The Hidden Costs of Remote Work

Remote work research has largely focused on knowledge workers and software engineers. Less studied: rural workers, caregivers, people holding multiple jobs, and anyone working in a context that doesn't fit the prototype. The study was designed with that gap in mind.

• What challenges do remote workers face with boundary and time management —and how do they feel about them?
• What solutions are they already using, and how well do those solutions actually work?
• Where might technology reduce friction without adding to the burden?

The study's data collection—a 14-day diary study with 35 participants, followed by semi-structured interviews with 12 of them—was completed before I joined the project. I did some interview transcription, though most had been completed by the time I arrived.

Everything after that was mine: qualitative coding, thematic analysis, synthesis, and writing the thesis. I worked independently throughout, with faculty supervision from Dr. Jason Wiese.

Working with the data

The dataset I inherited was substantial: 35 diary participants (25 of whom completed 10 or more days), 12 interviews, and partial transcription. My first task was completing the transcription and getting everything into ATLAS.ti for coding.

I did a single initial pass through all transcripts, generating codes as I went. Early on, I discussed a sample of my coding with a second researcher to align on methodology. After completing the first pass, I did a second sweep to merge, clarify, and tighten codes—the process of living inside data for weeks tends to shift your vocabulary, and consistency matters for defensible analysis.

The initial coding produced 987 codes across 67 code groups.

Thematic analysis

With coding complete, I exported everything to Miro and began clustering. The goal was to surface themes that represented patterns across participants — not just one person's experience repeated. A theme that five people brushed against matters more than an experience one person described in vivid detail.

Three rounds of analysis:

• Round 1: 987 codes → 121 themes (168 codes excluded—48 not specific to remote work, 120 tool references without context)
• Round 2: 121 → 29 themes (merging and collapsing redundancies)
• Round 3: 29 → 6 high-level themes

Each round involved review with a second researcher before moving forward.

What the analysis surfaced

The six themes aren't a tidy list of problems and solutions. They're a structured account of where remote work shifts burden—and onto whom.

1. Collapsed personal boundaries—Personal-life interruptions during work hours split attention and pushed workers to compensate by working late. For many participants, the boundary didn't blur so much as cease to function.
2. Flexibility as a double-edged sword—Schedule autonomy felt like a benefit until workers had to accommodate everyone else's flexibility too, often extending their own hours in both directions. Participants holding multiple jobs simultaneously had it hardest.
3. Invisible labor creep—Remote work added implicit demands that don't show up in a job description: being visibly productive, over-communicating progress to skeptical managers, sourcing equipment and support that would simply exist in an office.
4. Uneven access and equity—Remote work genuinely opened doors for rural workers and caregivers who couldn't feasibly work in person. But in-person workers retained structural advantages—hallway conversations, spontaneous visibility—that remote workers had to consciously engineer.
5. Disrupted communication norms—Availability cues broke down. Status indicators were widely reported as unreliable. Workers were unsure when to call, when to message, when to wait, and the ambiguity created real friction around asking for and offering help.
6. Training gap—Remote work requires distinct skills and management approaches. Most organizations hadn't built the infrastructure to support either. Several participants described having to advocate strongly — sometimes threatening to quit—just to stay remote after pandemic-era policies expired.

Design implications

The analysis didn't stop at description. The thesis proposed technology directions grounded in what participants said they actually wanted from automated support: better boundary coordination tools, adaptive planning with explicit breaks built in, AI-assisted administrative task handling, and better status indicators—all framed around reducing friction without adding surveillance.

The thesis was approved and defended in May 2023. The six themes provide a structured account of where remote workers bear hidden costs—specific enough to inform design, grounded in a sample that was deliberately diverse across geography, occupation, family context, and remote work experience. The purposive sampling decision to include 13 of 35 participants from rural communities produced findings that extend beyond the software engineer contexts that dominate this literature.

The most important methodological lesson from this project wasn't about coding or analysis—it was about what it means to take over someone else's data. The study design and interview script weren't mine. The sampling decisions weren't mine. I had to understand them well enough to analyze honestly without overstating what the data could support.

That constraint made me more careful about claims. Every recommendation in the thesis is explicitly tethered to a finding; the implications section names what would need to be true for each technology proposal to work. I'd rather be specific and limited than broad and undefended.

If I were approaching this differently: I'd want to understand the advisor-side workflow more deeply before drawing design conclusions. The study captured the remote worker experience well, but the organizational and managerial side of boundary enforcement was underrepresented. Some of the most important design levers—company policy, manager training, notification architecture—sit on that side of the relationship.