The 20-Minute Focus Ignition
Beat the hardest part of deep work — starting — by force-blocking distractions for 20 minutes
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 3
- Confidence
- 92%
Orosz's consistent problem isn't sustaining focus, it's starting. His reliable fix: set a 20-minute timer, physically block every distracting site so he literally can't reach them, and push through the first grumbling minutes until a mental switch flips around the five-minute mark and he drops into flow.
Origin
Gergely Orosz's personal method, built from a small Python script he wrote to override his machine's hosts file; he notes the same site-blocking can be done via browser extensions for non-technical people.
Core principles
- 01The barrier is starting, not persisting — remove the 'I'll just quickly check Twitter' escape hatch
- 02A hard block beats a soft nudge; apps you can dismiss stop working once your brain learns to work around them
- 03Focus tricks decay over time and need rotating
- 04Roughly five minutes in, resistance flips into flow
How to run it
- 1
Hard-block your distractions
Run a script (or browser extension) that blocks LinkedIn, Twitter, Facebook and the rest so the sites are physically unreachable, not just discouraged.
Pro tip Make unblocking require a deliberate second action (Orosz has to re-run his script) so it 'puts a ticket' between you and the distraction.
Watch out Distracting sites you use for research (Twitter, for Orosz) are the most dangerous — they justify staying open.
- 2
Set a 20-minute timer with zero other tasks allowed
Commit to just 20 minutes on one thing, nothing else permitted during that window. The short duration lowers the activation cost of starting.
- 3
Push through the first five minutes
Expect to grumble and wish you could check your feeds. Around the five-minute mark a switch flips and you go heads-down. Trust that this happens.
Pro tip Pair the timer with a co-working tool (Orosz uses Centered's camera feature — the sense someone might be watching forces work).
In the wild
Orosz wrote a simple Python script that kills/blocks LinkedIn, Twitter and Facebook via the hosts file. He runs it, sets a 20-minute timer, spends the first few minutes wishing he could look at Twitter to 'research', then around five minutes in flips into heads-down work.
→ This is the one technique that has 'consistently worked' for him, letting him ship two in-depth posts a week.
Common mistakes
Treating research tools as safe to leave open
Orosz keeps Twitter open for research, but it pulls his attention 'a little bit like slack, but worse' — leaving a justified-but-distracting tab open sabotages the whole session.
Expecting one trick to work forever
He finds a method works for a few months, then 'my brain just learns to work around it.' Assuming a single system is permanent leads to silent productivity decay; the fix is rotating methods.
Is it for you?
Best for
Self-employed knowledge workers and writers who struggle to start deep work amid social-media pull
Not ideal for
People whose work genuinely requires constant real-time access to the sites they'd want to block
From the transcript
“I set a timer of 20 minutes and then I say all right no distractions I have a script where I just kill”
“I just kill all LinkedIn Twitter”
“about five minutes in there's a switch and I'm like I'm now actually like heads down and doing it”
“I find it hard to start”
From the episode
Leaving big tech to build the #1 technology newsletter
Gergely Orosz (The Pragmatic Engineer)