30 Days Without Social Media: What Actually Changed

A documented experiment in intentional disconnection.

The Hypothesis

Removing social media (Twitter/X, Instagram, LinkedIn) for 30 days would:

1. Improve sustained focus and deep work capacity
2. Reduce baseline anxiety
3. Make me better informed (counterintuitive, but deliberate slow-news could beat the feed)

Methodology

Deleted apps from phone. Blocked domains on desktop browser using a hosts file. Kept a daily log of mood (1–10), focus quality (1–10), and notable observations.

Not a controlled experiment. Sample size of one. Take accordingly.

Week 1: Withdrawal Is Real

Days 1–4 were noticeably uncomfortable. The urge to check was habitual, not informational — triggered by transitions (waiting for coffee, finishing a task, sitting on the toilet). The behaviour wasn't about wanting information; it was a conditioned response to low-level discomfort.

By day 5 the urge frequency had dropped significantly. This was the clearest signal that the behaviour was more compulsive than I'd wanted to admit.

Focus scores: inconsistent. Anxiety: slightly elevated (adjustment).

Week 2: The Boredom Dividend

Something unexpected: genuine boredom returned. Not the anxious, distracted not-quite-boredom of doom-scrolling, but actual boredom — the kind that precedes creative thinking.

By day 12, ideas were arriving differently. Less reactive ("I should respond to this thing I saw"), more generative. Hard to quantify, easy to notice.

Focus scores: improving. Anxiety: noticeably lower.

Week 3: Recalibration

The information gap was smaller than expected. I was reading more slowly and more deliberately — one long piece in the morning instead of fifteen fragments throughout the day. My understanding of the topics I cared about was, if anything, better.

What I genuinely missed: specific people's thinking. The medium and the people were bundled together, and I'd removed both.

Focus scores: consistently high. Anxiety: baseline lower than pre-experiment.

Week 4: The Returns Slow

The dramatic improvements of weeks 2 and 3 plateaued. Week 4 was just... normal. Which might be the point — a better normal.

Focus scores: stable. Anxiety: stable improvement.

Results vs. Hypothesis

| Hypothesis | Result | |---|---| | Improved focus | Confirmed — significantly | | Reduced anxiety | Confirmed — moderately | | Better informed | Partially confirmed — topic depth improved, breadth reduced |

What I Brought Back

Reinstalled LinkedIn (professional necessity), left off Instagram, replaced Twitter with an RSS reader curated to specific writers I'd actually been missing.

The experiment outcome wasn't "social media bad." It was: the default configuration of these tools is not designed with my interests in mind, and using them by default means accepting their defaults.

I'd rather set my own.