The story

A founder I worked with — bootstrapped, seven people — decided to “finally get organized” and sat down one Saturday to document his processes.

He opened a blank doc, titled it “How We Do Sales,” wrote four paragraphs, got bored, opened a second doc titled “Company Culture,” wrote a mission statement he never used again, and quit by lunch. Two documents. Neither one ever answered a single real question, because neither one was about a question anyone was actually asking.

Six weeks later he was still fielding the same interruptions he’d sat down to eliminate.

Here’s what went wrong, and it’s the most common documentation mistake I see: he wrote the SOPs that sounded important instead of the SOPs that were actually being asked about. “How We Do Sales” is a nice thing to have. But nobody was Slacking him at 4 p.m. asking for the philosophy of sales. They were asking “what’s our refund policy again?” and “who approves this discount?” and “where’s the onboarding checklist?”

The right first SOPs aren’t the impressive ones. They’re the ones that plug the leaks you already have.

The principle

Read more

The brag

There’s a sentence founders say with their chest out, and I want to retire it.

“Honestly? Nothing happens around here without me.”

I’ve heard it in pitch meetings, on podcasts, at dinners. It’s delivered as a flex — proof of hustle, proof of indispensability, proof that this person is the engine. And for a while, it’s even true in a way that feels good.

Then one of two things happens.

They get the flu. Not a strategic long-weekend flu — a real one, the kind where you can’t open a laptop. And they watch, from bed, as the company just… slows to a stop. Deals wait. Questions pile up. Their phone becomes a small glowing emergency.

Or they don’t get sick, and instead they hit a quieter wall: they realize they physically cannot grow the company past the number of hours they personally can work. They’ve become the load-bearing wall in a building they wanted to add floors to.

Same diagnosis, two ways of discovering it. The founder isn’t the engine. The founder is the bottleneck.

Why it felt good (this part matters)

Here’s the thing most “stop being a bottleneck” advice skips, and skipping it is why the advice never sticks: being the bottleneck felt genuinely good, for genuinely real reasons. You can’t fix a thing you’re pretending you never enjoyed.

Early on, being the person everything flows through is a superpower. You have all the context. You make fast decisions because the whole picture lives in your head. Nothing gets lost in translation because there’s no translation — it’s just you, knowing. When someone asks a question, you produce the answer instantly, and that instant is a tiny hit of I am good at this, this is mine, they need me.

That’s not vanity. That’s the legitimate high of competence, and it’s real.

The trap is that the exact behavior that wins the early game loses the next one, and there’s no alarm that goes off at the transition. The skills that got you here — do everything, know everything, decide everything — are the skills capping you now. Nobody tells you the win condition changed. You just keep playing the game you were winning, and slowly start losing, and it feels like you’re not working hard enough, so you work harder, which makes it worse.

You were rewarded for becoming the bottleneck. Of course it felt good. That’s precisely why it’s so hard to give up.

The principle

Let me put the reframe as plainly as I can:

Your job was never to be the person who answers every question. It was to build the thing that answers them.

A bottleneck and a leader can look identical from the outside on a busy Tuesday. Both are in the middle of everything. Both are essential. The difference shows up only under one specific test — what happens when they’re gone.

When a bottleneck steps away, flow stops. When a leader steps away, flow continues, because the leader spent their time building the channels the flow runs through.

You don’t have to stop being central to the company. You have to stop being the only path through it.

What to actually do this week

You can’t platform-ify your whole company in a week. But you can run the diagnostic that tells you exactly where you’re the bottleneck, and take the first stone out of the wall.

1. Run the “if I vanished” audit. Open a notes file. Write the honest answer to: if I were unreachable for the next 10 business days, what specifically would break? Not vague (”everything”). Specific. “Client invoices wouldn’t go out.” “Nobody could approve a discount.” “The new hire would be stuck on day 3.” List them.

