Upwork • Client Growth
Q3/2021
Background
Upwork is the work marketplace where companies can find and hire freelancers, and freelancers can find work. The Upwork.com homepage is the front door for different types of clients and freelancers with various needs and characteristics.
There are clients with different industry domain knowledge, hiring intent, urgency level to get work done, the duration of their work, the scope of work etc. They have different experience with hiring remote talent, and have different ways in which they integrate remote talent within their teams.
Similarly, there are different types of freelancer visitors. From individuals to agencies, with various expertise level, experience working independently, and experience working remotely.
To address different hiring needs, Upwork had introduced multiple ways to hire on Upwork. There are self-service options - Talent Marketplace, where you post a job and hire, and Project Catalog, where you can buy a pre-scoped fixed-price service. And, there is an assisted hiring option - Talent Scout, where recruiter can find freelancers for you.
Early in 2021, Upwork had launched a new industry category (work marketplace) and went through a total rebrand. It updated the visual style and messaging throughout the site, including the homepage. As part of the new brand story, homepage included a multi-door design solution that introduced different ways to hire on Upwork as mentioned previously.
The goal of the homepage was to build on an experience that translates the brand promise into a concrete understanding of how clients and freelancers can leverage Upwork. Guiding them to their “Aha” moment sooner to convert.
PROBLEMS
Business problem
With Upwork’s rebrand and introduction of multiple hiring options on the homepage, the Client Registration went down. The business goal at that moment was to fix it and increase the Client Registration on the homepage.
Data problem
Unfortunately, the quantitative results of homepage updates were contradicting. My team could not rely on quantitative data to build new hypothesis for homepage improvement. However, analytic partners were able to point to a change in traffic pattern such that 82% less traffic was going from homepage to the category pages. The data suggested that this could have been one of the reasons for the decline in client registration.
Shift in the framework
Together with the Product Management Director, we first decided to focus on the low-hanging fruit and prioritize the audit & improvement of the category interaction experience on the homepage. I created a strategy for category experience improvement on the homepage and shared it with stakeholders and upper management.
The category interaction updates were received positively. But looking at the impact the proposed updates could drive, it was evident that more changes would be needed in order to increase registration to the desired level. And so, I changed the approach to work on the future homepage vision and growth strategy instead of doing piecemeal updates.
Project goal and success
The goal of this project was to create a North Star for the homepage that explains the high-level narrative about the future vision; uncovers the growth opportunities; and guides the client growth roadmap.
Client Registration was a success metric prioritized for the homepage vision design.
DISCOVERY
kICK-OFF
I began the work by gathering information about the latest homepage strategy - the target audience, business, and user problems it was solving.
Then, I looked at the heat map and the engagement data to better understand how users engage with the content on the homepage.
By analyzing the collected data, I saw two problems that I wanted to explore further:
1. The target audience. There was an equal focus to clients and freelancers. The homepage content tried to address almost all visitors at once. I felt it could have been one of the leading problems for low client conversion.
2. Low engagement with the client-focused content (the multi-door solution with different hiring ways, and the category section on the page).
In collaboration with a PM and UX researcher, I had set up 2 initial discovery pathways: comparative analysis and the baseline usability study on the latest homepage design.
Comparative Analysis
The data analysis thus far led me to believe that focusing on all types of visitors is the leading problem for the low client conversion. In order to stay relevant to all page visitors, the information on the page was quite generic and shallow. My hypothesis was that content was confusing and did not build trust. Trust is the key element for creating the successful first-time relationship.
“If you try to please everyone you'll please no one.”
I did a comparative analysis of various landing page designs. I specifically focused on how leading brands address different types of visitors on their sites. This analysis paired with the best design practices, helped me to pitch for target audience re-evaluation. I suggested focusing on clients first.
Comparative Analysis Results
I presented my analysis and recommendations to the product management team. As a result of this presentation, the target audience for the homepage vision was updated as below:
Primary: Prospective clients 
Secondary: Expert talent looking for new work opportunities
QUALITATIVE BASELINE USER RESEARCH
I worked with the UX researcher to set up an unmoderated usability study to better understand WHAT visitors think when they land on the page and WHY they take certain actions on the page.
User Research Goal
Evaluate the homepage qualitatively in order to improve future iterations and add insight to quantitative metrics.
User Research Findings
The research helped in identifying the key questions that prospective clients asked after they saw the homepage. Those questions were then grouped into priority groups of "deal-breaking" questions and “good-to-know”.
The “deal-breaking” questions were around the need for seeing the talent sooner and learning about Upwork cost. The “good-to-know” questions were about the hiring process.
We also discovered that the multi-door approach created confusion in understanding what each hiring option offered. Clients also faced choice paralysis when asked to pick a hiring option that they would consider going with.

User journey analysis
Once I was clear on the questions that users had asked when they visited the homepage, I mapped the user journey and tracked the ease of access to the key information. This exercise helped in the development of content strategy and user journey updates beyond the page.
Findings
The key information was 2-3 clicks away from the homepage in the best-case scenario, or completely missing on the site. Also, the homepage didn’t have clear indications on where clients can find the information that they were looking for.
Define
Based on the baseline discovery and user journey analysis, I defined few opportunities for the homepage vision and create a design checklist for design explorations (content checklist, and client characteristics checklist).

