Upwork AI recruiter
My role: Lead Content Designer
My contributions: Research, content strategy, taxonomy, user testing, end-to-end user flows, UI/UX content design
The challenge: 26% of clients think freelancers require excessive oversight after hire, leading to dissatisfaction and cancellations
The solution: An AI recruiter that detects at-risk projects and helps clients replace freelancers
The outcome: A 20% reduction in subscription cancellations and 15% lift in positive project outcomes
The challenge: improve satisfaction and reduce cancellations
Upwork is a global, two-sided platform that connects clients and independent freelancers across thousands of skill sets.
For this 0-to-1 MVP, the goal was to improve client satisfaction and reduce subscription cancellations. Here’s why: 26% of clients thought freelancers required too much oversight after hire, a major conflict with the quality messaging themes at the top of the funnel (right). Meanwhile, 58% of clients had 10+ contracts at a time, further compounding the issue when freelancers missed deadlines or didn’t meet expectations. Year over year, Upwork saw increased client dissatisfaction, an uptick in NPS detractors, and high subscription cancellations.
The solution: AI and a human in the loop
In the initial discovery process—including a listening tour, quantitative research, and external data from McKinsey and Nielsen Norman Group—the team found that more than 90% of organizations wanted a human in the loop for their AI solutions. These data informed our strategy to combine an AI agent with an Upwork recruiter, helping us balance the needs of top-tier Business Plus clients with the organizational cost of human recruiters.
So how did it work, exactly?
When a back-end signal indicated that a project was at risk—a missed deadline, for example—Upwork’s LLM Uma crawled the project metadata to find comparable freelancers who were available immediately. Uma created a shortlist for the client, then generated bulk messages to remove some of the admin work. Most clients self-serviced via the happy path and replaced their freelancers with Uma’s help; for the remaining 10% or so, a human recruiter stepped in with the same offering. This way, we improved client satisfaction without significantly increasing business costs.