Master's thesis: Applicant Tracking System

UX/UI

SaaS

My master's thesis was done in cooperation with Teamtailor, an Applicant Tracking System (ATS). I conducted user research about how recruiters use an ATS and created a design proposal for a candidate profile in their ATS.


What I did:

• User research (contextual inquiry and content analysis)

• User journey-like graphic of users (recruiters) throughout the recruiting process

• Wireframing

• Hi-fi design in Figma

• Usability testing in Figma

• Written report and oral presentation in English

Although the project was a full design cycle from user research to a finished UI design including usability testing, the emphasis was on the initial parts of the project, i.e. user research and ideation, and to find solutions to UX issues. While working with Teamtailor during this period (5 months) I learned a lot about B2B SaaS companies and how they work. This was an individual project.

Problem description

In the applicant tracking system in Teamtailor, the current design of the candidate profile is filled with information and functionality that has been added as the platform has grown over the years. Currently, the candidate profile displays a lot of information in a crowded space requiring a lot of scrolling. New information such as results from competence quizzes, meeting feedback from interviews, survey responses, and references is added as the candidate proceeds in the recruitment process. Therefore, a redesign of the candidate profile is necessary, where the display and structure of the page are optimized for an efficient recruitment process for the recruiter.

Process

A user-centered design approach to designing a candidate profile was applied. User research consisted of contextual inquiries and content analysis of user reviews, to understand what activities and information are relevant for the recruiter throughout the recruitment process. The contextual inquiries focused on how recruiters screen an application, and the activities tied to an interview with a candidate.User needs were defined, ideas were sketched on a whiteboard, and wireframes of the design were created in Figma and tested by users. Finally, a hi-fi design was created based on the usability test feedback, and a conclusion with recommendations for designing a candidate profile in an ATS.

Results

The result from the user research revealed that many recruiters have their own way of working with the recruitment process, leading to one of the recommendations for designing an ATS being to let the users customize parts of the candidate profile. Additionally, some recruiters use ranking or scoring of a candidate, while others do not. For those recruiters that use scoring, the score should be based on the job application and not stick to the candidate if they apply for another role.

The results from the user research also revealed that reducing clutter, unnecessary information, and scrolling in the candidate profile was requested by users. Right before an interview was held, the recruiters needed to remind themselves who the candidate was. Additionally, the recruiters had candidate-specific questions to ask during the interview. This finding led to another recommendation for designing an ATS, to have a feature to let the user highlight interesting parts of the candidate's application.

For the design deliverables, I listed around 80 qualitative user insights, a user journey-like diagram about relevant actions and information thorughout the user journey, as well as improvements for the ATS to increase the users' efficiency. These deliverables was aimed for the development and design team to use when they tackle the full project of the candidate profile redesign. 

Additionally I designed a hi-fidelity design proposal for the candidate profile using the recommendations and insights I discovered, as seen below.

To dig a bit deeper into one specific part of the project, the toolbar got a complete makeover. In addition to collecting some of the most important actions such as scheduling meetings and opening the messages, a customizable button where the user (recruiter) can choose what action they often use, in order to become more efficient when handling many applications. The user can then choose from moving the candidate to a specific state that they can choose (e.g. a "not interested" column), performing an action that isnt available immediately at the toolbar (i.e. send NPS suvey), or a a dynamic action suggested by the system depending on the recruiting stage the candidate is in.

The initial user research showed that many recruiters use the ATS in their own way. By allowing the recruiters to customize the toolbar accordiing to their own needs, the ATS can improve the users' efficiency.

Future work

Future research could focus on which scales (linear or non-linear) a recruiter could best use to assess a candidate, based on the impression gathered either when screening an application or after an interview. Another area to further research is more reliable data on what candidate information is prioritized by the recruiter. 

You can find the full thesis at: https://umu.diva-portal.org/smash/record.jsf?pid=diva2:1782888