2. Sort each one into WHY it would break. Next to each item, write the real reason. You’ll find they almost all fall into one of three:

• It’s undocumented — the how-to only exists in your head.

• It’s un-permissioned — someone could do it, but they don’t have the access or the authority.

• It’s un-decided — there’s no rule, so it genuinely requires your judgment each time.

3. Take out one stone. Pick the single item that would break soonest and is undocumented (not un-decided — start with the easy category). Write the how-to. Give someone the access. Tell them it’s theirs now. One item. This week.

4. Notice the flinch. When you hand it off, you’ll feel a small resistance — but I do it better / faster / they’ll get it slightly wrong. Name that flinch, because that flinch is the bottleneck defending itself. “Slightly wrong and off your plate” beats “perfect and dependent on you” every single time, at company scale.

The thing nobody tells you

You’ll do this, it’ll work, and you’ll feel the relief of one thing running without you. Then you’ll notice the next problem: you have ten of these, and writing them one-at-a-time as they catch fire is its own kind of bottleneck.

That’s the difference between plugging a leak and building the plumbing. There’s a specific, boring, high-leverage set of processes that stops the most 2 a.m. questions — and it’s roughly the same ten for almost every founder.

That’s Wednesday.

Cager Unlimited builds the systems that let a founder step out of the middle without the company slowing down. If you’re the load-bearing wall in your own business, that’s exactly what we fix.

Know a founder who says “nothing happens without me” like it’s a good thing? Forward this. Gently.

Why your last wiki died

You’ve done this before. Almost every founder has.

Somewhere in your Drive there is a folder called Company Docs or Processes or — my favorite, and I have seen this one at three separate companies — Notion (NEW). It has eleven documents in it. Four are drafts. Two contradict each other. One is titled onboarding-v2-FINAL-actually-final. Nobody has opened any of them since the quarter they were written.

That wiki didn’t fail because you’re undisciplined. It failed because of three specific, diagnosable structural defects, and if you don’t fix all three, the next one dies exactly the same way:

Defect 1: No architecture. Documents were created reactively, one at a time, wherever felt convenient. There was no shape to the thing, so nobody could predict where anything lived. If a person has to search rather than navigate, they will ask a human instead. Search is a failure mode. Navigation is the product.

Defect 2: No owner. A document without a named owner is a document nobody updates. It becomes stale, then someone gets burned by the stale version, then everyone quietly agrees the wiki can’t be trusted — and once trust is gone, the wiki is dead even though every file still technically exists. Unowned documentation rots in about ninety days. Reliably.

Defect 3: No cadence. You wrote the docs in a burst of energy and then... nothing. There was no recurring, scheduled moment where anyone was obligated to look at them again. Documentation is not a project. It’s a practice. A project has an end date; that’s precisely the problem.

Everything below exists to solve those three defects, in that order. Architecture. Ownership. Cadence.

The principle

A knowledge base is not a place where you put documents. It is a machine that answers questions.

Read more

The teardown

I asked a founder — call him Marcus, seed-stage, eleven people — to guess how much of his week went to admin.

“Maybe fifteen percent,” he said. “Twenty on a bad week. It’s annoying but it’s not the story.”

Then he tracked it for five days. Not perfectly. Just a running note on his phone, every time he switched contexts, with a timestamp. Ugly data. Good enough.

Here is what the ugly data said.

What Marcus thoughtWhat Marcus did~8 hrs/week admin31 hrs/week admin“A few Slack questions”86 distinct questions answered from memory”I’m mostly on product and sales”4.5 hrs on product. Total. In a 58-hour week.

He was off by nearly four times. That is not a Marcus problem — that is the normal result, and I have watched it play out enough times to stop being surprised by it. Founders are catastrophically bad at estimating their own admin load, and the reason is structural: admin doesn’t arrive in blocks. It arrives in interruptions.

