape of the arc matters more than any single quarter. Revenue peaked in 2021, collapsed 36% in 2022, then ground sideways for two years before reaccelerating in 2025 — a five-year U with a steady, structural margin expansion underneath. Management's tone moved with revenue; the underlying story moved on its own track from "digital DTC" to "premium science."

2. What Management Emphasized — and Then Stopped Emphasizing

Topic-frequency across transcripts and 20-Fs (0 = not mentioned, 3 = heavy emphasis). Read top to bottom: themes that faded sit on top; themes that emerged sit on the bottom.

The chart reads as a single sentence: everything that justified the IPO valuation has been quietly removed from the pitch, and a new set of premium-skincare/R&D talking points has been bolted in. Two reframings stand out:

1. The "32 million DTC customers" metric vanished after FY2021. Management invoked the Personal Information Protection Law as the reason it could no longer disclose the figure — a regulatory excuse that conveniently arrived as the customer base was no longer compounding. It has not returned in any form (DAU, MAU, repurchase rate, cohort retention) in five years.
2. "Direct-to-KOL" / proprietary MCN, the entire IPO competitive moat, is no longer described as a moat. It is mentioned operationally (a livestream here, a partnership there) but never as the structural advantage that originally justified Yatsen's premium-to-peers multiple. The science-backed/patent narrative has fully replaced it.

3. Risk Evolution

Risk-factor emphasis across the five 20-Fs (FY2020–FY2024). 0 = absent, 3 = explicitly elevated.

Three patterns from the risk language are worth flagging.

The IPO-stage risks (KOL dependence, COVID, ICFR material weakness, HFCAA) all faded — some because the underlying problem genuinely receded (HFCAA after PCAOB inspection, ICFR after remediation), some because management stopped narrating them as central. KOL dependence in particular went from a top-three risk in 2020-21 to a perfunctory mention by 2024.

The acquisition-era risks (goodwill impairment, integration, color cosmetics weakness) showed up exactly when they should have. Goodwill-impairment language strengthened in FY2022 and FY2023 — and in both cases the impairment then materialized ($49.8M in 2023, $55.2M in 2024 against Eve Lom). The risk factor was honest; the conference-call narrative around Eve Lom was not (see Section 4).

Going-concern language has never appeared. Across five 20-Fs, despite cumulative net losses near $1.1B since IPO, no auditor or filing has flagged substantial doubt. This is consistent with the balance sheet, which retained material net cash through the entire crisis.

4. How They Handled Bad News

Three episodes show the pattern: external blame first, financial detail second, narrative continuity third.

FY2022 revenue collapse (-36.5%). The official explanation in the FY2022 20-F: "recurrent COVID-19 outbreaks and related lockdowns," intensifying competition, weak consumer sentiment, and "color cosmetics market faced prolonged headwinds." Notably absent: any discussion that the 32M-DTC-customer cohort was no longer compounding, that Perfect Diary had crossed a brand-fatigue inflection, or that the masstige positioning had been competed away by Douyin-native challengers. The five-year transformation plan — announced in early 2022, before the worst of the collapse — was repeatedly invoked to argue that the strategy was already in motion, recasting a forced retreat as a planned pivot.

Eve Lom goodwill impairments ($49.8M FY2023, $55.2M FY2024 — cumulative ~$105M against the 2021 acquisition price). Both writedowns were disclosed with the same boilerplate language and immediately followed by a brand re-affirmation:

"The carrying value of the Eve Lom reporting unit exceeded its fair value… due to weaker operating results than expected at the time of acquisition. Despite challenges in the market environment and ongoing competition, we still see potential in the brand." — Q4 2023 release

What makes the language revealing is the context. In Q3 2022 management had specifically called out Eve Lom's "robust growth despite the challenging industry environments," and in Q4 2022 cited 99.3% combined revenue growth across Galénic / DR.WU / Eve Lom. The brand was being celebrated on the call while it was simultaneously being impaired on the balance sheet. The second impairment a year later was barely discussed beyond the financial walk-through.

Skincare disappointments in 2023-2024. When skincare itself stuttered — Q3 2023 declined 4.1% YoY, Q1 2024 was flat, Q2 2024 forced a mid-year guidance cut — the explanations again externalized: Abby's Choice phase-out (Q3 2023), seasonal low (Q1 2024), "softer performance of China's beauty market" (Q2 2024). The skincare-led recovery thesis was never re-litigated; the next quarter's results were always framed as the resumption of the original plan.

