Connecting the Dots
With a limited budget, and a diverse student population, how do you make connections? How do you know which students to reach out to?
This is the problem I found, pitched, and tried to solve, while working as a Data Coordinator at Stephens College. Each new semester we had a meeting that pointed to a disconnect between student population and data. This problem would appear in conversation, on a list of needs, bookmarked as action items.
We were not able to access and utilize the data on student populations that we had. With a restricted budget the department had to choose which student populations to reach out to, and I could see that we had useful trends buried in our archive of data. I wanted to find a way to help, to strategize a solution to help the team with this problem.
Questions: Starting Point:
· Which students do we focus in on? For which of our programs?
· At which events (School Visits? Fairs? in person? virtual?)?
· In which geographical areas?
· Which student populations (Inquiries? Previous unfinished apps?)?
· What’s the most cost effective way to find and utilize student populations?
First I split that question into two parts:
· Find (where are they coming from)
We have a complicated Admissions Application system. Built out with custom applications, lots of data, with no easy way to connect it all to get the solutions we needed to pinpoint these key points. We need to see the trends of where students are coming from geographically in our data.
· Utilize (what data we already have).
In the past we’ve relied on previous student trends, program stats, and Department memory. There was no future-forward process in place to know where students are coming from, a process to log it to be searchable in the present, or even to ensure future capture. We need a way to access what we have from past logs, and what trends are coming in each new semester. And we need a process in place to tag and capture this moving forward.
More Questions: Stakeholder Interviews
I met with the Director of Admissions (my boss), VP of Enrollment (her boss), specific program Department Directors, and student Recruiters.
Key findings:
· Recruiters and Directors have a limited budget to reach out to students. (Reality budgets are in place, there wasn’t room to create a new system, we needed to be cost effective and utilize what we have).
· There is low enrollment for specific programs (HIA & MED). Strategic focus on who to target within those programs is a major goal. It would help to know where previous students came from, where successful pipelines already exist, and which demographics are engaging (to lean in and build out contacts where they are most needed).
· New programs (MHIM) and existing connected department programs (HIA) have an online student base. Students can and are coming from a variety of places and populations (returning students vs new students, local populations vs. national).
· Also we have Various Student Pipelines (New Inquiries via Event Contacts, Referrals, Local Sponsorships, and more).
Sub-Problem:
For each new App there was no way of knowing where they came from if the students don’t directly report it (add it to their application). The fields existed for them to add-in this data, but often they did not. If that information wasn’t logged, we couldn’t search it via our system.
We have a department memory of trends and student contacts (through recruiters, Dept Directors). But our process was dependent on that knowledge base to make decisions on target student trends moving forward. Adding to that setup, the new digital platform was an amazing and complicated customized asset, not built out for easy access to the data that we had in system. Strategic decisions were made via enrollment numbers, logs, but we are not able to utilize the incoming admissions data (these additional fields) that we did have in place.
How Might We … solve this?
· Hire someone new to build a custom database for our data? (Using current and past recorded data, student trends logged accessible) (See: Limited Budget)
· Buy an expensive custom system add-on, customization? (Less expensive, more access to database existing data to see trends)(See: Limited Budget)
· Force students to add that data in, as a application requirement. Create more friction for them in the app process. (Works with Budget, but potential to hurt us on student app melt).(Director said no to adding more friction to app process).
· Use the data we already have. Find a way to connect it without adding in additional steps for students (On budget, no additional friction, but how?)
Our System:
TargetX, built on a Salesforce platform.
Custom Apps, custom applications for each program, fields and setup established by program directors.
New (4 years old) integration. All previous system data was transferred, scanned, and fed into a series of logs, Recruiter notes, and misc via batch process. Recent transition to digital applications, digital first, and digital logs.
My Role:
As a Document and Data Coordinator I worked with all Graduate and Online program incoming student apps, returning student apps, and legacy apps. All of that data came through my weekly and daily ques, where I logged and tracked all required student documentation, Resumes, Statement of Purpose, Reference Letters, and more.
I saw what data we received from students, and noted the required data. Given those requirements I found that we collect:
· Resume (required for all programs)
· Statement of Purpose (required for MED, MFA, MHIM)
· Student Demographics: Gender (optional), Race (optional)
· Location: Current Mailing address (required for all programs)
· Identify verification: (required later in process, outside of admissions dept.)
· Current Mailing Address (required for all programs)
I used that required data as the starting point. What data are students already giving us? What do they have to provide? (No additional friction).
· Past and Current : Work History
· Current Location
In-App there are fields for both of these criteria, to be tagged and searchable, but students rarely fill them in. Part of my job was to scan Resumes (ensure that they are in place and legible), and it was a simple step to cut and paste that work and location data into the correct field. So I started doing it. Logging work history, work location, and current location, and gender (if known) into the form fields for all current students. By plugging this data in, it made it searchable in system.
Then I backtracked. I found the data (if existing) and plugged it into our fields, semester by semester, for every student app that we had for the HIA program. I started with that program as a test. It had low enrollment, and we were currently strategizing on which students to reach out to within that population. The data process was fairly simple, plugging in what we alreayd had to searchable fields that we had in place (but were not using). I also built the process into my work routine moving forward. For each new App I added that data into place as I processed it.
Prototype:
After I got through the first full year of semesters (backtracking), I created a copy of the report that recruiters used to gauge enrollment numbers. I wanted to use something familiar to make it easy to understand (mental models in place) but also mod the criteria to fit the setup we needed to focus the data. Three prototypes later, after tweaking the fields, checking to see if the data set would pull from all the programs.. it did! Success!
Through trial and error I found that I could set the criteria to pull from all students within a program. And you could easily toggle between category to see trends (gender, address and area, work title, etc). All of this using the data we already had, in a familiar format, searchable via the system we were using for our work everyday.
I dug in, using any free moment that I had to backtrack further. I also created a similar dataset for the MED program (the other high priority program). It took me 6 months, but I made it through 4 years of student data on both the HIA and MED programs! Once all known data was logged, and the files were setup, I presented them to my manager, who shared them with our VP.
Result: Access to student demographics and trends
The HIA and MED data sets were used by the Admission department to identify key areas for strategic partnerships. Program recruiters and Department directors were use the dataset to toggle student criteria and see in real time the past, present, and future of our population trends.
In two critical low enrollment programs (HIA, MED) our department was able to focus in on specific regions to target (utilizing budget strategically), and to identify new opportunities in student demographics. The logs were searchable, and able to be utilized by anyone with the link within the systems we already used. Since each data set was made of information that we already required, no additional friction was created for students. They provided their application, and we were able to utilize that information to identify key areas and interests to focus our admissions strategy.
Utilizing the the data set:
Burrell Behavioral Health Partnership
The MED report was specifically used by Admissions to identify and strategically create a partnership with Burrell Behavioral Health, a local counseling organization. That partnership made a valuable new student pipeline which increased revenue for the college. Revenue through increased MED program enrollment, cross-promotions, and a valuable MED specific application funnel.
