Talent Plus+
The Challenge
Talent Plus+ is a global talent discovery firm that is facing the challenge of having an overloaded interview team that continues to shrink, despite having over 400 clients. The team is burdened with manual processes, complicated excel sheets, too many distinct tools, and outdated UIs, which is leading to burnout and employment churn.
The Solution
Over a period of three months, I worked with the project management team and directly with the interviewers to assess specific blockers and update their toolkit. The solution was to unify each stage of the talent discovery process into one platform, which involved building new UIs for scheduling, conducting, and transcribing interviews. The focus was on expediting workflows to reduce burnout and consequential employment churn. By implementing this solution, Talent Plus+ was able to streamline their talent discovery process, reduce their reliance on manual processes and outdated tools, and increase efficiency and productivity for the interview team.
First Research Method: Immersion
One of my favorite parts of this project is how closely I got to work with the actual users. As part of my onboarding, I was scheduled to meet with most groups in the company using their questionnaire rubric to get to know the other employee’s histories and personalities. This really helped provide social lubrication for future conversations that required compromise. Onboarding also included observation sessions; watching the interview team actually do their jobs in real-time was also invaluable.
Throughout my tenure with Talent+, I was engaged in weekly product huddles and review sessions. Observing the behavior and reactions I was able to divide the interview team into two demographic groups. Group A were 50+ and had decades-long careers with Talent+ and intended to stay until retirement; this group was smaller but more likely to be in training positions and had lighter workloads. In terms of behavior, this group was very change-resistant and inflexible in review sessions. Group B were mainly straight out of college, were much likely to leave Talent+ in a span of < 1 year, and were more excited about updating the product.
Second Research Method: Tooling Pain/Gain Analysis
The interview team used quite a few disparate tools to workaround the shortcomings of the Talent+ platform. Namely, they used Salesforce for receiving incoming interview orders and scheduling (context-switching with Outlook for the later), Excel for candidate scoring and Clickup for ticket tracking.
After determining that the best solution for productivity would be to streamline all tasks into one platform, I reviewed each of these applications with the users to learn which functions they used most often and what sort of elements they liked/disliked so I could figure out what would be worth porting over to the new Talent+ platform.
Research Synthesis: Jobs-Be-Done Profile
Claire and other users of the new Talent Plus+ platform will require the following:
Employee Dashboard
The ability to:
See order interviews in all stages of the life cycle - my interviews, scheduled interviews, interviews that still need to be scheduled, and all interviews. Filters and selective information will need to be displayed for each.
Set your schedule - including a general schedule and a schedule specific to certain days.
See a given interview’s order information, general information, and work history (including comments).
Schedule, reschedule, and cancel interviews.
Manually move interview tasks through the various queues.
Interview Engine
The ability to:
Quickly annotate (with typed comments as well as static mark types) the candidate’s responses as I walk them through the predefined questions.
Edit both these annotations and the resulting scores.
Preview the candidate’s results.
Submit a recommendation with a written explanation.
Transcription Tool
The ability to:
Manipulate audio controls to playback the interview recording.
Rearrange speech-to-text portions to make sure the audio is ascribed to the right speaker.
View which portions of the transcription are coded as less likely to be accurate.
Quickly cycle through those portions to correct them more quickly.
Mark speech-to-text portions with predefined custom tags and toggle the tag view on and off.
User Flow
Visual Design: Creating a Starter Design System
Because the Talent+ UI was so outdated and broke most gestalt principles and accessibility rules, it was easier to start from scratch. It was also important to me to build a design system intentionally because I knew I would only be engaged with Talent+ temporarily and I wanted their next designer to have building blocks to work from or at least context for elements they might want to change.
There were a few areas that I did not have choice over due to branding that was already set in stone. For example, I would not have chosen the yellow and brown colors due to them being hard to work with accessibility-wise and are generally unflattering when paired with the brighter colors.
Click Below to View the Design System in Figma!
Wireframes
Employee Dashboard
Translation Tool
Note: I did not create lo-fi mockups for the interview engine because the request was on-the-fly .
We conducted A/B testing with 6 participants to evaluate the efficiency of the two components. In this case we already had old recordings of interviews to use as a scenario and we had the old system and new in different qa environments. And to be clear, this wasn’t done in a single go - I would finish a section, dev would build it, and then we could introduce it to a tester.
Employee Dashboard
Measuring Time Saved in Interview Scheduling Stage
Still in Dev! :(
Transcription Tool
Measuring Time Saved in Audio Correction Tasks
Old Workflow: 130:40 minutes
New Unified Platform: 20:14 minutes
Difference: 110.26 minutes; 6.5x more efficient.
Interview Engine
Measuring Time Saved in Post-Interview Candidate Scoring
Old Workflow: 57:55 minutes
New Unified Platform: 38:04 minutes
Difference: 110.26 minutes; 1.5x more efficient.
Qualitative Feedback
Favorite addition across the board was the in-platform scheduling feature.
Product loved the manipulation of branding color guidelines to be more accessible/visually pleasing while not technically breaking the rules.
Dev liked the consolidated element library and guidelines.
Iteration and Conclusion
Although the results of our A/B testing are very promising, we need to consider the small sample size and the fact that some enhancements were not under my responsibility. For instance, the significant difference in transcription tool results is partly due to the new deepgram functionality, which replaces the manual typing process previously used by the transcriptionists. Nevertheless, our testing shows a definite positive trend across the product.
Talent+ has been the most enlightening project I've had the pleasure of working on. It provided me with new opportunities to innovate a more significant part of the solution, experiment with designing various types of functions that were previously unfamiliar to me, and most importantly, collaborate more closely with all tech teams and users. As an introvert, the most challenging aspect of this project was not the design but the social aspect of teaching the teams about the role of UX. Since many team members were new to working with a UX designer, part of my responsibility was to coach them on design thinking and the art of providing effective feedback.