Students at the University of Washington weren't frustrated with the bookstore. They were frustrated with everything around it. What to buy was scattered across Canvas, MyPlan, course sites, and instructor slides; no single tool connected course registration to textbook acquisition. Students consistently delayed buying until the first week of class, not from disorganization but because the ecosystem gave them no reason to act sooner.
That insight redirected the project. "Improve the bookstore's ordering process" became a challenge about bridging fragmented systems, and produced DubsBooks, a portal that connects a student's registered courses directly to their required materials.
Context
The University Book Store is a student-led non-profit founded in 1900. In 2025 they brought our HCDE capstone team in to modernize the textbook ordering experience, alongside renovations to their physical store and website. We started broad:
How might we improve the textbook ordering process with the University Book Store to give students the best possible experience?
Once the research was in, we narrowed it:
Refined: How might we bridge the gap between course registration and textbook purchasing, when students don't know what they need until the first week of class?
Findings
Four themes emerged consistently across the research.
- 4 research themes
- 25 survey respondents
- 21 intercept interviews
1 · Satisfaction with the bookstore is high
In both surveys and intercept interviews, students were frequently satisfied, crediting the knowledgeable, approachable staff they could turn to whenever they got confused.
Implication: the problem wasn't service quality; students liked the bookstore. The friction was in everything around it: the fragmented set of platforms they had to navigate before ever reaching the store.
2 · Purchase decisions peak in the first week
Most students decided during week one, even though Washington law requires course materials to be listed a month ahead. They took their cues from syllabi, not the bookstore's own listings, so the bookstore's early availability went unnoticed.
3 · Information gaps on both sides
Students often didn't know which books to get; some asked for titles professors never sent to the store. Operationally, contextual inquiry surfaced a compounding bottleneck: roughly 250 orders stacked in the Day-1 morning backlog and about 15 minutes of inventory lag, feeding timing mismatches where students arrived before their orders were ready.
4 · A fragmented experience
The journey spanned disconnected platforms: identify materials on one site, buy on another. About a third of survey respondents used third parties like eBay, independent bookstores, or classmates, fragmenting it further.
The research pointed to four design directions: unify purchase and discovery, improve communication and transparency, support early decision-making, and surface affordable options.
Process
We used four methods over two quarters, starting in the field during the bookstore's busiest week and broadening out from there.
| Method | Scope | Purpose |
|---|---|---|
| Contextual inquiry | 5 observers, 16 hrs | Shadowing staff and observing customers on-location |
| User surveys | n = 25 | Textbook buying behaviors and preferences |
| Stakeholder interviews | n = 4 | Sponsor needs and constraints (CEO, COO, and others) |
| Intercept interviews | n = 21 | Quick at-the-moment feedback during spring peak hours |
A critical detail surfaced in the stakeholder interviews: the bookstore was already complying with the state listing law, but students weren't finding the information. The problem wasn't availability; it was discoverability. As one sponsor framed their goal:
"Here is what to expect. Yes, this is normal. We can help you."
What I owned
On a five-person team I served as UX Research Lead and the primary point of contact with bookstore leadership across both quarters.
| Area | My contribution |
|---|---|
| Research design | Scoped the contextual-inquiry protocol; designed the intercept guide and co-designed the survey instrument. |
| Sponsor coordination | Primary contact with leadership (CEO, COO, Chief of Campus Operations); coordinated site visits and stakeholder interviews. |
| Synthesis | Led the affinity-mapping sessions that turned field notes into the four research themes. |
| Field & design | 2+ hours of shadowing, 21 intercept interviews; contributed to sketching, service blueprints, and usability testing. Design execution was led by our two UX designers. |
From research to design
We built three buyer personas (a physical-textbook buyer, a digital-first student, and a last-minute shopper) and mapped each to a service blueprint tracing their journey from course registration to textbook pickup.
A timed speed-sketching sprint narrowed seven opportunities to two concepts: viewable textbooks during course registration, and an editable textbook portal with push notifications.
The professor-portal pivot
Our low-fidelity designs explored both a student-facing and a professor-facing portal, where instructors would add required textbooks and have updates sync to students. During stakeholder discussions, the bookstore told us a professor-facing portal wasn't feasible within current constraints.
This was a defining moment. We'd invested in a dual-sided system, and half of it wasn't implementable. Rather than treat it as a setback, we pivoted to focus solely on the student experience, and scoping down in response to real constraints became one of the most valuable decisions of the project.
Prototyping and testing
In mid-fidelity, students could import their registered courses via NetID (auto-populating their textbook list) or upload a list manually, with return-deadline reminders on the homepage. Usability testing produced three specific changes: an expandable course dropdown so students see all required and optional books up front; external pricing (Amazon, eBay) moved to a dedicated tab to keep the focus on bookstore offerings while staying transparent; and a redesigned marketplace card that surfaces price, condition, and seller at a glance.
The solution
DubsBooks is a textbook portal that unifies the fragmented journey: one place to get the right materials at the right time.
Smart course integration
Students log in with their existing UW NetID to instantly see required materials for their registered courses, or import courses manually. They can also export the list to bring to the physical store, bridging digital and in-person.
Integrated alerts
Real-time order updates, payment and return reminders, and clear Day One Access information keep students informed from purchase to the first day of class, directly addressing the information gaps and timing mismatches we observed.
Student-to-student marketplace
A peer exchange lets students buy and sell textbooks directly. This was a proactive call: our survey showed about a third of students already used third-party channels, so rather than ignore that behavior, we brought it into a trusted, centralized platform aligned with the bookstore's mission.
Step 1 of 6:A UW NetID login replaces the fragmented path across Canvas, MyPlan, and course sites; students without one can import courses manually.
Returning to the service blueprint, the fragmented website navigation was replaced by a single NetID login that surfaces everything in one place, with much of the bookstore's existing backstage translating directly into DubsBooks' internal system.
Outcome
DubsBooks was presented at the HCDE Capstone Showcase on June 2, 2025, and delivered to the bookstore as a final prototype with supporting research documentation. It was a capstone deliverable, not shipped to production. What the research produced was a reframing: the problem wasn't the bookstore's service (which students rated highly) but the fragmented ecosystem around it. Backed by 16 hours of contextual inquiry, 25 surveys, 4 stakeholder interviews, and 21 intercepts, that reframing gave the bookstore a research-grounded lens for its ongoing website redesign.
Reflection
Two quarters was less time than it sounds. Coordinating five people, sponsors, and shifting research plans compressed the timeline more than I expected. The professor-portal pivot was the lesson that stuck: scoping down isn't failure, it's a design decision. When stakeholders told us half of our dual-sided system wasn't implementable, we didn't lose work; we gained clarity about what could actually be built. Learning to treat that kind of constraint-driven adaptation as progress rather than loss was the most valuable takeaway.
The honest limitations: our samples (n=25 survey, n=21 intercepts) were enough for thematic saturation but limit any quantitative claim, and both skewed toward engaged students who showed up; the intercepts, by design, missed those who'd already abandoned the process. The prototype was tested for directional feedback, not benchmarked against task metrics.