You don’t remember the four minutes you spent finding a contract, or the six minutes explaining the reimbursement policy, or the eleven minutes untangling who was supposed to send the invoice. Each one is too small to remember. Together, they eat the company.

Your calendar is a record of your intentions. This audit is a record of your life.

The principle

You cannot delegate what you have never measured, and you cannot measure what you refuse to write down.

Read more

The email

It’s 2:04 in the morning and Dana is awake, which is not unusual, but she is awake for a stupid reason, which is.

Her newest hire — six days in, eager, three time zones east — has sent a message. Not a crisis. Not a client escalation. Just this:

Hey, sorry to bug you — where do I find the client onboarding checklist? I looked in Drive but there are like four folders that could be it.

Dana knows the answer. Of course she knows the answer. She is the answer. She types it out from her phone in the dark, adds a link, adds a caveat about the version that’s out of date, adds a second caveat about the version that replaced the version that’s out of date, hits send, and lies back down.

And then she does the math, because that’s what founders do at 2 a.m. instead of sleeping.

This is the fifth time this month she has answered this exact question. Different person, same question. Before that it was the reimbursement policy. Before that it was who approves a client discount. Before that it was where the contractor agreement template lives, which she answered by sending a screenshot of her own desktop.

Dana is not running a company. Dana is running a help desk with one employee, and that employee is the CEO.

The principle

Here is the part almost nobody says out loud, so I’ll say it:

Being the person who knows everything feels like competence. It is actually a liability that has been dressed up as a virtue.

There’s a specific kind of pride in being indispensable. When someone asks and you answer instantly, you get a tiny hit of usefulness. You built this. You know where every wire runs. Of course they need you.

But notice what’s actually happening in the mechanics of that exchange:

• The question gets answered once, for one person, in a private channel, from memory.

• The answer leaves no trace. It doesn’t accumulate. It doesn’t compound.

• The next person to have that question will have to ask a human being — and the only human being who can answer is you.

You’ve created a company where information can only travel through you. Every question routes to you. Every answer routes from you. You are not the leader of this system. You are the bottleneck in it — and worse, you’re the single point of failure.

Take a sick day and watch what happens. Not a vacation — vacations get planned around. A real, unplanned, flat-on-your-back sick day. Things stall. Not because your team is incapable, but because you never gave them anything to be capable with.

The 2 a.m. email is not a people problem. Your new hire did everything right — they looked first, they couldn’t find it, they asked. The 2 a.m. email is a systems problem wearing a people costume.

And systems problems have one useful property: they can be fixed once.

The test

Here’s the line I want you to keep:

If a question has been asked twice, the answer belongs in a document — not in your head.

Twice. Not five times. Not “once it becomes a real pattern.” Twice.

The first time someone asks, it’s a conversation. The second time, it’s a signal — you have now proven that this information is (a) needed by more than one person and (b) not findable anywhere. That is the entire qualifying test. It costs you eleven minutes to write the answer down properly. It costs you eleven minutes every single time to keep answering it live, forever, including at 2 a.m., including on the day you have the flu, including on the day you finally tried to take a real vacation.

What to actually do this week

Not a transformation. Four small moves, and you can do them in about ninety minutes total.

1. Run the seven-day scrollback. Open Slack, your texts, and your email. Scroll back one week. Write down — literally in a notes file — every question a team member, contractor, or client asked you that you had to answer from memory. Don’t judge them yet. Just list them. Most founders find between twelve and thirty.

2. Mark the repeats. Go down the list and put a star next to anything you have answered more than once, ever. These are your leaks. This is your homework, and it wrote itself.

3. Write the top five. Badly. Not a beautiful wiki. Five documents, each one screen long, in whatever tool you already have. Title, the answer, who owns it, the date you wrote it. That’s the whole format. A rough document that exists beats a perfect one that’s still in your head. Perfectionism is how documentation dies before it’s born.

