When Business Picks Up, the Instinct Is to Hire. Here’s How That Can Backfire.

The first time someone told me to hire, I did.

We had just taken Bud Authority from nothing toward a real book of business. The work was piling up — content that needed to go live, client sites that needed a technical look, broken links, a list of findings that had to get back to us and then to the client.

The obvious move was a person. Someone whose job was to post, crawl, check and come back with a report. I thought of it as an assistant seat. It was not. It was a messenger seat. We hired for it.

Capacity was the plan. A queue was the result.

On paper, the seat looked clean. Open the CMS. Push the pages. Walk the site. Note what broke. Put it in a doc. Bring the doc to the people who could actually change the work. That is a full Tuesday if you write it that way. It is also a description of a courier route, not a production role.

About two months in, the work was still slow. Not because the person was a villain. Because the job we had designed was: Go find out, then come tell us. Every finding had to wait on a human to notice it, write it down and walk it across the room. The seat was supposed to add capacity. It added a queue.

I watched the queue form in small ways. A broken link sat until someone remembered to check. A missing page sat until someone opened the list. A client question waited on the next report cycle because the only path from “known” to “ours” was a person carrying it. The operation was not short on eyes. It was short on a standard that made the finding show up without a handoff.

We did not need a better hire. We needed the findings to exist without a messenger.

There is a version of this story that starts with the number. We went from $0 to $1 million in less than a year. I will not lead with that, because the number is what made the bad hire feel responsible. When revenue moves, you start buying seats so the shop looks like the number. That is how you inherit a slower company than the one that made the money.

The problem wasn’t headcount

I had already watched a different company try to scale output by adding writers and then developers to hang the pages. More drafts. More tickets. More people standing between the idea and the live page. Throughput did not rise with headcount. It rose when the path from write to publish stopped depending on a chain of “tell the next person.” This was the cousin of that mistake. Not more producers. A courier. Same instinct: If we are behind, add a human to the gap.

That is the part most agency owners get backward when the book starts to grow. You feel behind, so you buy a pair of hands. The hands do not fix the delay. They become the delay. The report is only as fast as the person who has to remember to make it.

The messenger seat trains the whole shop to wait. Producers wait for the list. Clients wait for the update. You wait for the meeting where someone reads the findings out loud. Everyone gets busy. Nothing gets earlier. Busy is not the same as closed.

I am not going to dress this up as a software story. We changed how the operation ran. What we built is our business. What I will say in public is the decision: If a job exists only to notice, collect and deliver what the work already knows, that job is a delay with a salary.

A loop is different from a report. A loop means the check happens on a schedule or a trigger, the result lands where the fixer already works, and the next action does not require a briefing. A report means someone has to assemble the truth, then schedule the room where the truth gets heard. One of those is ops. The other is theater that feels like ops because there is a person attached to it.

The $1 million year is not the lesson. Plenty of firms hit a number and hire their way into a slower shop. The lesson is that the first ops seat we bought was a person standing between the problem and the people who could fix it. Two months was long enough to see it. We did not need that person. We needed the loop to close without a meeting.

Hire for the work that changes the client outcome

If you run a services business, look at the last role you opened because you were underwater. Write down what that person will actually do on a Tuesday. If the list is post this, check that, tell us what you found, then you are not hiring capacity. You are hiring a delay and calling it help.

Be specific. Separate the work that changes the client outcome from the work that only moves information about the outcome. Publishing a page changes the outcome. Logging that the page is live is information. Fixing a 404 changes the outcome. Pasting the 404 into a weekly sheet is information. Hire for the first kind. Automate, standardize or kill the second kind before you put a salary on it.

Ask an uglier question than “Can we afford them?” Ask: “If this person disappeared on Wednesday, would the work stop existing, or would the report stop existing?” If only the report stops, you do not have a people problem. You have a loop that depends on someone remembering to run it.

That question is uncomfortable when you like the person. Like them anyway. The design of the seat is still the design of the seat. Loyalty to a hire does not turn a messenger role into production. It just makes you slower to admit what you bought.

Do not confuse a client update with a client result. A finding that sits in a doc until Friday is not quality control. It is latency. The client does not buy your headcount. They buy the broken thing getting named and fixed. If naming it requires a full-time messenger, you will always be late, and you will always think the answer is one more body.

I still hire people. I hire for judgment, for craft and for work that dies when the human walks out. I do not hire people to carry information the operation should already have. That sounds obvious until you are the one who posted the job, onboarded the seat and spent two months wondering why the work still felt slow.

Fix the loop

Speed is not a personality trait. It is whether the work has to wait for a person whose only job is to notice it. Cut that wait, and the seat often disappears. Keep the seat, and you can grow the revenue and still feel like a $0 shop that just got louder.

The standard matters more than the hire. Once the check is required, timed and visible without a courier, you stop arguing about who forgot to look. You argue about the fix. That is the right argument. The wrong argument is staffing a hallway so truth can walk from one room to another.

Operators love to say they hire slow and fire fast. Fine. The useful version is: Hire only for work that disappears if the human does. Do not hire for work that is actually a report.

If your next hire is a messenger, do not hire them. Fix the loop. If you already hired them, two months is not a failure. It is the first honest look at the job you actually bought.