You're looking at the calendar, the backlog, the general sense that everyone's stretched thin, and the conclusion feels obvious: you need to hire. More hands, more capacity, problem solved. It's the instinct almost every founder reaches for first, and it's not wrong exactly — it's just usually early. Before the hire, there's a cheaper question worth answering: what is your current team actually spending its hours on, and how much of that is work a system should be doing instead?
I ask founders this constantly and get some version of the same surprised pause every time: nobody's actually looked. Everyone knows the general vibe of being busy, but almost nobody's traced where the hours specifically go. And when you do trace it, it's rarely one big dramatic inefficiency. It's three or four small, repeated, manual tasks that nobody's gotten around to fixing because each one, alone, feels too small to justify the effort.
That's the actual pattern underneath we need to hire: not too little capacity, but capacity quietly being spent on work that shouldn't require a person doing it by hand every single week.
There's a reason this is so easy to miss from the inside: busy and inefficient produce the exact same physical sensation. Both feel like there aren't enough hours. The only way to tell them apart is to actually look at what's filling the hours, which almost nobody does under deadline pressure, because looking feels like a luxury you don't have time for right when you need it most.
A new hire doesn't fix a system problem. It just gives the system problem a second person to run it manually.
Where the redone work actually hides
It's rarely in the visible, exciting parts of the business. It's in the client intake form that gets manually copied into three other places. The invoice that gets built from scratch every month instead of from a template with the numbers dropped in. The status update that gets typed out fresh to five different people who could've just been looking at the same shared dashboard. None of it looks like a crisis. All of it adds up to hours a week that never show up on anyone's radar as the actual problem.
None of these tasks would survive being pitched on their own. Nobody would approve a line item that said pay someone to retype the same information three times a week. But nobody's pitching it — it's just quietly happening, task by task, absorbed into the general shape of the job until it stops being visible as a decision at all.
The reason it stays invisible is that each individual task is small enough to just do. Nobody's going to escalate I copy this form into a spreadsheet every day as a business emergency. It's a two-minute task. It's just a two-minute task that happens every single day, to someone, forever, until somebody finally adds it up and realizes it's actually eight hours a month of a real person's attention spent on something a system could be doing invisibly, correctly, every time.
Why hiring doesn't actually solve it
Bring on a new person before fixing this, and you haven't removed the manual work — you've just handed it to someone new, who now has to learn the exact same clunky process everyone else has been quietly tolerating. You've added a salary to a problem that was never about headcount. The bottleneck was never the number of people. It was the number of things still being done by hand that shouldn't require a hand at all.
This is also usually where founders lose the most ground on hiring itself: onboarding someone into a system full of manual workarounds takes far longer than onboarding them into one that's actually built. You end up training your new hire on the same duct tape everyone else has been quietly maintaining, instead of handing them something that already works, so they can focus on the part of the job you actually hired them for.
I see this constantly with founders who are certain their growing pains are a headcount problem: they hire, the relief lasts about six weeks, and then the new person is just as underwater as everyone else, because the actual bottleneck — the manual process nobody fixed — is still sitting there, now serving two people badly instead of one.
Fix the system first, and the hire you make afterward does the job you actually needed. Hire first, and you've just found someone new to do the job you should have automated.
How to actually find the three things
Ask your team directly: what's the task you do most often that feels the most repetitive and the least like it should require a human? Not the hardest part of their job — the most repeated, most mechanical part. You'll usually hear the same one or two answers from multiple people independently, which is exactly how you know you've found something real.
Then build for those specifically, before touching anything else. You don't need a full systems overhaul to start. A template that auto-fills instead of getting built from scratch. A form that feeds directly into the tool you already use instead of getting copied by hand. A shared dashboard instead of five separate status updates. Small, specific fixes to the exact three things people named — that's the whole first move, and it's usually enough to change what hiring actually needs to solve next.
Do this before writing the job listing, not after. A role built to absorb three redundant manual tasks looks completely different on paper than one built to do the actual, higher-value work you're hiring for in the first place — and candidates can tell the difference between a job description that's really about mopping up busywork and one that's about a real seat on the team.
The takeaway
This week, ask each person on your team one question: what do you do the most often that feels the least like it needs a person doing it? Write down the answers. 🗂️
That list is your actual next hire's job description — except half of it might not need a hire at all. Some of it just needs to stop being done by hand.
Run this exercise every time growth makes you reach for the job listing again, not just once. The redone tasks change as the business changes — what's manual today won't be the same thing that's manual in a year — so the question is worth asking on a schedule, not as a one-time fix you check off and forget about. 🤍
A Digital Systems Scan finds exactly which tasks are costing your team hours before you post the job listing. Start here: Discover