01 / 12
Executive Summary
People can be surrounded by glucose, medication, food, activity, sleep, and wearable data—yet still face a harder question: what matters now?
The opportunity may not be another stream of information, but a clearer path from data to context, understanding, action, and care.
This one-day project challenges the assumption that consolidation alone is the answer. It uses secondary research to define questions, expose assumptions, and shape the smallest prototype capable of testing where the real breakdown occurs. The expected outcome is not a finished healthcare product; it is a credible direction for primary research.
02 / 12
Why this matters
Type 2 diabetes management can involve monitoring glucose, medicines, food, physical activity, and health checks—not one isolated task.[2][3][6]
Connected tools can add continuous streams, reports, and alerts. ADA guidance treats diabetes technology as part of individualized care and emphasizes education, training, and ongoing evaluation—not technology in isolation.[4]
The design question is therefore not simply whether information exists. It is whether a person can interpret it in daily context, decide what deserves attention, and communicate what is useful to their care team.
A small, single-site mixed-methods study described diabetes self-management as cognitive, logistical, and emotional “work” integrated into daily life.[7]
Useful direction; not a population-wide conclusion.
Interviews with 20 people with diabetes identified extra tasks introduced by digital tools, including managing devices and their data.[8]
Small mixed Type 1 / Type 2 sample; hypothesis-generating only.
A systematic review found mixed evidence that patient-portal features alone improve diabetes outcomes.[9]
Published in 2020; a baseline, not the current state of connected care.
What an authoritative source directly supports.
What a cited study observed in its specific sample and setting.
A proposed explanation still requiring primary research.
03 / 12
Initial assumption to challenge
My initial assumption is that fragmented data makes diabetes management harder—and that bringing it together would make the next step clearer.
That is a starting belief, not a finding. Each assumption needs a question capable of overturning it.
People need all diabetes information in one place.
Challenge: When did scattered information last prevent someone from understanding or acting—and when did consolidation add noise?
Data fragmentation is the main problem.
Challenge: Is the breakdown access, interpretation, confidence, action, or communication?
People need more reminders.
Challenge: When a routine breaks down, was the cause forgetting—or timing, access, environment, technique, or competing priorities?
AI-generated explanations would help.
Challenge: What explanation is useful and trusted, and what uncertainty must remain visible?
A physical device, adhesive, or new placement would improve the experience.
Challenge: Is the recurring problem caused by the object, body location, environment, interaction, behavior, or timing?
More connected data produces better decisions.
Challenge: Which data changed a past decision, which was ignored, and what context was still missing?
04 / 12
Target users
Primary users
Adults living with Type 2 diabetes
The focus is context, not an invented demographic: routines spanning medication, glucose monitoring, meals, activity, sleep, work, family, travel, healthcare interactions, and existing technology.
Secondary users
The people around care
Family members and caregivers, clinicians, and diabetes educators—where their role intersects with daily support, shared information, or care conversations.
No persona yet. Without primary research, a named persona would create false specificity. Recruitment should seek variation in routines, treatment contexts, monitoring practices, digital confidence, work, family, and access.
05 / 12
Research questions
Ask for specific past events. Do not ask people to endorse a proposed assistant, device, or feature.
Data
Walk me through the information you collected the last time your glucose changed. What did you use, ignore, or have to find elsewhere?
Interpretation
What did you think the change meant? What context did you look for, and what remained unclear?
Action
What happened next? Where did you hesitate, and what workaround did you use?
Medication
Tell me about the last time your medication routine did not go as intended. What was happening around you?
Physical / placement
When has a monitoring or delivery object been difficult to use? What role did placement, clothing, movement, environment, or timing play?
AI
Think of a health explanation you trusted—or rejected. What made it credible, and what should a system never decide alone?
Care
What did you last need to explain to a clinician? Which details helped the conversation, and which created noise?
06 / 12
Journey to map
Notice
Potential breakdown to investigate—not an observed finding.
Understand
Potential breakdown to investigate—not an observed finding.
