Upwork Talent Scout
Client Onboarding Experience
The Talent Scout client intake flow is Upwork’s onboarding process that gathers client needs and signals to help recruiters find the right talent. This service saves clients time and addresses the challenge of finding high-value talent not available in Upwork’s Marketplace.
Recruiters in Talent Scout act as intermediaries, assisting clients in finding pre-screened, skill-certified candidates for roles in development, design and creative, marketing, and product management.
Role
Design lead
Collaborator
PMs, Engs, Data analysts, Content designer
Design Process
Pain points and problem statement > Ideation > Solution/concepts exploration > User testing > Design and content iteration > E2E experience mapping out > Design system update
Duration
5 weeks
Tools
Figma, User testing.com, Miro
Outcome
Increase user adoption, better engagement, and an 73% improvement in client conversion. Reduced the no-show rate from 40% to 16%
Business Pain Point
Business Goal + Hypotheses
- Increase client sign-up conversion
- Reduce drop-off rates during sign-up and intake flow
- Boost first meeting conversion
- Cut no-show rates
- Signup is burdensome for clients
- Clients drop off due to long wait for first meeting
- Intake experience is not engaging and intuitive
Understand Users’ Problems
Understand Experience Issues
- Left and right columns increase cognitive load
- Low-domain clients unsure of role categories
- No back button to return to Talent Scout landing page
- Non-tech-savvy clients unsure of needed roles/skills
- Clients wanting a website may not know they need an HTML developer
- Suggest search corrections for misspellings
- Clients unaware of remaining steps
- Willing to provide more info if it helps the recruiter
- Need visual confirmation for added skills, the selected skills are added to the same “cloud”
- Too much info required
- Clients unclear on why info is needed outside of development
- Illustration doesn’t match non-development categories
- Users lose interest/patience with excessive info requests
- Some drop off due to phone number requirement
- Small businesses worry Talent Scout favors big companies due to logo placement
- Want clarity on first meeting benefits
- Mobile users missed date and time picker initially
- Clients want to share more about the role, doubting recruiter’s ability to find the right talent
Ideation
> User Journey Map
We compiled the data and created a journey map with stakeholders to address problems and brainstorm solutions for the end-to-end experience. The team identified and voted on key HMW questions, with the top one being:
“How might we collect as much information from the client as possible before the recruiter call, without it feeling too cumbersome?”
Potential solutions:
- Only show the primary value prop to reduce cognitive load
- Provide options to shortcut questions for non-tech-savvy clients (e.g., “Just want to talk to someone”)
- Keep the client curious and excited about the possibilities
- Gather enough info to present a shortlist of candidates at the first meeting
- Craft a compelling value proposition and motivation to encourage clients to schedule their appointments and show up on time
- Add the progress bar showing where they are in the process
The user journey map clarifies specific user problems at each phase and suggests opportunities and solutions.
Information Architecture
Based on our findings, I collaborated with the engineering team to create the information architecture and user flow, ensuring there were no dead ends.
Low-fidelity Prototype
I listed, visualized, and prioritized potential solutions with stakeholders. The primary solutions after discussion and voting are:
- Add questions on budget, hourly rate, and timezone to shorten waiting time for best-match talent (Persona A & B)
- One-click “skip all” CTA to let low-domain clients (Persona A) talk to a recruiter directly
- Real talent profile samples to engage clients based on their info (Persona A & B)
- Simplify sign-up by removing phone number and company name, and adding a pre-fill function
High-fidelity Prototype
I created a high-fi prototype for user testing to confirm if the solutions effectively solve our users’ problems.
Think-aloud Test and Iterations
To see if our solutions address Persona A & B’s problems and fit their mental models, we conducted a think-aloud test. The right picture shows our key findings and recommendations.
Reorder intake flow from category -> intake option -> create account -> scheduler to category -> scheduler -> intake option -> create account to match clients’ mental model/expectations (get help from recruiter).
Merge input fields (name, company name, email) into the meeting scheduler. This helps clients understand why the info is needed and auto-fills the sign-up form afterward.
Add a summary page as a “waiting room” for optional info.
Keep users focused on intake questions by removing the interactive map and profile sample (show them later) to reduce cognitive load.
Conclusion and takeaways
Upwork’s focus on improving user experience really paid off. They saw a big jump in user adoption and engagement, which is awesome. But the real highlight? A 73% improvement in client conversion rates. And the new experience reduces the no-show rate from 40% to 16%. Besides, the recruiter’s work efficiency of finding the right talent has been enhanced from 3 days to 1 day. That’s huge! It just goes to show how important it is to keep users happy and engaged. Upwork’s commitment to making things better is clearly working, setting them up as a top player in the freelance market.
- Inviting the engineering team to the loop earlier would be a great approach to talk about the technology limitation and feasibility
- Fork experiences for different user types are essential in the onboarding experience to retain the user
- The hypothesis should be built on the data or be test proved
- We need to find a balance between providing enough content to intrigue users and eliminating some content to reduce their cognitive load
Like what you see? Hit me up ：)