Imagine logging in to find out how your job applications are progressing & being faced with this wall of information.
The dashboard active job seekers use to track the status of their job applications on Expert360 was confusing and needlessly complicated, causing users to rely heavily on the support team for clarification and updates.
It also wasn't mobile-responsive, difficult to scan, and offered very little in the way of updates and feedback — the core purpose of the page.
My goal as the sole Product Designer
To lead and facilitate research & design with multi-disciplinary squad to define the needs of contract workers seeking work and deliver a quick intuitive solution to those needs.
Key goals:
-
Reduce number of related support enquiries by 50%
-
Reduce bounce rate by 60%
-
Give users the context and guidance needed to understand where they stand in hiring funnel quickly
-
Work within the constraints of 2 sprints (4 weeks), using as many existing components from the existing Design System as possible
EXECUTIVE SUMMARY
How we got there — scope narrowing & many iterations
This project swiftly turned from a fancy dashboard vision to a simple project update page thanks to the combination of quantitative and qualitative research I conducted at kick-off.
-
Users didn't want to spend a lot of time on this page, they just needed to know what had changed, and if there was anything they needed to do. No frills.
-
Thanks to this sharpening of scope, we were able to work as a squad to workshop rapid prototype solutions for the core 'how might we' statements we aligned on, and test them early.
-
V1 was released without the project progress bar, which was added in V2, 2 weeks later.
Qualitative Analysis
User feedback reflected that users' experience of applying to a job was really poor, as only 3% of users win work and the rest are left in the dark — a lack of updates accounting for 81% of all related user feedback.
After analysing and categorising feedback from Intercom and the customer support team, I found that users didn't know where they stood in relation to the project, or what actions to take.
65% of the problems in user feedback related to shortcomings in project status updates, across users who were winning and losing jobs.
Application counts were misleading
"I assume I can disregard this project? There are already 5 people at interview stage"
Page was difficult to scan & key information was limited to email communication
"It doesn't show or say anything about being selected for an interview. If I didn't get the email from Expert360 I wouldn't be aware of it"
Users were heavily reliant on support for updates & clarification
"Is this project still active and taking applications?"
Project status was confusing because process is non-linear
“When you see the Interviewing words appear on a job, does that mean they have already selected the people they want to interview? And if you aren’t notified you didn't get selected?”
Quantitative Analysis
My goal was to understand user behaviour on the platform when in the phase of seeking work. I always this to understand more about the user goals by how users are currently interacting with the page.
I looked at data from Google Analytics, Fullstory and Mixpanel to understand the quantitative data behind the qualitative data I'd collected:
-
Who was visiting the page, and how often?
-
What did users do on the page? What actions did they take?
-
Which devices were being used?
-
What were the baselines for page load speed, bounce rate and traffic for the page?
-
How many job applications do users make at any given time?
Baseline statistics
-
Device: 90% users on desktop.
-
Average # applications: 3, max 15.
-
Speed to interview: average 3 days for hired projects.
-
Average Page Load Speed: 7.42 seconds (lots of rage clicking observed during loading times.)
-
Bounce rate: 73.79% (very high)
-
No bells and whistles — any information that doesn't directly meet these needs, was excluded from the design for first release.
Which means...
The user goals emerge
Active job seekers use this page to understand:
-
where they stand in the hiring funnel for their projects
-
what's changed, and;
-
what the next steps they need to take are.
"Who wants to have to look at 5 projects they lost every time they log in?!"
Consultant user on Expert360 platform
Workshop
I facilitated a workshop with the squad to align on some 'how might we' statements and early concept sketches that were technically feasible.
These 'how might we's' represented our best bets at solving the problems and providing value to users. They weren't intended to be rigid solutions that couldn't be altered or re-interpreted. Some of them were just too prescriptive and couldn't be tied to the problems. These were ice-boxed.
I also posted proto-personas on the wall for the team to gain the context of these consultant users.
-
It was important for the team to understand that most users were passive, and only visited the page when prompted.
-
This guided us away from bell and whistle solutions like real-time notifications and badges, towards simpler solutions that just solved the core problems.
How might we...
-
Help users find what they need, when they need it?
-
Update users on what has changed?
-
Make it obvious where consultants stand with a project and what to do next?
-
Provide a better feedback loop to a consultant that has lost?
Lo-fi wireframes of filters & project cards
I like to use whiteboards (as well as digital) to sketch. It allows me to iterate rapidly and explore concepts in a tangible way, instead of the team trying to picture things.
The biggest visual design challenge with this project, was finding a way to represent the project status, as well as the user's application status, in the same card.
-
Previously, we overused little pills with statuses, or relied on navigation to infer status, & relied on notifications to communicate the user's progression.
-
We ditched all that, and landed on the new concept of a process bar, with a user indicator to represent the project and user stages, and some kind of updates bar.
User Testing & iterations
This project ended up with the most iterations of any other project I've worked on — primarily because we needed to create robust new components for the design system for filtering, project cards and statuses.
I created several prototypes using sketch and marvel to test in person with 12 users over 3 rounds of testing — 5 who were actively seeking contract work, 6 others who had in the recent past. I tested different variables by tweaking components one thing at a time, like filters, project status and information.
Finding 1: Sidebar filter more discoverable
Users noticed the sidebar filter concept much more often over the linear.
Decision
The sidebar filter also made use of an existing mobile component in our design system, so we went with that.
Finding 2: Company details useful
Users loved seeing the company name for the first time & found the extra information about the company as clutter.
Decision
We resolved to surface just the company name and remove the industry and location to review again once we get feedback after release.
Finding 3: Progress bar easier at a glance
Users found the progress bar much easier and faster to understand. They also said they perceived the site to be more polished and trustworthy with this addition.
Decision
This would require significant tech effort. Because of this, we negotiated with the engineers to release the progress bar in V2, in an effort to deliver 70% of the value in the first release.
Some of the many iterations of the project card
The Solution
Seamless tailored dashboard for job seekers to build, track & manage their pipleline of work.
How might we make it obvious where an applicant stands, and what to do next?
-
Project progression bar — user knows where they stand relative to the project's progression at a glance, removing all confusion caused by ambiguous statuses.
-
Client updated time — users expressed a strong desire for signals as to whether the client had been active on the project or not, as a tool to deduce whether the project was likely to progress at that time.
How might we enable users to find what they need, when they need it?
-
Filter & refine - users can focus on their real pipeline, and filter out stale projects.
-
Active/inactive status bar — users wanted to get a quick snapshot of their true pipeline at any given time.
-
Personalised action bar — users can now see they have key actions pending for that project.
Results
They speak for themselves.
63.76%
Reduction in bounce rate
With traffic of 1.3K visitors a day, this was a huge win. Lots of positive feedback.
+31.8%
Increase in pages per session
A sign that users might have perceived more value in the site.
70.9%
Reduction in page load speed
From 7.42 seconds down to 2.16 seconds. All credit's due to the engineering team!
Oh my god finally! This is so much easier to understand.
Management Consultant and E360 user
"I love that I can control my view now. I can just focus on my pipeline and clear out all that other stuff."
Management Consultant and E360 user
TITLE OF THE CALLOUT BLOCK
Learnings
I learned to break things down into a true MVP. Otherwise we as a team can tend to lose sight of the core problems we're solving for and it'll slow us down in delivering value to users earlier to get that valuable feedback.
In this instance, there were so many problems and opportunities we could address, we decided to keep each other in check with our guiding principle, illustrated well in this graphic: