Wix.com (WIX) Q2 2026 earnings review
Base 1 LLM Sparks Unit Economic Breakthrough, But Bookings Growth Decelerates
Wix delivered a solid 15% YoY revenue growth in Q2 2026, but the real story is a massive inflection in AI unit economics. The launch of the proprietary Base 1 LLM is structurally fixing Base44's margin drag, with management projecting Base44 non-GAAP gross margins to surge to ~60% in 2H from 'near-zero' entering the year. However, beneath the surface, total bookings growth decelerated to 12% (from 15% in Q1) due to continued softness in the Partners channel. Furthermore, heavy investments and a sudden $27M restructuring charge squeezed Q2 Free Cash Flow to just $52.6M (9% margin), a sharp drop from historical 30%+ levels.
🐂 Bull Case
The deployment of the proprietary Base 1 LLM lowers compute and inference costs dramatically. Base44 is expected to reach ~60% non-GAAP gross margin in H2, adding roughly two points of margin improvement to the consolidated business.
The core Creative Subscriptions segment remains highly resilient, accelerating to 15% YoY revenue growth in Q2 ($398.4M) despite broader macro pressures on SMBs.
🐻 Bear Case
Total bookings grew just 12% YoY (down from 15% in Q1). Management explicitly cited 'Partners softness' as the primary drag, suggesting a longer-than-expected recovery for this critical B2B channel.
A sudden $27.1 million restructuring charge contributed to a widened GAAP net loss of $76.4 million. Coupled with elevated R&D and S&M, GAAP operating loss fell to $58.8M.
⚖️ Verdict: ⚪
Neutral. The structural fix to AI compute costs via Base 1 is a massive long-term win that validates management's heavy Q1 investments. However, the deceleration in bookings and ongoing weakness in the Partners channel present near-term execution risks that cap immediate upside.
Key Themes
Proprietary Base 1 LLM Transforms AI Unit Economics
Wix successfully launched Base 1, a proprietary LLM for its Base44 app-creation platform. This gives Wix direct control over AI inference and compute—its largest cost driver. This reverses the severe margin compression narrative from Q1, with management guiding for Base44 non-GAAP gross margins to hit ~60% in H2, a massive leap from near-zero earlier in the year.
Partners Channel Softness Drags Bookings
Total bookings growth decelerated to 12% YoY (from 15% in Q1), lagging revenue growth by 300 basis points. Management explicitly noted that this lag is a direct result of the 'more immediate impact of Partners softness on bookings.' While Partners revenue grew 17% to $213.8M, this represents a slowdown from the 20%+ growth rates seen in 2025, confirming that the product timeline delays and macro pressures cited in Q1 persist.
Reinvesting AI Savings into Aggressive Base44 Marketing
Instead of letting all the Base 1 LLM cost savings flow to the bottom line immediately, Wix is raising its Time to Return on Investment (TROI) threshold for Base44. This means Wix will actively channel these structural cost savings into elevated Sales & Marketing spend in H2 to aggressively capture market share in the AI-native application building space.
Unexpected Restructuring Hits GAAP Margins
Wix recorded a $27.1 million 'Restructuring and other costs' charge in Q2 2026, which did not appear in the prior year. This heavily impacted GAAP profitability, pushing the GAAP operating loss to $58.8 million (vs a $44.3M profit a year ago). While excluded from non-GAAP metrics, this indicates organizational realignment costs are eating into available cash.
Creative Subscriptions Prove Resilient
Despite concerns over AI cannibalization of traditional web building, the core Creative Subscriptions business continues to accelerate. Q2 revenue grew 15% YoY to $398.4 million, boasting an elite 80% non-GAAP gross margin. This steady cash cow is providing the necessary runway to fund the aggressive Base44 expansion.
Other KPIs
Up 15% YoY. This forward-looking metric demonstrates stable underlying subscription strength, confirming that despite bookings deceleration, the existing customer base continues to expand and retain value effectively.
Up 12% YoY. A sub-segment of Business Solutions, primarily driven by Wix Payments. This represents a deceleration from the 19% YoY growth seen in Q1 2026, pointing to continued macroeconomic pressure on SMB gross payment volumes (GPV).
Down slightly from 70% in Q2 2025, but a 100 bps sequential improvement from 66% in Q1 2026. This signals that the worst of the AI compute cost pressure is in the rear-view mirror.
Guidance
Decelerating. This implies a slowdown from the 15% YoY growth delivered in Q2, likely reflecting the flow-through of Q2's softer bookings into recognized revenue.
Accelerating/Reversing. A massive structural improvement from 'near-zero' entering the year, entirely driven by the deployment of the proprietary Base 1 LLM. This is expected to lift total consolidated gross margins by two points in 2H vs 1H.
Stable. Reiterated from prior guidance. This excludes acquisition and restructuring costs. Reaching this target will require significant FCF generation in H2, given the severe drop to 9% unadjusted FCF margin in Q2.
Stable. The company maintained its full-year outlook, suggesting confidence that Base44 scaling and H2 seasonality will offset the observed Q3 deceleration.
Key Questions
Restructuring Charge Details
What specifically drove the $27.1 million restructuring charge this quarter? Is this related to AI-driven headcount reductions in R&D mentioned in prior quarters, or a broader realignment?
Partners Channel Timeline
With bookings decelerating due to 'Partners softness', what are the specific leading indicators management is tracking to call a bottom in this channel, and are the previously delayed product roadmaps now back on track?
TROI Threshold Strategy
You noted an intention to 'moderately' raise the TROI threshold for Base44 to reinvest AI savings into marketing. Can you quantify what the new TROI target is, compared to the 7-9 months cited in Q1?
Base 1 LLM Capabilities
How does the proprietary Base 1 LLM compare to frontier models in terms of reasoning capabilities for complex app development, and what limits exist on internalizing this inference?