4. Change your reflex. This is the hard one, and it’s the one that actually compounds. The next time someone asks you a repeat question, do not answer it in the DM. Instead:

“Great question — I’m writing this one up so it’s findable. Give me ten minutes.”

Then write it, and send the link, not the answer.

You will feel slower this week. You will be dramatically faster in six.

The thing nobody tells you

You’re going to do the four steps above and it’s going to work — for about ninety days. Then the documents will go stale, nobody will know which ones are current, and you’ll be back to answering from memory because your own wiki has become untrustworthy.

That’s not a failure of discipline. It’s the difference between writing some documents and building an actual system — one with an architecture, an owner for every piece, and a maintenance cadence that runs without you nagging anyone.

That’s where this month is going.

Cager Unlimited builds customized digital HR infrastructure for founders — the systems, documentation, and processes that answer the 2 a.m. email so you don’t have to. If you’re tired of being the only API in your own company, let’s talk.

If this hit close to home, forward it to the founder in your life who is definitely also awake right now.

Subscribe now

If you’re a founder, you already know the feeling this newsletter is named after.

It’s 9:40 on a Tuesday and you’ve been “working” since 7. You’re tired in a way that doesn’t match what you actually produced, because the day got eaten — not by strategy or product or the big scary thing you keep meaning to tackle, but by a hundred small things only you could apparently do. Approving an expense. Finding a file. Re-explaining a process for the fourth time. Answering a question at an hour no one should be awake to answer it.

None of it was hard. All of it was yours. And you can’t quite point to where the day went.

That’s a Founder Problem — and that’s what this is about.

Who I am, and why this exists

I run Cager Unlimited. We build HR and operational infrastructure for founders — the systems, documentation, and people-ops backbone that let a business run on something other than the founder’s memory and stamina.

I’ve watched the same pattern play out across a lot of small companies, and it’s remarkably consistent: the thing that makes you a great founder in year one — doing everything, holding it all in your head, being the person nothing happens without — is the exact thing that quietly caps your company in year two. You don’t have a growth problem. You have a you’re-the-bottleneck problem, and nobody warned you it was coming because it arrives disguised as being helpful.

I started this newsletter because most advice for founders is either 30,000-foot strategy that doesn’t tell you what to do on Monday morning, or generic HR content written for companies with a People team of twelve. There’s a huge gap in the middle: the founder who’s doing HR and ops by accident, at night, with no template and no time. That founder is who I write for.

What to expect

Every week, I take one real founder problem and go at it from three angles:

• Monday (free): the story. Relatable, honest, sometimes uncomfortable — the trap and why it feels so reasonable while you’re walking into it. Free forever, made to be shared with the founder friend who needs it.

• Wednesday (paid): the deeper cut. The actual framework, template, or checklist to fix the thing — the stuff worth paying for because you can use it the same afternoon.

• Friday (founding member): the complete build. The full system, the done-with-you version, the whole set of assets for the founders who want to implement, not just read.

Across a year, these ladder through the entire founder journey: solo overwhelm, your first hire, building a repeatable hiring machine, onboarding, compensation, the boring-but-lethal compliance stuff, running a remote team, culture, performance, the hard conversations and exits, delegation, and finally scaling real HR infrastructure. It’s a map of every people-and-ops trap ahead of you — and how to get out of each.

The Wednesday and Friday tiers are for the people who want the templates, systems, and depth — and, honestly, for the people who want to support this and keep it going.

One small ask

If any of this sounds like your Tuesday night, do two things:

Subscribe — so Monday’s story lands in your inbox and you don’t have to remember to come back.

Share it with one founder who’s currently the bottleneck in their own company and doesn’t quite realize it yet. That person is exactly who I wrote this for, and they’ll thank you.

The first real post goes out Monday. It’s called The 2 AM Email That Should Have Been a System, and if you’ve ever answered the same question at an hour you’re embarrassed to admit, it’s going to feel a little personal.

See you then.

— Cager Unlimited