How might we show freelancer profiles early in the experience?
How might we educate clients about the cost of hiring on Upwork?
How might we instill trust in the quality of freelancers and reduce the fear of getting unsatisfactory work outcomes?
How might we introduce and describe different ways to hire on Upwork?
How might we make content useful and usable for clients with any domain knowledge and hiring urgency?
Exploration & ideation
Various initial ideas
I explored design solutions, from minimal design updates to brand new ideas; with low to high implementation efforts; from safer choices to bolder experiments. Then I shared the initial design concepts with the stakeholders:
to inspire the team with possibilities,
to probe the scale of changes that the team wants to pursue, and

to gather preliminary feedback on the directions.   
The team decided to pursue big ideas to improve the UX.
Guiding principles
As I was exploring design solutions I crafted the guiding principles for the page content design. These principles clarified the persona priorities and execution rules.
Experience beyond the homepage
I explored how the homepage vision could work with the onboarding flows. And explored how the onboarding flows can be improved to provide cohesive storytelling and better UX.
I also explored how we can change the information architecture to make the key information easier to find.
I shared design concepts with my peers (PMs, researchers, and fellow designers) for critique before sharing the design with the stakeholders and the leadership team.
Low-fidelity prototype
I designed a low-fidelity prototype and demoed it along with the guiding principles to the stakeholders and the leadership team. During this presentation, I gathered the feedback on the design and created an open discussion to gauge the level of concerns that the team might have. This helped me to prioritize feedback and work on the wireframe updates accordingly.
I grouped the feedback into few buckets:
1 - concerns that cannot be answered without tests.
2 - opinions about the certain content emphasized on the page.
3 - nice-to-have or minor variables to the design that do not affect the homepage vision as a whole.
I tweaked the design and shared it with the stakeholders and leadership team again for the final review. I also communicated the next steps and address why certain feedback should be validated during user testing.
High-fidelity homepage design
Once the wireframe with the homepage vision was ready I connected with the research partners to set up the study. UX researchers recommended bumping up the fidelity of the design to collect more accurate responses. While the research team was busy with the test plan and recruiting test participants, I was working on the high-fidelity homepage vision design.
Validation
Homepage vision user research
I worked with 2 UX researchers to set up an unmoderated usability study to better understand WHAT people think when land on the page. We had set up this study for different personas - prospective clients with high and low domain knowledge, and prospective freelancers. We wanted to collect feedback from different types of prospective visitors.
Goals
1. Get directional validation of the main hypotheses:
"If the homepage content primarily focuses on client needs, it will provide more clarity to clients on how they can leverage Upwork, and so improve the client conversion.”
"Client-focused content is useful to freelancers. Freelancers still gain the understanding of what Upwork is and how they can leverage it.”
2. Evaluate the homepage design updates qualitatively in order to improve future iterations and setup quantitative test priorities.
Research findings
We got directional validation of both hypotheses and received many helpful insights for further design improvements.
The homepage came off as professional and compelling for talent, with clear action pathways. The content had built interest for the client test participants. They wanted to return to the live site to take an action.
Results
A/B TEST PRIORITIES
The homepage vision usability research findings equipped me and my team with the data to plan and prioritize the first few A/B experiments with high confidence.
ROADMAP UPDATES
When the team was equipped with the validated homepage vision, we worked on the growth opportunity exercise. In partnership with the product manager, I organized the opportunity mapping workshop for the client growth team. We looked at the homepage findings and other relevant studies and came up with test experiments to improve client conversion.
This exercise helped my team to uncover growth opportunities and have a blueprint of experiments for the roadmap.
First TEST RESULTS
Below are the results of a few A/B tests that were prioritized after homepage vision validation. Test results provided a quantitative validation of key hypotheses and also provided numerical data on the type of content that moves the needle when it comes to the Client Registration and Starts. 

The new section "Why Upwork"
Success metric Visit to Client Registration saw 6% lift. Design treatment got defaulted on the homepage.
Carousel with mini case studies
Success metric Visit to Client Registration saw a 3% lift. Design treatment got defaulted on the homepage. Visual design for this particular component was done in collaboration with my peer-designer.  
Categories section under the hero banner
The primary success metric of Client Registration saw an insignificant lift, but the secondary success metric of Visit to Any Hire saw 14% lift. Design treatment got defaulted on the homepage.
___
“Anna’s leadership on the homepage user research and vision design work brought out not only the best in the team but produced a framework for us to solve some key concerns for client prospects.”​​​​​​​
Daniel S., Senior UX Researcher
“Anna, thank you for your strong thinking and work on the Homepage Vision. Excited for that to come to life!”
Matt W., Director, Product Design
Project summary

Timeline
3 months from discovery to the blueprint of A/B tests
Cross-functional team effort
1 Senior Product Designer (myself)
1 Director, Product Management
1 Data Analytic
3 UX Researchers
1 Copywriter
Stakeholders
Senior Director, Product Management
Senior Director, Engineering
Director, Product Design
Lead Product Manager, Client Onboarding
Senior Product Manager, Client Acquisition
Fellow product designers
My responsibility
Data synthesis. Analyzing qualitative and quantitative data to form recommendations.
Research guidance. Setting up the direction for user research and usability studies.
Strategy development. Creating design & content principles. 
UX/UI design. Designing solutions that improve user experience and support Upwork design system.
Prototyping the experience.
Copywriting guidance. Setting up the messaging direction.
Feedback gathering. Conducting design reviews and critique sessions, interviewing and collecting feedback from stakeholders and the leadership team.
Collaboration. Work with cross-functional partners to gather data, prioritize and explore A/B experiments.
Back to Top