5. Guidance Track Record

Yatsen guides one quarter forward in revenue (a YoY % range). Eleven guides issued; ten now have outcomes.

Near-term revenue guidance has been credible. The single clean miss (Q2 2024) was acknowledged in real time with a mid-year revision. Beats clustered through the post-collapse recovery (Q1–Q4 2023), exactly when management most needed the credibility — easier to read as conservative guides during a base-low recovery than as systematic outperformance.

The bigger promises track less cleanly:

Credibility: 6/10. Near-term revenue guidance is reliably hit, often by setting a conservative range. The skincare-led pivot has, on the second derivative, delivered: gross margin from 64% to 78%, skincare from 4% of mix to 53%, non-GAAP profitability achieved. But the guidance track record on multi-year promises is mixed — "near to mid-term" profitability took three years, "the decline has bottomed out" was premature by a year, and two Eve Lom goodwill writedowns followed quarters in which management was actively celebrating the brand. There has not been a single earnings call in which an earlier strategic claim was explicitly retracted; setbacks are absorbed into a smooth narrative arc. The pivot worked; the messaging around it has not been audited.

6. What the Story Is Now

The pitch as of Q4 2025: a multi-brand, premium-skincare-led, R&D-driven beauty platform with proprietary patent IP, AI-assisted molecular research, and a global footprint anchored in France (Galénic), Britain (Eve Lom), and Taiwan (DR.WU). Skincare is 61% of Q4 revenue and growing materially faster than color cosmetics. Gross margin sits at 78%. Non-GAAP profitability has been delivered for the full year. A $200M buyback is complete; a $120M convertible private placement (Trustar plus founder Jinfeng Huang) signals internal confidence and provides capital for "research, logistics, international expansion, and acquisition."

"From trend-driven to science-backed, we now offer products across all major segments." — CEO David Huang, Q4 2025

The line is the entire repositioning in one sentence. It is also the line that the rest of this tab should make a reader careful about.

What has been de-risked.

• The skincare-share question. Skincare actually hit and exceeded the 50% threshold management invoked in Q4 2022.
• The gross-margin question. The mix shift to clinical and premium brands has compounded into structurally higher GM (64% → 78%).
• The going-concern question. Despite five years of losses and two material impairments, the balance sheet has held; no auditor language to date.
• The HFCAA / ICFR / China-listing tail risks that dominated the 2021-22 risk factors have largely receded.

What still looks stretched.

• The Eve Lom thesis. ~$105M of cumulative goodwill impaired against the original 2021 purchase price. The brand is still being repositioned ("emotional skincare," Oxford summit, dermatology congresses) — the investor pitch outpaces the financial recovery.
• The "AI / cellular / patents" frame is new. It appeared in Q4 2025 transcripts after being silent for three years. Investors should treat it as a freshly added narrative and look for it to be quantified (specific patent revenue, specific AI-discovered molecules in market) before crediting it.
• Color cosmetics has been deemphasized but is still ~40% of revenue. The structural decline of that segment is being managed, not solved — Perfect Diary's repositioning is a Biolip-driven niche play, not a return to mass relevance.
• The capital architecture. A $120M convertible private placement to the founder and a controlling-shareholder vehicle (Trustar) is unusual and dilutive; it is being framed as a vote of confidence but it is also a discounted insider transaction that bypassed the public market.

What the reader should believe vs discount.

Believe. The mix shift to skincare. The gross-margin expansion. The cost discipline. The near-term guidance. The fact that management can produce non-GAAP profitability when it chooses to.

Discount. The new "science-backed / AI / cellular" framing — until it appears in segment economics rather than press releases. Any forward statement about Eve Lom specifically. Any claim about international expansion until it shows up as a disclosed revenue line. The implicit suggestion that the 2020 IPO-era story merely "evolved" — most of it was retired, and the current portfolio of acquired premium brands is the second strategy, not a refinement of the first.

The current Yatsen is a real company doing real things, in real margins, with real if modest profitability. It is not the Yatsen that came public in 2020 — and the value of reading these filings back to the beginning is precisely the realization that almost none of the original thesis is still load-bearing.