Redesigning outdated mobile app with user experience focus.
Kariyer.net is the biggest recruitment platform in Turkey, founded in 1999. It has a huge job seeker CV database and also companies enrolled around Turkey.
Kariyer.net's vision is to help job seekers not only find the best jobs but also help them through their career journey.
The technology infrastructure was outdated and the user experience needed to better align with the vision of helping job seekers find the best jobs.
I was the UX designer and worked in close collaboration with a UI designer and a product manager. I was responsible for entire mobile app's user experience.
One of our challanges was time. We have had 6 months to release the new app from the beginning of the project with requirements gathering, and design. Other challange was that we had to keep the critical features going as the current app has been used for a long time. Beside that, we had to create an MVP from the current app along with some improvements.
We have come together with people from customer service, marketing and also some managers to put the pre-existing insights together.
Key outcomes were:
I detected some usability problems based on 10 usability heuristics by Jacob Nielsen. The problems were mostly about lack of consistency and standards and lack of aesthetic, minimalist design. You can see a short video for a sample problem.
This is one example of issues. Here it's really difficult to type and find what you need as the system controls every letter you type and so you need to wait to type another letter. One other important thing on this filter page is that filter titles and text alignments were not carefully designed.
Many users were complaining about the inconsistent experience on different devices. They were able to handle some key features on the desktop website but cannot achieve on the mobile applications.
Job seekers have different abilities, different needs, wishes. However, app treats all the users as a single type of a user. Such as, not personalised search results. For intstance, users saw results from all industries and skill levels, rather than personalised to their professional background.
Top issue caused by disorganised information architecture was about user flows. While some information which was used by a little group of users were only one click away, some frequently used information were more difficult to reach.
The impacts of works done have not been measured. Teams were assigned to create features instead of fosucing on user experience. That situation resulted in creating lots of infrequently used features.
I quickly interviewed with 6 users whose needs and goals might differ from each other. I focused on user behaviours and device usage difference such as frequency of searching for jobs, the way they search for jobs (filtering, keywords etc.), difference on habits between app and desktop usage.
I got one important insight from every user. They complained about that they applied for the jobs but nothing got back to them. Any company did not get in contact or they did not see any updates about the job status if they hired another candidate or whatever. Another big issue was about irrelevent job results. They also pointed out that it was always hard to update the information on resume/profile through mobile app and also some of them said that they needed some resume sections (for example "certificates") which already exist in desktop version.
I conducted user testing with the current app based on different scenarios. I chose the samples by considering existing personas. Some of the user profiles were new graduated and looking for a job, students looking for part-time jobs, students looking for internship, professionals with 5+ years experience.
Scenarios were based on job search, saving a job, creating/updating the resume/profile, creating/updating the cover letter.
Users complained about irrelevant search results. They had some difficulties while updating/creating resumes, finding saved jobs because of usability issues and poor designed information architecture.
The old personas were demographics centered. To learn more about our users' patterns, I worked with people from marketing and the product manager to rewrite the personas based on roles and needs. These helped us align on who are users are and build emphaty and shared understanding.
I decided to use proto-persona template based on Lean UX not to lose so much time on creating personas.
We agreeded on some focus points for solutions to user problems. We needed to focus on personalisation, informing users more through their career journey (such as informing about closed job posts they had applied), focusing on providing different benefits instead of only helping users find the best jobs.
I wanted to see the images in our users' mind to design a better journey. We have gathered some of our users in a session which they can freely draw their own Kariyer.net mobile application.
What I achieved from this session were multiple outcomes. It was beneficial in terms of designing a better information architecture, the language that the platform should speak to their users, what they have frustrated so far or what they have wished to have through their journey. This research method also helped me identify the user types closely.
One of our challange was that we had to sustain many features in the new app. So, MVP had to include a lot of features.
There must have been some "must have" features in the initial product and the ones "not necesarry". Here we had to detect these stories by the help of insights and data.
We created a red route table depending on insights and data to see the frequent usage of functions into the app. The output helped us determine the priorities better.
We, as a product team, have collaborated with people from all departments in the company for storymapping session. Getting all together helped us develop a shared understanding.
After storymapping, we got together again for collaborative design session. As lean UX methodology suggests, I tried my best to encourage keeping the collaboration between departments to maintain shared understanding.
All the people contributed remarkably to the session. They all sketched out wireframes on the paper and everyone presented their masterpieces. After the session, product manager, UI designer and me came together to review the sketches and ideas.
I had already studied the old information architecture and detected some issues especially by means of user testings with the current app but I needed further studies to better shape the IA.
Both close and open type card sorting gave me clues about what was wrong with the current IA. This has showed me also that users were confused because of wordings on the current system. That negatively affected people's way finding abilities through navigations. I collaborated with content managers to find solutions about writing issues.
I wanted to see if my early assumptions forming the IA goes well. So, I quickly created prototypes for some flows and got out of the building to test them. The flows I've focused on (see the flowchart below) were based on search and application views by recruiters.
I collaborated with UI designer to shape our early assumptions and create prototypes. We worked on navigation, search flows, filtering options. In this ideation phase, our aim was to test our assumptions mostly based on information architecture solutions.
Here are some iterations on high fidelity wireframes. As I stated in the card sorting section, users had problems with some wordings and so we tested even some wordings through iterations to catch the best language we needed to speak to users.