Note: this story is published on @theuxblog, view it here.
Telling The Tale of Finding Freelancers
A UXD Case Study.
Kent graduated with a Bachelor of Accounting, a field he never really intended to pursue. Fortunately for him, his side project — selling cameras on eBay began to take off.
Now a successful business a year later, the website he built himself needs to be replaced. Without any outside experience in the process and without a network he can lean on, he’s not sure where to begin.
Kent may be fictional but his circumstances aren’t.
Accenture estimates that by 2020 43% of the US workforce will be freelance. With an increasingly fragmented marketplace, how will SMB owners without specialist skills or knowledge like Kent find the right talent for projects?
This is the problem area explored in the development of Telltale.
Discover & Define
Research on blogs and forums framed an idea of the issues that SMB owners may be having, giving the hypothesis:
The existing ways of finding freelance work (e.g. Upwork) are full of problems. SMBs need a solution that helps them discover trustworthy and talented freelancers and agencies, not just cheap ones.
Interviews with four owners of small and medium business in Australia revealed something surprising.
Finding a good Agency isn’t really a problem.
Sure, it’s a long process that involves a lot of retracing of steps. “It was easier buying my house” one of the users even remarked. Ultimately though they eventually find good agencies and freelancers.
There went my problem , or so I thought.
Slightly deflated, I went back to the transcripts with a question.
If SMB owners are ultimately finding quality agencies or freelancers, why wasn’t the process shorter?
Something that stood out was how circular the task was. Owners frequently get right up to the point of signing on with an agency or freelancer then mysteriously withdraw to start the whole process again.
As it turned out, there was a clue to this behavior that I’d overlooked. SMB owners speak to people in their network a lot during the process for advice and validation of their ideas. Their network is made up of people in similar positions who, like them are not experts on the topic. This is a problem.
These owners aren’t getting the expert advice that they need to correctly scope and brief a project. When they eventually spoke to an agency with expertise they would discover issues with their brief that took them back to the beginning of their search.
I realised the challenge isn’t to help SMB owners find agencies, it’s helping them find the right advisors.
With this, the problem statement became clear:
How might we give SMB owners accurate knowledge on-demand so that they can make the right hiring and procurement decisions?
At this point I’d typically take a broad, iterative based approach towards a solution. This project however was for a UX course at General Assembly (which I highly recommend) and a learning objective for myself was to develop my app design skills so the solution had to come in the form of an app.
Develop & Deliver
User goals and pain points from the research were fed into the primary persona, Adam Briggs. Potential features were mapped on a two axis grid of execution difficulty versus user benefit giving a list of prioritised features:
- Consultant matchmaking with filtering on location, price, experience and skills
- Set, clear pricing
- Evaluation indicators such as: consultant quality assurance, social proof and a project history
- In-app native live chat with file sharing
- Saved profiles
With the user goals in mind and MVP features listed, the user flow (above) could be mapped out along the happy path.
The first direction explored was conceptually similar to a dating app; understand the user then try and match them with an appropriate advisor.
To users this was the matchmaking equivalent of Google’s ‘I’m feeling lucky’ button. The decision is too important for them to trust some black-box methodology — they wanted to explore their options before committing.
Different flows were tested with users and it became clear that one focused on discovery best matched the user’s mental model and therefore the most intuitive. Rather than matchmaking relevant discovery became the focus.
The discovery and search screen became the most important part of the solution for users.
From paper prototyping through to hi-fidelity, designs were usability tested on users using clickable InVision prototypes. Linking with sketch allowed for on-the spot changes to the design between tests — rapidly developing the prototype. Some of the most important changes from these tests included
Labeling: From the paper prototype it became clear that labelling of the icons was extremely important for learnability.
Icons: The interface used a star rating system so a bookmark icon was used to represent the ‘saved’ profiles in an attempt to reduce confusion. Testing showed that in despite the overlap in meaning users were more comfortable with a star icon representing both a rating and a saved users.
Mystery meat: Some affordances just weren’t smelly enough: the user wasn’t sure what was clickable and what action it would result in. This was a particular problem for the search menu; they didn’t know it was clickable! To fix this I tested a more native iOS design pattern (grey search bar with darker entry field). This improved success rates although users were still getting stuck on searching versus filtering. Introducing colour to visually separate these affordances resolved this problem all together.
The entire project’s flow was usability tested with task based tests asrecommended by NNG through to the hi-fidelity prototype.
A hi-fidelity interactive prototype that users are delighted to use! Interact with the live prototype below or click here to view on InVision.