Connect
Potential breakdown to investigate—not an observed finding.
Decide
Potential breakdown to investigate—not an observed finding.
Act
Potential breakdown to investigate—not an observed finding.
Communicate
Potential breakdown to investigate—not an observed finding.
If a physical mechanism is involved: trace object, placement, interaction, environment, behavior, and timing before attributing the failure to the device itself.
07 / 12
Opportunity areas to validate
These are parallel territories, not a ranked roadmap. Open each one to see what would support—or eliminate—it.
08 / 12
Potential prototype direction
“Why did my glucose change?”
Prototype one event review—not an ecosystem. Show the observed change, nearby records, evidence considered, possible associations, missing information, and uncertainty. Let the person verify context, choose a safe non-prescriptive next step, and prepare a question for their clinician.
Critical journey
Kill condition: if primary research shows the recurring breakdown is physical placement or another concrete mechanism, change the prototype. Do not force an AI, mobile, or connected-device answer.
09 / 12
Success measures
Understanding
Can someone explain what was observed in their own words?
Context
Can they identify the information surrounding the event?
Interpretation
Can they separate an observation from a possible association?
Cognitive load
Does the journey require less effort than the current workaround?
Action clarity
Can they identify an appropriate, non-prescriptive next step?
Trust
Can they explain why information was shown and where uncertainty remains?
Relevance
Would this information be useful in the situation it is designed for?
Care communication
Does it help form a focused question for a clinician?
For a physical intervention, observe frequency, time, effort, repeated failures, before/after behavior, preference, and environmental impact. Set measures only after baseline research; none are invented here.
10 / 12
Design process
Morning
Observe · Define · Research
Review secondary evidence; document the original observation; prepare questions for real behavior.
Midday
Map · Challenge · Synthesize
Map the journey, challenge assumptions, and separate known evidence from open questions.
Afternoon
Identify opportunities · Select one hypothesis · Prototype
Explore without ranking, then choose the smallest intervention that tests the most important uncertainty.
Late afternoon
Test · Learn
Run a lightweight comprehension test; record what is observed without converting it into clinical evidence.
Before evening
Document
Tell the honest portfolio story: what is known, what remains a hypothesis, and what research comes next.
“Identify the most important problem worth investigating further and create a testable design direction.”
One day can frame and test a direction. It cannot clinically validate a healthcare solution.
11 / 12
Key takeaway
Before designing a new mechanism, understand why the existing system fails.
I began with a belief that fragmented data was the central problem. Secondary research supports the complexity of diabetes self-management and the need for individualized, supported technology use—but it does not prove that consolidation is the right intervention.
The most important next step is primary research around a recent, real glucose event: how a person noticed it, sought context, interpreted uncertainty, acted, and communicated. Only then should the project commit to a timeline, reminder, AI explanation, physical support, or another direction.
12 / 12
Sources
This case study uses secondary research only. No primary interviews, participant observations, measurements, or usability outcomes are claimed.
- 01
American Diabetes Association Professional Practice Committee
Standards of Care in Diabetes—2026
- 02
World Health Organization
Diabetes — Fact sheet
- 03
Centers for Disease Control and Prevention
Managing Diabetes
- 04
American Diabetes Association
Diabetes Technology: Standards of Care in Diabetes—2026
- 05
American Diabetes Association
Facilitating Positive Health Behaviors and Well-being: Standards of Care in Diabetes—2026
- 06
National Institute of Diabetes and Digestive and Kidney Diseases
Managing Diabetes
- 07
Spencer-Bonilla et al. · Mayo Clinic Proceedings: Innovations, Quality & Outcomes (2021)
Patient Work and Treatment Burden in Type 2 Diabetes: A Mixed-Methods Study
- 08
Al Zahidy et al. · Mayo Clinic Proceedings: Digital Health (2024)
Digital Medicine Tools and the Work of Being a Patient
- 09
Alturkistani et al. · Journal of Medical Internet Research (2020)
Patient Portal Functionalities and Patient Outcomes Among Patients With Diabetes: Systematic Review