You can just do things
In April I entered my first BJJ competition and came away with a bronze and a silver. The medals aren’t the part I think about.
The part I think about is that I almost didn’t enter, and the voice that nearly talked me out of it sounded completely reasonable. You’re not ready. You’ll get smashed. Wait for the next one. Every clause of that is defensible. All of it is a request for permission, addressed to nobody.
There’s a line that’s been going around for a while now: you can just do things. It reads like a slogan for people with no dependants. It isn’t. It’s the most practically useful sentence I’ve come across in two years, and it’s now written down in a private document of mine as the one thing I most want my kids to end up believing.
The permission never arrives
There’s a version of starting something that looks like quitting the job, raising capital, burning the boats. It makes good keynote material. It also requires permission - from your partner, your bank account, your risk tolerance - and for most people with a mortgage and two kids, that permission doesn’t arrive. So they wait. And the waiting quietly becomes the plan.
I know the shape of it because I ran it for years. I’ll start when the kids are older. I’ll start when there’s more buffer. I’ll start when the day job settles down. Each of those is reasonable. Each is an indefinite postponement dressed as prudence. The kids will always have something on. The number that feels safe keeps moving. The job never settles.
The need for conditions to be right before starting isn’t caution. It’s avoidance wearing a sensible disguise.
Start small enough that failing doesn’t matter
In June last year I walked into a Brazilian jiu jitsu gym at thirty-four, years after the last time I’d trained, and took a fresh white belt. That wasn’t a decision. It was a Tuesday. Nobody was consulted, no plan was written, and if I’d quit after three weeks the cost would have been a direct debit and some bruising.
That’s the whole mechanism. Not commitment - scale. Make the first thing small enough that failing at it is genuinely irrelevant, and the permission question stops being load-bearing. You don’t need a mandate to try something that costs you nothing to be wrong about.
Then it accumulates. Over five weeks last summer I trained forty-two times. I didn’t set out to. I set out to go on Monday.
The conviction arrives after the work
The belief you’re waiting for is downstream of the doing, not upstream. I don’t get confident and then build. I build, and the confidence turns up afterwards, and it decays if I stop. Mine regenerates through shipping, not through thinking about it at night. Rumination has never once resolved a doubt for me. Contact with the actual work does it every time.
So the business didn’t start with a leap either. By the time I handed in my notice, the thing already existed. Resigning was the least interesting step in it. The public launch came later again, and the first inbound enquiry, from a firm I’d never approached, landed three days after it.
That’s the opposite of the founder story. I didn’t jump and trust the evidence to appear. I gathered evidence until jumping was the boring option. Every step was small enough on its own to be survivable, and the whole sequence only looks like a decision in hindsight.
Between the last day of employment and the first day of self-employment, I competed again, and bruised a rib properly. Three to four weeks off the mat. So much for the clean narrative arc.
The ledger includes the failures
Since December I’ve created thirty-six git repositories. Eight of them have exactly one commit - opened, poked at, never touched again. One tool went from nothing to finished in a single day, twenty-nine commits deep. Another took two days. A third took thirty minutes to build and two and a half hours to polish, which taught me more than the tool did.
In my first full month working for myself I pushed 863 commits across fourteen repositories.
A couple of weekends ago I rebuilt a product’s entire front end on a different foundation. Three days: demo environment standing, the commercial case written up, the whole thing genuinely working. Three days after that I archived it and went back to what I already had, because building it was the only way to establish it was the wrong trade. A weekend is a cheap price for an answer you can’t get by thinking about it.
I’m not putting those numbers up as a flex. I’m putting them up because they’re the same ledger. The eight dead repos and the three that turned into client platforms were all started the same way, with the same amount of certainty, which was none. There was no moment where I could tell them apart in advance. If I’d waited until I could, I’d have written zero of them.
Volume is what makes the hit rate stop mattering.
And you can put them down
Years ago I started a garden share group in my area. A Facebook group and a monthly meetup - people turning up at someone’s house or a local park, swapping seedlings and surplus produce. There was no business case. There was barely a plan. I wanted the thing to exist, so I made it exist, and then people came.
It ran for years. I hadn’t been to a meetup in about eight months.
Recently it moved across to an established not-for-profit that’s better set up to carry it than I was. New name, new branding, and I’m no longer the founder or the admin of anything.
I don’t think the starting is what actually stops people. It’s the suspicion that starting is a life sentence - that anything you begin, you’ll be carrying in ten years whether it still fits or not. That’s a reasonable fear, and it makes the permission question feel enormous, because now you’re not deciding whether to try something. You’re deciding whether to be the sort of person who does this forever.
You aren’t. The things you start are allowed to end, or to outgrow you, or to belong to someone else. A thing you founded getting too big for what you wanted it to be is a success condition, not a loss.
Starting is cheap precisely because stopping is permitted. If it weren’t, none of the rest of this would be worth recommending.
It costs something
This is where the slogan usually stops, so let me not stop there.
The month I lost the mat to that rib was the exact month the work went vertical, and I didn’t spot the collision until someone pointed at it. Household income dropped the day I stopped being employed and hasn’t fully closed the gap yet. There are commit timestamps in my history at 1am and 2am and 3am that are not evidence of dedication - they’re evidence of a man trying to buy back the hours he spent being a dad that day.
At one point this year, working through what was actually driving the pace, the honest version came out as: you’re over-working because you’re scared about the gap, and the kids are paying for your fear.
“You can just do things” is not “and it will all be fine.” It’s a statement about who gets to authorise the attempt. It says nothing whatsoever about the bill, and the bill turns up regardless. Anyone selling it as a cheat code is selling something.
What it’s actually for
None of the above is why the line sits in that document.
I’ve got two kids who are watching how this goes. Not listening - watching. You can tell a child they’re capable of anything and it lands as noise, because it’s the sort of thing parents say. What doesn’t land as noise is a house where the adults enter competitions they aren’t ready for, start things that don’t work, and keep going anyway.
I can’t teach it in a conversation. I’ve tried. The only version that transmits is the one they live next to.
That’s the actual return on all of it. Not the medals and not the invoices. Two kids who grow up assuming that if you want to make a thing, you make it, and you don’t wait for someone to tell you you’re allowed.
Nobody is coming to give you the tick. That was never the obstacle - it’s the opening.