Project Background
After my acceptance into Georgia Tech's Undergraduate Consulting Club (UCC) the prior semester and the successful completion of a pro-bono project with Advocate (which you can read about here), I decided to take on greater responsibility and apply to become a team lead for a project in Fall 2020.
After my application was accepted, I was assigned to lead a team of seven students on a project with the talent management team within UPS Global Customer Solutions (GCS), a department that recruits logistics experts and consultants. Our principal client contact would be Art Moreno, a GCS talent manager.
In my initial one-one-one meeting with Art, we clarified his precise needs, outlined his ideal deliverable, generated a project scope and timeline, and established a system of direct and consistent communication — something I especially wanted to prioritize as team lead, as I felt client communication was lacking in my first project with Advocate. To address this, we arranged a date and time each week to meet so I could update him on our progress and clarify any questions or obstacles our team encountered over the course of the week.
Project Goals
Art's overall goal was to identify top candidates for GCS internships in a manner that was both effective and efficient. To do so, he tasked our team with:
Compiling research (i.e. making the system more effective)
What kinds of students would be the best fit for each of the GCS internships? What majors should these students have? Which schools best prepare students with these majors? Do applicants need prior experience? If so, what kind?
How should GCS measure intern success? What specific returns should GCS expect from interns, and which KPIs should GCS measure to evaluate them?
How can UPS not only meet but exceed industry standards for internships? How can GCS provide maximum value to its interns, and how can it attract the best applicants?
Building talent management dashboards (i.e. making the system more efficient). This required our team to consider the following in our design:
How should we quantify, standardize, and display intern performance?
What statistical analyses can we conduct to better predict an applicant's aptitude for a given role?
How can we simultaneously provide our client with a high-level overview of optimal candidates while also allowing them to examine individuals more closely?
With a better understanding of the project in tow, I scheduled an introductory meeting with my team so we could get to know each other, so I could familiarize them with the project and with my expectations, and so that we could together establish regular meeting times and a workflow that would allow us to be as productive as possible. We also delegated work for each of Art's research questions, with each person taking responsibility for 1-2 topics.
My Leadership Style
From the beginning, my priorities as team lead were structure, consistency, and communication.
Structure: Together, my team and I devised an overall project structure, segmenting our semester into three phases: 1) Research, 2) Data analysis, and 3) Dashboard construction. We also determined a week-to-week structure, establishing 90-minute blocks on Mondays and Wednesdays to devote to the project.
Consistency: Every Sunday, my team could expect a lengthy message from me in our team GroupMe (examples below), outlining the project status, our goals for the week, and our initial plan for our weekly meeting with Art. I found such a system helped to keep everyone on the same page, with a better understanding of what they needed to focus on that week.
Communication: I made a point to be as transparent as possible with my team, involving them in high-level decision-making and welcoming their input in our weekly meetings with Art. This included being honest with them if we ever got behind, and being clear about the expectations we had for each other. I found that leading with honesty and transparency helped my teammates feel comfortable communicating with each other in the same way, with the mutual understanding that we were all accountable to each other.
Analysis: Visualizing our Deliverables
After devoting the first three weeks of the project to research and compiling a 13-page summary of our findings for Art, we advanced to the analysis phase of the project, which involved exploring the data Art provided us and devising a plan for how we might use it in the dashboards.
Art sent us a series of Excel workbooks containing a host of information on individual (anonymized) applicants — including hometown, college, major, year, etc — as well as their application status — whether they were denied, admitted but declined, or accepted. Most importantly for us, however, was a workbook containing data on post-internship performance reviews consisting of both quantitative and qualitative manager feedback, which we could use in our predictive analysis of who GCS could expect to perform the best in each role. Combining this post-internship data with the scores the GCS talent management team gave to the applicants in their pre-internship recruiting phase, we could also evaluate how accurately recruiters currently evaluate the aptitude of applicants.
In short, we had a lot to work with, so we aimed to initiate the dashboard construction phase as quickly as we could.
Keeping in mind our goal of clear client communication, we arranged a meeting with Art to discuss his expectations for our deliverable, and to ensure that his expectations were in line with what we could produce given our time and resources (understanding his expectations would also help us exceed them!). Along with Art, we identified six principal questions that the dashboards should answer and assigned each team member a question. They were:
Who is turning down internship offers, and how is this changing over time?
Who is receiving return offers? Who is not receiving return offers? How is this changing over time?
Are current assessments/interviews good measures of intern aptitude?
Who are the best performers?
How does seniority play into performance?
Which disciplines perform the best in each role?
With our deliverable clearly defined, we were ready to advance to the dashboard construction phase.
Dashboard Construction
We constructed a story in Tableau consisting of seven dashboards — one for each of the questions above and one allowing the user to view individual applicants more closely.
The story was designed to provide a high-level understanding of each phase of the intern lifecycle — first, to summarize the types of students most frequently receiving offers, and who among them is most often declining their offer; second, to detail the types of students who perform best in each role (What school are they from? What do they study? How old are they?); third, to show the types of students who are most often receiving return offers; and fourth, to evaluate how effective GCS's current tools are at predicting intern success.
To evaluate GCS's current recruitment system, we constructed a Tableau dashboard to visualize any correlation existing between the "rank" assigned to applicants following their interview and online assessment, and their overall performance review score (reported as a numerical value between 0 and 5). We ultimately found that GCS talent management was very good at evaluating intern performance at the extremes — the best of the best and the worst of the worst — but that the numbers became a bit jumbled in the center. To correct this, we proposed introducing an additional recruitment round, through which these middle 50% would be filtered in an attempt to provide more structure and clarity to this block.
We also made a point to quantify the qualitative features of performance reviews. To accomplish this, we classified questions on the manager performance reviews into buckets representing different characteristics — ownership, teamwork, flexibility, positivity, and receptiveness —and used the reviews to give each intern a score for each. Using this metric, recruiters could see which traits would be most valuable for which positions, and determine which traits the highest-performing interns exemplify, informing them of what to look out for when interviewing candidates.
Outcome and Reflection
In the end, Art and his team were impressed with our project and even invited us to continue working with them through January. Because we had become so familiar with his data management practices, he wanted our insight into how to further standardize and centralize his data, and to help him transition into using Microsoft Access for database management.
While we were excited by the opportunity, we thought it best to have Art introduce it as a UCC pro-bono project the following semester, as the scope and timeline sounded more appropriate as a semester-long project. That way, the members of our team who were interested in continuing could do so, while others could pursue other projects if they wished.
On the scale of my undergraduate career, the UPS talent management project represented one of my early forays into leadership and project management. Though I was initially nervous for the added responsibility, I quickly found my footing and carved out my unique leadership approach. I developed positive, professional relationships with my teammates and my client contact, and I refined the Tableau skills that would proceed to help me land my first internship.
While it was easily one of my toughest semesters — especially in the context of the pandemic —it was ultimately one of the most rewarding. This two-sided coin —challenging yet rewarding — would come to be a common trend over the next year.
See below for the presentation we gave to Art and his team at the end of the semester.
