Signal to Noise
Steve Jobs had a filter. He called it signal to noise ratio.
It's an engineering term. In audio or communications, signal is the thing you actually want. Noise is everything interfering with it. The goal is simple: maximize signal, minimize noise.
Jobs applied this to everything. Products, decisions, his calendar, the companies he ran. He believed that most of what fills a day - most meetings, most features, most options - is noise. And noise doesn't just waste time. It actively drowns out the signal. You can't hear what matters when everything is competing for the same attention.
That's the part that stuck with me.
When Jobs came back to Apple in 1997, the company was weeks away from bankruptcy and had dozens of products in its lineup. His first major move wasn't to launch something new. It was to cut. He killed roughly 70% of Apple's product line and focused on four things.
Four.
The reasoning wasn't purely financial. It was philosophical. He believed that a company - like a person - can only do a few things exceptionally well. Spreading attention across everything means nothing gets the focus it deserves. The result is a lot of average output instead of a few remarkable ones.
He said Apple's job was to say no to a thousand things so they could say yes to the right one. That ratio - a thousand no's for every yes - is what signal to noise looks like in practice.
I think about my own days as a developer and how rarely they reflect that ratio.
The default mode is the Jira board. There's always a list of issues — bugs, tasks, tickets - and the day fills itself around resolving them. Moving cards feels productive. Closing tickets feels like progress. So you work through the queue, context switching between one issue and the next, and somewhere in the middle of all that movement is the one thing that actually mattered - buried under everything else.
The noise isn't always obvious either. Some of it looks exactly like work. A bug fix is work. A code review is work. Responding to a comment on a ticket is work. But not all of it is signal. Some of it is just queue management - motion without direction, effort without return.
The question Jobs implicitly kept asking was: does this move the thing forward? Not does this clear the board, not does this keep the tickets green. Does it actually move the thing forward.
Most things don't.
The hard part is that noise is comfortable. It fills time without requiring the kind of deep focus that real work demands. You can stay busy all day on noise and feel like you've done something.
Signal work is harder - it requires clarity about what actually matters, which means you have to decide, which means you have to commit.
Deciding is uncomfortable. Committing closes doors. So a lot of people - myself included - stay in the noise because it's easier than confronting the signal.
Jobs wasn't comfortable with that. He had a reputation for being difficult, for saying no abruptly, for cutting things that people had worked hard on. Some of that was personality. But some of it was the filter working. If it doesn't belong, it goes. Feelings about it are separate.
That clarity is rare. And it shows in the output.
I've started asking a simpler version of the question at the start of a day: if I only got one thing done today, what would it need to be?
Not a list. One thing.
Everything else either supports that or it's noise. It doesn't mean the other things don't get done - they usually do. But having the signal named means I know what I'm protecting the day for. It changes what I say yes to, what I let interrupt me, what I let run long.
It's a small filter. But it's the same instinct Jobs was working from, just at a different scale.
Most days have more noise than signal. The work is figuring out which is which - and being honest enough to act on the answer.
