New Credentials, Not New Courses: A Report on a Year-Long Pilot of Embedded Industry Credentials

Last spring, a recent graduate named David emailed me. David had worked in project management for more than a decade, but he had never believed the PMP, the profession's flagship credential, was within his reach. In his final semester, he took an industry certification exam that we embedded in his project management course—a proctored, credential-bearing exam, folded into a class he was already taking. That exam, he wrote, gave him the confidence to sit for the CAPM in January. On March 19, he passed the PMP. Ten years of experience, and the chain that finally converted it into recognition began with an exam he took because it was available inside his course, at no additional cost, with no separate program to enroll in. Today, David teaches project management for my department.

That email is the shortest argument I know for the work my team did this past year. Here is the longer version.

The problem we were trying to solve

Employers increasingly screen for demonstrated skills, and a degree alone does not always carry that signal. Most institutions recognize the gap and respond by building something new—a certification bootcamp, a micro-credential program, a badge academy. Those efforts tend to run into the same two walls. Students working toward a degree rarely have the time or money for a separate certification track, so the new offering competes with coursework rather than complementing it. And building credential-bearing programs from scratch requires curriculum approval, staffing, and budget cycles measured in semesters.

There is a third obstacle, quieter and more structural than the other two. At most comprehensive institutions, the vendor relationships live in continuing education while the students live in academic affairs. This is not an accident of the org chart; it is the organizational expression of a long collision between an industry logic and an academic logic in public higher education (Gumport, 2019), encoded most durably in the boundary between credit and non-credit programming. The two sides of the house rarely build together, and the opportunity sits in that seam.

The model

We chose not to build anything new. In Fall 2025, Continuing & Professional Education partnered with more than ten faculty members to embed industry certifications directly into existing courses: Generative AI Foundations in our introductory computing course, PMI Project Management Ready® in our project management course, and Intuit Personal Finance in a business section.

The design principle was a strict division of labor. The CE unit acted as broker—holding the vendor contracts, funding and distributing exam access, running proctored testing through our Certiport-authorized testing center, and absorbing the administrative work a certification adds. Faculty kept full ownership of their courses; their role was to map the credential objectives to what they already taught and to frame the value for their students. No course was rebuilt. No curriculum approval was triggered. No additional tuition or fees reached a student.

The credential rides on the course. The course never changes hands.

What happened

Across two semesters, 536 students were given access to a credential, we administered 424 certification exams, and students earned approximately 380 industry credentials—for many of them, a first professional certification.

The second semester produced stronger results than the first, and that improvement matters more than either semester alone. First-time pass rates rose from 88.8% to 90.6%; the final Spring pass rate, after retakes, reached 93.4%; staff hours per exam declined as our processes matured. This spring, students began posting their Credly badges to LinkedIn on their own initiative, and the chair of the host department publicly announced her cohort's completions. A pilot that improves on its second pass is not a lucky launch. It is a repeatable system.

The finding I did not expect

The most useful observation of the year had nothing to do with pass rates. In Spring, participation across nine sections of the same course ranged from 90.9% down to 58.6%—same credential, same materials, same support, same semester.

The difference was integration behavior. The highest-participation sections connected the credential explicitly to course outcomes, set firm deadlines, and addressed the material in class. The lowest-participation section left the credential self-paced and loosely scheduled, and never explicitly connected it to the course. Face-to-face sections outperformed online sections on participation across the board.

Organizational scholars would recognize the pattern. A program, like a policy, is not self-executing; it is inhabited by the people who carry it out, and the same formal structure produces different outcomes depending on how local actors take it up (Hallett & Ventresca, 2006). Birnbaum (2003) drew the relevant distinction between hard governance—the formal structures and instruments—and the soft governance of norms and informal practice that determines how those instruments are actually used. Our program was a hard instrument, identical across all nine sections. Participation was a soft-governance outcome, produced classroom by classroom. Meyer and Rowan (1977) described organizations that adopt formal structures while decoupling them from operational practice; what we observed was a section-level version of the same phenomenon, where a credential could be formally present in a course and functionally absent from it.

The practical consequence is that participation is not a property of the credential and not a measure of student appetite. It is a property of enactment, and the levers are identifiable: early access to practice materials, explicit syllabus connection, firm deadlines, and in-class touchpoints. We adjusted what we could between semesters—practice-platform access moved from late in the term to mid-semester in response to faculty feedback—and the results moved with us.

What did not survive, and why

One of our three credentials did not continue into Spring, and the reason is worth stating plainly. The Intuit Personal Finance pilot ran in a single section; the host instructor asked to continue it and expand it into additional sections; the credential ended anyway, because no funding existed to renew the seat licenses.

I include this because it was the most instructive event of the year. Faculty demand does not sustain a credential, and neither does student interest. A funding model sustains a credential, and in its absence, demand does not scale—it disappears. Any institution considering this model should decide who pays before the first exam is proctored, not after the pilot succeeds.

The last mile

One student comment from our feedback has stayed with me: "I don't know what to do with my certification. Do I put it on my resume and LinkedIn or not?"

We had built the full pipeline to credential attainment and left the final step—converting attainment into a career signal—to chance. Some students figured it out on their own, and their badges are on LinkedIn now. Others earned something valuable without knowing how to use it. The remedy is inexpensive: a brief claim-your-credential step, built with career services and delivered at the moment of passing. If you build an embedded credential program, build the last mile into it from the start.

What I would tell a colleague building this

Start where enrollment already lives. Choose two or three high-enrollment courses whose learning outcomes already overlap substantially with a recognized certification; fit matters more than novelty. Let your CE unit hold every vendor contract, so that faculty never touch a vendor portal. Recruit faculty by subtraction rather than addition—their course stays theirs, and your unit absorbs the administration. Fund the first year deliberately and centrally, so that the student's answer can be yes at no cost to them. And measure everything from the first exam date, because the numbers are what turn a pilot into a program.

A note on what this is and what it is not. These are program results, reported by the administrator who ran the program—practice, not research. I am a staff member describing what one model produced at one regional public university across one academic year, in the hope that the account is useful to colleagues building similar programs. It is not a controlled study, and it does not claim to be one.

What I can say is this: the model works, it improves with iteration, and its failure modes are knowable and preventable. And somewhere in your institution right now is a student—perhaps one with a decade of experience and nothing on paper to show for it—who would take the exam if it were simply available, and who might, a few months later, send you the best email you receive all year.