Add Headcount for Volume. Fix the Process for Rework.

The request arrives with evidence attached. A team is behind, the backlog is growing, the people on it are working evenings, and the manager has a spreadsheet showing that one more person would close the gap. Every line of it is true. What the spreadsheet cannot show is whether the gap is a shortage of hands or a shortage of design, and the choice between adding headcount and fixing the process turns on exactly that difference.

The case that always looks sound

A hire is the easiest response a growing company has to strain. It is legible, it fits the planning cycle, and it comes with an owner who will be grateful for it. A process fix is none of those things. It crosses team boundaries, it asks someone to give up a step they defend, and its benefit is an absence: work that stops arriving. So the hiring case is usually made well, and the process case is usually not made at all.

The planning cycle asks whether the team can justify another person. The question underneath is whether the work the team is drowning in needs to exist in its current form. In most of the growing companies we work with, the first question gets a careful answer and the second gets none, because no forum owns it.

Some strain is genuine volume. More customers, more transactions, more accounts, each needing the same competent work as before. Adding people to that is correct, and delaying the hire to run a process review first is its own kind of avoidance.

Much strain is not volume. It is rework, waiting and handoff, and each of those grows with headcount rather than against it. A new person joins the loop, inherits its habits, and produces more of whatever it already produces. If a large share of what a team handles is correcting cases that arrived incomplete, another analyst corrects more of them, faster, and the cases keep arriving incomplete.

How it shows in onboarding

Digital banking onboarding gives a clean version. A manual review queue builds up behind identity verification, and the case for more reviewers is strong: response times are slipping and applicants abandon while they wait. Read the queue item by item, and much of it often traces to one cause upstream: an automated document check that rejects images a person would accept, or a form field that is never validated at entry. Each additional reviewer clears that queue at the same rate as the last one. Nobody is assigned to the rule that fills it.

A hire answers the question of how much work there is. It never asks why the work exists.

Checkout and payment operations show the same shape in disputes and refunds. A team handling chargebacks grows with transaction volume, which looks proportionate, until it turns out that a recurring share of disputes comes from a billing descriptor customers do not recognise on their statement. That is a one-line fix in another team's backlog, and it sits there while the disputes team keeps hiring.

Busyness is the wrong signal, because every strained team is busy whether the cause is volume or design. The useful signal is where work sits idle, and for whom. A team short of capacity has work waiting for its own people. A team caught in a broken process has work waiting for someone else: an approval, a missing input, a decision deferred to the next meeting.

Signs the hire is going around a broken process

Four checks separate the two, and each can be run against the last month of real work rather than against anyone's account of it:

• the share of items that come back for a second pass, and where they came from
• how long work waits between steps compared with how long it takes to do
• how many people touch a single case before it closes, and whether that number grew with the team
• whether the last hire into the same role changed the backlog, or only the size of the team holding it

The last check tells the most. When a team has hired twice into the same strain and the strain is unchanged, headcount growth has stopped solving the problem. The second hire was an answer to the failure of the first, and nobody read it that way at the time.

Handoffs grow faster than people

Coordination cost also rises faster than the team. Each new person adds handoffs, a seat in the standup, and one more party who needs to hear when something changes. In a sound process that cost is small and worth paying. In a broken one it compounds the original fault, which is how a team can grow by half and deliver roughly what it delivered before. It is the arithmetic of growth that has become noise, applied to people instead of dashboards.

The answer is rarely one or the other. It is usually an order: remove the cause that generates rework, measure the volume that remains, then hire against that. The process fix writes the job description. A hire made after it is a hire for the work the company actually has, not for the work its broken loop manufactures.

Sequencing under pressure

The objection is time. The team is overloaded now, and finding the cause takes weeks the backlog does not have. The objection is fair, and the answer is to separate the two decisions rather than choose between them. Temporary capacity, a contractor or a person borrowed from an adjacent team, holds the line while the cause is found. The permanent hire waits until the fix has run long enough to show what is left. That order only holds if the fix is a decision with an owner and a date rather than a deferral.

The order matters because a permanent hire into a broken process tends to become its defender. The new role is built around the rework, and when the cause is later removed, the role has to be redesigned or ended, which is a harder conversation than the one that was skipped. The fix itself often needs a decision about ownership, since the upstream rule that fills the queue usually belongs to a team that never feels the queue. That is the territory of decision rights that were inherited rather than assigned.

Sales capacity follows the same rule

Revenue teams run the same pattern under a different name. A pipeline shortfall gets answered with more account executives. If deals are stalling at a pricing approval that takes three weeks, more representatives produce more stalled deals, and the approval chain becomes the busiest step in the company without anyone hiring for it.

The diagnostic that separates volume from design is short. It reads a month of real cases, finds where they wait and why they return, and puts a cost on the upstream cause next to the cost of the hire. That is part of our consulting work under growth discipline, usually done before a headcount plan is approved rather than after it has been spent.

The outcome is often still a hire, but a smaller one, made later, into a role that will not need to be undone. Where a team wants to run the same test on its own planning cycle, we take it up as training for the managers who write the hiring cases.