I decided to actually pay attention to my screen time. Not glance at the Sunday report and wince. Pay attention. For 30 days, I logged every session — which app, how long, what triggered it, how it ended. I wanted to run a proper screen time experiment on myself.
The assumption going in: I'd find one or two problem apps eating my hours. I'd cut them, feel virtuous, and move on. That's the narrative we all carry — there's a villain app, and if you just deal with it, the problem shrinks.
It didn't work that way. Not because I didn't find problems. Because the problems weren't where I expected.
The first week: the numbers aren't the story
My total daily average came in right around where I expected — roughly four and a half hours, give or take. Not shocking. The weekly summary would have told me the same thing, and I would have sighed and moved on.
What was shocking was the shape of the usage. When I looked at individual sessions instead of daily totals, a pattern emerged immediately. Most sessions were under two minutes. A notification glance, a quick check, a reflexive open-and-close that barely registered as "using my phone." But scattered among those tiny interactions were a handful of sessions — two or three a day — that ran 20, 30, sometimes 40 minutes. And those few long sessions accounted for most of the time I actually regretted.
This isn't unusual. Session durations tend to follow a power-law distribution: many very short interactions, a few very long ones. The average flattens this shape into a single number that describes neither the quick checks nor the long drifts accurately.
The total isn't the problem. The distribution is. "I spend four hours on my phone" means something completely different if it's 120 two-minute checks versus 8 thirty-minute drifts. The first pattern might be fine — fragmented, maybe, but largely intentional. The second one usually isn't. Once I could see the shape, the weekly summary felt almost useless. It was answering the wrong question.
The app I blamed wasn't the problem
Going in, I had a suspect. The social media app I felt worst about — the one I would have named if anyone asked "what's your phone problem?" But the data told a different story. The app I worried about most, I actually used with relative intention. I opened it, I scrolled for a bit, I closed it. Sessions were moderate and fairly consistent. Not ideal, but not the problem I'd imagined.
The real time sink was an app I barely thought about. A news reader I'd have called "productive" or at least neutral. Because I never thought of it as a problem, I never noticed when sessions stretched long. There was no guilt to trigger awareness. It just quietly accumulated minutes while I told myself I was staying informed.
This tracks with what researchers have found. Andrews et al. (2015) tracked 23 college students for two weeks, measuring their actual phone use against their self-reported estimates. The finding was stark: self-reported use didn't correlate with measured use. We are, in a precise sense, unreliable narrators of our own phone habits. Separately, Parry et al. (2021), as discussed in Anderl et al. (2024), found that self-reported screen time shares only about 20% of its variance with objectively measured use — and the errors go in both directions. Some people think they use their phone far more than they do. Others far less.
The lesson is uncomfortable: the guilt-driven narrative — I'm addicted to this one app — often misdirects attention. The real patterns are quieter and less dramatic than the ones that make you feel bad. You can't trust your gut about which app is the problem. You have to look.
The triggers were boring
The most surprising finding had nothing to do with apps at all. Almost every long session started from the same two or three triggers. Not emotional crises. Not boredom spirals. Just transitions — the moment between finishing one task and starting another. Waiting for the coffee to brew. Putting off an email I didn't want to write. The tiny gap between intentions where the phone fills the silence.
I'd expected to find dramatic patterns — stress-driven scrolling, loneliness binges, late-night spirals. Those existed, occasionally. But the ordinary triggers outnumbered them ten to one. The phone wasn't a coping mechanism. It was a default. The thing my hand reached for when my brain hadn't yet decided what to do next.
This reframes the strategy entirely. Blocking one app doesn't help if the trigger just redirects to another. I tested this accidentally — I removed one app from my home screen for a few days, and the time simply migrated to two others. The problem isn't the app. It's the unstructured moment. The app is just whatever's closest to your thumb.
Week three: the watching changed the behavior
By week three, my usage had dropped noticeably — maybe 20 to 25 percent. Not because I'd set rules or installed a blocker. Because I was watching. The act of paying attention made the automatic opens feel conspicuous. I'd reach for my phone, remember I'd be logging it, and sometimes just... not. The observer effect, applied to myself.
This is a well-documented phenomenon. When people know they're being observed — even by themselves — behavior shifts. The mere act of measurement changes the thing being measured. And for about two weeks, it worked remarkably well.
But the drop wasn't even. The short, reflexive pickups — the under-two-minute checks — declined the most. Those are the sessions driven by pure automaticity, and even a thin layer of self-consciousness was enough to interrupt them. I reached for my phone before I had a thought about why, but knowing I'd have to write it down gave me just enough pause.
The long drift sessions barely changed. The ones where I opened something intentionally and then lost track of time — 15 minutes became 25, became 40 — those kept happening. I knew they were happening. I could see them in the data every evening. But knowing after the fact didn't help me in the moment when I was fifteen minutes deep and the session still felt like five.
This is the gap. Self-tracking gives you hindsight. It teaches you your patterns in retrospect. But the pattern you most need to interrupt — the session that's drifting right now — requires awareness during, not after. A log you review at night can't tap you on the shoulder at minute fifteen.
This is where detection as a real-time layer becomes something more than a concept — it's the answer to the specific failure I kept experiencing. Tracking told me what happened yesterday. What I needed was something that could notice what was happening now.
What thirty days actually taught me
Here's what I'd tell someone starting their own screen time experiment.
Track sessions, not totals. The daily number is nearly meaningless. The per-app, per-session shape tells you everything. Two people with identical daily averages can have completely different relationships with their phone. One is checking quickly and moving on. The other is drifting for long stretches without realizing it. The weekly report treats them as the same. They're not.
Watch the triggers, not the apps. The app you blame is rarely the whole story. The moment you reach for your phone — what was happening right before — tells you more than which icon you tapped. If you only track apps, you'll play whack-a-mole forever, chasing usage from one icon to the next. If you track the moments — the transitions, the gaps, the small avoidances — you start to see the actual pattern underneath.
Awareness fades. The observer effect is real but temporary. By week four, I was paying less attention to the log. The automatic habits were creeping back. Sustained awareness can't be maintained by willpower alone — it needs to be built into the system, not carried in your head. That's the thing I kept circling back to — and why we built Threshold to do what this experiment does, but continuously: learn per-app patterns on-device, signal when a session drifts past your baseline, and let you choose the level of intervention per app. Not a manual log you'll abandon. A quiet system that watches for the drift.
For practical strategies on building friction into the right moments, we've written a guide to reducing screen time on iPhone that pairs well with what I learned here.