Problem
A large number of first-time candidates on Reed.co.uk were dropping off from the registration to job application journey due to the long application process.
Solution
Role
Product Designer and User Researcher
Team
Project Manager, Product Analyst, Development Team
Design process
Discovery
Analytics
Competitor research
Qualitative usability testing
Exploration
Sketches/wireframes
Lo-Fidelity Prototype
Qualitative usability testing
Quantitative testing
Solution
Final UI designs
Quantitative testing
Handoff for development
Analytics
Year on year, there has been a reduction in job applications submitted on the Reed.co.uk company website. We were aware of contributing external market factors but knew the design and experience of the application journey could be improved.
76% of applicants drop off from the application process while trying to create an account and complete their first application in the existing registration to application journey.
Overall, only 1.9% of users without an account who land on the job details page then convert to registering and applying for a job. Hotjar recordings showed users were struggling to sign in or register and some users eventually dropped off.
With 70% of traffic coming from aggregator sites, these users will apply for a job via Reed and will need to create an account in order to do so.
Research objective
We ran a research study to investigate why the jobseeker registration journey via an application is performing poorly. We wanted to understand if users find the registration process too long due to the interruption of their workflow (applying for a job), whereby we hypothesise that we ask for too much personal information. The results will be used to improve the design of the current experience.
I conducted moderated usability testing sessions with five users. I asked them to search for a job and apply for it by registering for an account.
Key user insights
All participants were confused as to why they had to re-enter the information that was on their CV
Users were surprised to find the journey wasn’t as bad as they had anticipated, but we know from the analytics that users still dropped off
The problem
The current journey presented the user with multiple obstacles.
The user has to re-enter the same information into the form which is already on their CV
The form is too long and takes too much time to complete
I created how might we statements which would help me to find solutions to the problem.
How might we require less manual input from the user when inputting their information?
How might we get a user to create an account and submit a job application faster?
From this, I compiled all the feedback and insights and grouped similar ones. This helped me brainstorm and develop potential ideas and gave me a clearer view of what is important to the user.
Concept development
We tested a concept to explore how the order of the journey can potentially be rearranged by putting the CV upload at the front of the journey, followed by a shortened form.
This proposed journey will reduce
The number of steps in the journey
The amount of information asked
Manual input from users with the help of the CV parsing feature
Further user testing
The concept was tested with participants again via qualitative research methods.
Participants said that the information asked of them was expected and the CV parsing function helped them to prefill their information
One participant said the form “Gets straight to the point as opposed to making an account and filling in your details. It seems quite quick”
AB testing
The concept was validated via an AB test whereby the current journey was simplified to mirror the intended new application journey. The screens were reordered and secondary fields were removed. This test showed there was an uplifted registration to the application conversion rate of +4.6%. This increased confidence that the new journey would see an uplift in applications during the registration process.
The solution
Reordered journey
The CV upload would be the first step after the user has to register for an account (inputting an email address and password). This would enable them to provide their information upfront or build a CV if they do not have a CV document.
CV parse
This feature will scrape their CV and auto-populate the requested form fields with data uploaded from the candidate’s CV. The CV parsing feature will scrape all the information it can into the appropriate form fields. This will remove the user's pain point of having to re-enter the same information again, saving them both time and effort.
Simplified form
The form will be simplified by requesting only primary fields including name, mobile number, location followed by what they are looking for to allow Reed.co.uk to send the candidate job recommendations after submitting their application.
Improved interface and experience
The overall CV uploading and form completion experience will see an improved user experience and interface that will guide the user through the journey.
Design refinement and handoff
After the research phase, I spent the next few weeks iterating and refining the responsive screen designs for both mobile and desktop, ensuring the best user interface and experience across all touchpoints in the journey. This also includes an option for a desktop user to drag and drop or manually select their CV file.
The new experience needed to support the technology powering the CV parsing tool, so I worked with the engineers to understand these constraints. Following the candidate’s CV upload, there is a short pause as the system scans the CV for viruses, uploads the CV to the candidate database and then extracts the relevant information into the application form. Effectively communicating to users that this delay is normal and part of the process was addressed in the messaging on the progress modal, reducing the risk of users dropping off from the journey.
While the form was simplified, certain additional fields including the 'What are you looking for?' section could not be removed due to their relevance in collecting essential data for other business areas. Consequently, a compromise was reached, balancing the streamlining of the form with the need for data collection.
As part of the handoff process, I regularly communicated with the developers to ensure the designs were implemented correctly on the new technical stack under the technical transformation strategy.
The impact
The new journey was AB tested to ensure there was no harm to application levels.
33% increase in conversion rate from registering to applying
As this test confirmed that the new journey was performing significantly better than the old journey, the new experience was pushed live to 100% of users. The same journey and experience have been reflected on the organic registration journey (registering from the navigation header). This means that all users who register will be able to have their CV scraped into the form to speed up the process of creating an account.
This case study was submitted to The Digital Experience Awards by Contentsquare for the “Best Digital Experience Innovation” award. The case study was shortlisted and won a silver award.
Conclusion and learning
It would have been harmful to the business if the journey was released knowing there could be a reduction in the number of applications made, therefore the team needed a high level of confidence in the journey performance before releasing it to all Reed users. The extensive testing efforts not only increased team and business confidence but also validated the hard work and patience involved in this project, as seen by the positive results.
Additionally, I also learnt how important communication between different product teams is to make sure the changes made to the journey would not harm other business areas. This slowed down the release but the communication was necessary to ensure the new experience had a positive impact on both the business and the user.
