Fiverr International Ltd. Q3 FY2026 earnings call

• Strategic Shift to AI Era: Management emphasizes that the traditional web/search era is ending, with organic clicks falling and acquisition costs rising. Fiverr is pivoting to become the most trusted platform for high-value, high-trust work in an 'agentic' world where AI assistants discover talent.
• High-Value Work Growth: High-value projects (>$1,000) now represent 16% of completed projects' gross order amount on a trailing twelve-month basis, up from 15%. Verticals like Programming/Tech and Graphics/Design are seeing mid-teens growth in this segment.
• Product Innovations: Launched 'Mira', an AI headhunter that converts vague briefs into precise matches, resulting in median order values five times larger and 25% lower cancellations for first-time matches. The Knowledge Graph performs 3.3 million skill extractions monthly to improve matching accuracy.
• New Distribution Channels: Fiverr is now integrated into Claude and ChatGPT via public APIs and MCP integrations. AI-sourced Gross Merchandise Value (GMV) for first-time buyers grew ~64% quarter-over-quarter.
• E-commerce Services Momentum: AutoDS drove significant upside in services revenue through a major campaign launched in late August, which performed better than expected and is expected to continue into Q4.
• Operational Discipline: Non-GAAP operating expenses were $62.2 million, down from $66.7 million year-over-year, due to tight management of discretionary spend, internal AI efficiencies, and reductions in R&D and Sales & Marketing.
• Leadership Update: Chief Business Officer Jinjin stepped down due to personal relocation plans; the search for her successor is advanced.

The company operates as a unified business rather than distinct product segments, but financial performance can be broken down by revenue source. Marketplace revenue was $55.4 million, representing approximately 61% of total revenue, and declined 24.7% year-over-year due to traffic headwinds from search engine shifts and a move away from low-value work. Services revenue was $35.5 million, accounting for 39% of total revenue, and grew 3.4% year-over-year, driven primarily by the e-commerce offering (AutoDS). Total revenue reached $90.9 million, beating guidance.

• Full Year 2026 Revenue: Raised to $374–$380 million, up from the previous range of $356–$372 million.
• Full Year 2026 Adjusted EBITDA: Raised to $61–$65 million (16.7% margin midpoint), up from the previous range of $52–$62 million.
• Q4 2026 Revenue: Expected to be $80–$86 million.
• Q4 2026 Adjusted EBITDA: Expected to be $6–$10 million.
• 2027 Outlook: Management anticipates a steeper decline in marketplace traffic through 2027 due to the shift from search to AI-driven discovery. While operating expenses are expected to decrease year-over-year, this reduction will likely be offset by traffic/demand headwinds, creating downward pressure on 2027 adjusted EBITDA margins. Full 2027 guidance will be provided on the Q4 call.

• Search Engine Disruption: Continued compression in low-value transactional work and declining active buyers (-25.8% TTM) due to SEO disruption and reduced organic traffic from search engines.
• AI Agent Adoption Uncertainty: The transition to AI-agent-led discovery is uncharted territory with no established best practices. Agents lack brand loyalty, potentially making customer retention and acquisition more volatile.
• Traffic Deterioration: Expectations of worsening traffic trends continuing into Q4 and throughout 2027, posing significant headwinds to marketplace growth.
• Brand Loyalty Erosion: Risk of losing direct customer relationships and brand affinity as transactions increasingly occur through third-party AI agents rather than directly on the Fiverr platform.

Q: Analyst asked about the quality and trajectory of AI-sourced GMV from LLM channels, noting a 64% QoQ growth. / A: CEO described the shift to AI agents as an 'earthquake,' noting that while volume may be lower initially, the traffic quality is higher because users have formed intent. Agents lack brand loyalty and prioritize output maximization, so Fiverr must optimize for outcomes within agent ecosystems (like ChatGPT/Claude) rather than relying on traditional search or ad clicks. The goal is autonomous agent interaction without leaving the assistant.

Q: Analyst questioned if AutoDS momentum was a one-time campaign pull-forward similar to prior quarters. / A: Management clarified that AutoDS growth is multi-vector, driven by software solutions, AI content generation tools, and efficient marketing campaigns. The specific Q3 beat was aided by a successful partnership campaign launched in late August, which is expected to sustain momentum into Q4, justifying the upward revision in full-year guidance.

Q: Analyst asked about the economic model for LLM partnerships (take rate, customer ownership) and repeat buyer handling. / A: CEO stated the economics remain unchanged; these integrations serve as new top-of-funnel discovery layers focused on outcome optimization. It is too early to define specific take rates or long-term repeat buyer dynamics, as the primary objective is meeting customers where they are shifting their discovery behavior—from browsing to agentic action.

Q: Analyst inquired about the phase-in of 'Mira' (AI headhunter) across categories in 2027 and whether spend-per-buyer growth showed signs of weakness. / A: CEO explained Mira acts as a sophisticated matchmaker, qualifying both talent and clients to reduce friction. Expansion in 2027 will be cautious and data-driven, surfacing Mira only when it improves success metrics. CFO noted that while spend-per-buyer continues to grow due to higher-value projects ($1k+), the pace has slowed slightly due to broader traffic headwinds affecting overall transaction volume.

Q: Analyst asked about the future of SEO/GEO and risks to brand loyalty as AI agents handle transactions. / A: CEO predicted traditional SEO will become obsolete as search engines increase ad density to compensate for lost organic traffic, rendering GEO ineffective since agents prioritize output over process. Regarding brand, he acknowledged high risk as agents have no loyalty; however, Fiverr aims to differentiate through its data assets, trust, and fulfillment reliability, ensuring its platform is chosen by agents even if the user doesn't interact directly with the brand.