Negative Points

• Upwork Inc ( NASDAQ:UPWK ) faces persistent near-term AI and macro headwinds, with an acceleration in AI-related automation impacting demand for low-complexity work.

• Google SEO changes have dampened new customer acquisition, with accelerated negative impact in Q2 and expectations of further deterioration in Q3 and Q4.

• Active client count declined to 763,000, pressured by AI automation, search trends, labor market conditions, and the company's strategy to focus on client value over volume.

• Full-year 2026 revenue guidance was lowered to $730 million to $750 million, reflecting the heightened pace of AI automation and no improvement in the labor market.

• Q3 2026 revenue guidance of $176 million to $184 million implies a sequential decline, and adjusted EBITDA margin is expected to be lower at 29% midpoint.

• The pace of AI-related automation has accelerated, leading to a pull-forward in erosion of the business segment most exposed to AI, with no clear timeline for stabilization.

• Free cash flow in Q2 was impacted by one-time restructuring payments, and the company incurred restructuring charges related to the $70 million OpEx reduction plan.

Q & A Highlights

Q: How is Upwork navigating the challenging operating environment, and what is the company's view on visibility and the SEO challenges? Additionally, how might the new MCP server impact the business going forward? A: Hayden Brown (President and CEO) explained that the company is in the midst of a work transition, with an acceleration in AI automation of low-complexity work and further deterioration in Google SEO, particularly in the non-brand channel. However, he emphasized that the strategy is working, citing green shoots like the $330 million GSV run rate in AI categories, eight consecutive quarters of increased GSV per client, and record average hourly work contract length of 100 hours. Regarding the MCP server, Brown stated it opens Upwork to integrating agents on both sides of the marketplace at scale, allowing clients and freelancers to use Upwork within their preferred AI tools. This is a response to agents already attempting to work on the platform, and it positions Upwork uniquely to offer humans at scale on demand, which no other marketplace can do.

Q: What is the expected duration of the transition to AI-driven work, and how much of the current active client base will transition to higher-quality AI-driven work versus needing to be rebuilt? A: Brown acknowledged that predicting the exact timing is difficult but framed it as an evolution where Upwork will emerge as a structural beneficiary due to its unique ability to serve the market where it's going. He noted that the company is focused on executing strategies that are working, such as enterprise growth, Business Plus (up 174% year-over-year), and AI tailwinds. On the client base, Brown described it as an expansion, not a replacement, noting that small business customers are shifting their workloads to be more AI-oriented, and there is strong resonance with customers wanting the combination of Business Plus features and AI transformation. He cited the significant growth in AI Strategy Consulting as evidence that customers are using new tools for different types of work.

Q: Can you provide more detail on the slowdown in AI category work growth (from 50% to 40% to 22% year-over-year), and what drove the higher-than-expected marketplace take rate? A: Brown clarified that the AI category metric is an undercount of actual AI work on the platform. Customer behavior is changing; they no longer specify AI in job posts because it's now an expected capability. A recent survey showed nearly half of talent reported their most recent jobs were AI-related, while only 16% of job posts explicitly referenced AI. On the take rate, Brown attributed the overperformance to successful monetization efforts, including dynamic pricing on freelancer fees, and stated the outlook is for the take rate to continue stepping up in the back half of the year.

Q: What is the latest on Lifted's progress, including customer migrations and pipeline building? A: Brown reported great progress in Q2, with customer migrations starting in June and ramping up due to positive feedback. The pipeline remains healthy, with customers advancing through the funnel. He highlighted a recent win of a multimillion-dollar RFP with a large global company for a staff augmentation contract, which would not have been possible before Lifted. The EOR solutions, enabled by Ascen, grew 29% year-over-year and are improving gross margins. The company remains on track for approximately 25% GSV growth for Lifted in 2026.

Q: Has the impact from Google SEO changes stabilized, or is it still ongoing? A: Brown stated that the impacts accelerated in Q2 and the company's outlook assumes they will continue to worsen through Q3 and Q4. He noted that while there is volatility in the industry, they have not yet seen stabilization, and any stabilization would be better than the current guidance, which anticipates further deceleration.

Q: What is the monthly cadence of when things started to weaken, and has the thinking changed on the portion of GSV susceptible to AI? A: Brown explained that at the time of the last earnings call, some trends had stabilized, but over the past few months, they have seen an acceleration in the erosion of the part of the business previously identified as more exposed to AI automation. He characterized this as a pull-forward in that activity versus prior expectations, rather than a change in the overall assessment of susceptibility.

Q: How is AI impacting freelancer work in the enterprise segment, and how does Lifted help insulate the business? A: Brown explained that Lifted was built to tap into the full $650 billion enterprise market, not just the 10% historically spent on freelancers. By serving customers across all work types, the business becomes more insulated from AI-related risk. If enterprises change their workloads and bring AI to different parts of their business, Upwork can bring AI talent to bear and help them lean into those strategies, while also being protected from any single part of the business being automated at a faster rate.

Q: What are the near-term customer acquisition channels being leaned into, and how was the $5 million to $10 million incremental marketing spend determined? A: Brown identified international SEM as a strong area with good returns, along with other channels like Meta, Reddit, podcasts, and connected TV. The incremental spend was determined through extensive testing in the first half of the year, arriving at an optimal range based on performance and degradation curves in different channels. The company is confident in deploying capital efficiently and expects to see benefits even in Q4 of this year.

Q: How is Upwork managing monetization levers like Connects and dynamic pricing in the current environment? A: Brown acknowledged that the marketplace has become more competitive for jobs, which has both positive and negative factors. The company balances how much it charges on the talent side for job access while focusing on using these levers to drive match quality and fill rate. In places where the marketplace is more congested with talent, there is an ability to adjust pricing beneficially for both take rate and matching quality. The company will continue pulling these levers in the back half of the year within its guardrails around match quality.

For the complete transcript of the earnings call, please refer to the full earnings call transcript .