ODDITY Tech (ODD) Q2 2026 earnings review
Algorithm Crisis Moderates, But Profitability Remains Suppressed
ODDITY's Q2 results confirm the severe damage inflicted by the Q1 ad-platform dislocation, with net revenue plunging 25% YoY to $181 million. However, the top-line trajectory is reversing: management guided Q3 revenue to decline just 5%, signaling that algorithmic retraining and the shift away from the 'Try-Before-You-Buy' model are finally yielding results. Despite the revenue shock, the company remained EBITDA-positive ($13M) and aggressively capitalized on the crisis, repurchasing $80 million in stock and retiring debt at a discount. The core question now shifts from 'can they fix the algorithm?' to 'how permanently are unit economics and margins impaired?'
🐂 Bull Case
With Q3 revenue guided to only a 5% decline, the catastrophic 25-30% drops of H1 are ending. Tests to bypass signal distortion with their largest ad partner are working.
SpoiledChild remains on track for 35% growth to ~$350M, while METHODIQ is currently beating SpoiledChild's historic year-one trajectory, proving the platform can scale brands independently of IL MAKIAGE.
🐻 Bear Case
Gross margin collapsed to 68.7% from 72.3% a year ago. The spike in customer acquisition costs and deleverage on lower sales are materially degrading unit economics.
The ~50% drop in new customer acquisition in H1 mechanically chokes off the highly profitable repeat order flow required for H2, making the Q3/Q4 recovery fragile.
⚖️ Verdict: ⚪
Neutral. The operational turnaround is progressing faster than feared (Reversing trend into Q3), and management is brilliantly executing counter-cyclical capital allocation. However, structurally lower margins demand caution.
Key Themes
Algorithmic Dislocation Recovery
The primary drag on the business—a catastrophic spike in Customer Acquisition Costs (CPA) driven by an ad partner's algorithm change—is Reversing. Management noted a 'moderating impact' from the dislocation, driving a sequential revenue improvement from a 25% drop in Q2 to an expected 5% drop in Q3. This validates earlier claims that the issue was a 'technical' edge-case involving their Try-Before-You-Buy model rather than brand fatigue.
SpoiledChild and METHODIQ Defying Gravity
Growth in emerging brands is Accelerating despite the core IL MAKIAGE issues. SpoiledChild delivered double-digit Q2 growth and is guided to grow 35% in 2026 to ~$350 million. More impressively, the newly launched telehealth dermatology brand, METHODIQ, is tracking to deliver first-year revenue ahead of SpoiledChild's historic year-one ramp.
Gross Margin Compression
Profitability metrics are Decelerating sharply. Gross margin dropped 360 basis points YoY to 68.7% (from 72.3%). Adjusted EBITDA margin plummeted to 7.1% from 29.0% a year ago. This compression reflects brutal deleverage on fixed costs, higher promotional/CPA intensity to force new user acquisition, and potential mix shifts.
Concentration Risk on Ad Platforms
The stable, underlying risk that derailed 2026 remains entirely intact: ODDITY is overly reliant on a single dominant advertising platform. While they are successfully 'retraining' the algorithm, the company's 100% direct-to-consumer model leaves them permanently exposed to arbitrary external platform changes.
Aggressive Counter-Cyclical Capital Allocation
Management is treating the current operational crisis as a massive financial opportunity. In Q2 alone, they repurchased 5.6 million shares for $80M (retiring ~10% of the float in the first half of the year) and opportunistically bought back $50M of 0% exchangeable notes for just $35M, instantly capturing a $15M gain. Liquidity remains a fortress at $561 million.
Other KPIs
Reversing. After posting a negative $7 million print in Q1, Adj. EBITDA swung back to a positive $13 million, comfortably beating the $8-10 million guidance. While this is down 81% from $70 million a year ago, it proves the business can stabilize its bottom line even while revenue is down 25%.
Net income fell 73% YoY from $49 million, but it perfectly matched Adjusted EBITDA ($13M). This convergence is unusual and implies substantial non-operating gains offset standard depreciation and share-based compensation (likely the $15M gain from retiring exchangeable notes at a steep discount).
Guidance
Reversing. This is the most important number in the release. After back-to-back quarters of 25-26% declines, a 5% drop implies sequential stability and validates that algorithm retraining efforts are restoring the top of the acquisition funnel.
Accelerating sequentially. An improvement from the $13 million delivered in Q2, indicating that as revenue run-rates normalize, operating leverage is beginning to return to the model.
Stable (negative). Implies full-year revenue of roughly $656 million (down from $810 million in 2025). Achieving this requires Q3 and Q4 to vastly outperform the H1 disaster.
Decelerating violently from $163 million in 2025. Given that H1 delivered $6 million in Adj. EBITDA, the guidance implies a massive reliance on H2 profitability ($24-$26 million expected in H2) to bail out the year.
Key Questions
Repeat Revenue Mechanics
With new customer acquisition down roughly 50% in Q1, how much of the Q3 and Q4 revenue recovery is dependent on new user acquisition versus harvesting the older, pre-2026 repeat customer cohorts?
Gross Margin Floor
Gross margin has contracted for four consecutive quarters, dropping to 68.7%. Where is the structural floor for margins as METHODIQ scales and the Try-Before-You-Buy mix normalizes?
Platform Diversification
Given the existential threat posed by the single ad partner dislocation, what tangible steps have been taken in Q2 to diversify customer acquisition across new platforms or retail channels?