Don’t Approve The Hire. Approve The Experiment.

“We need another person.”

Those four words can bring a small-business budget meeting to a halt.

The manager making the request sees overdue work, exhausted employees, and missed opportunities. The owner hears salary, benefits, payroll taxes, equipment, supervision, and a commitment that may last for years.

Both are responding rationally. The manager is experiencing the cost of insufficient capacity. The owner is imagining the risk of fixed overhead.

The conversation usually becomes a contest of confidence.

“I know we need someone.”

“I’m not convinced.”

There is a better question: what is the smallest sensible experiment that would give us evidence to support the request for hiring another person?

Annual salaries make poor opening bids

Suppose a manager proposes hiring an administrative coordinator for $50,000 a year. Leaders immediately debate whether the company can afford $50,000.

But the company does not need to risk the full annual amount to learn whether additional support creates value.

It might begin with a contractor for ten hours a week. It might hire a part-time employee for 90 days. It might use a bookkeeping service, virtual assistant, or temporary placement. The initial decision could be $5,000 or $10,000 rather than $50,000.

It sizes the risk to the uncertainty.

Many small businesses understand this principle in marketing. They test a campaign before committing the annual budget. They sample a product before ordering a container. They pilot new software with one team.

Capacity deserves the same discipline.

Write the hypothesis

Every hiring experiment should begin with a falsifiable hypothesis.

For example:

“If we add ten hours a week of administrative accounting support, the operations director will recover at least eight hours for inventory planning and supplier management, overdue reconciliations will decline, and supplier payments will be prepared accurately each week.”

That is much stronger than “Matt is overwhelmed.”

The hypothesis identifies the intervention, the expected capacity returned, and the business outcomes. At the end of the test, leaders can evaluate what happened.

Without a hypothesis, temporary support merely makes everyone feel better for a while. That may be worthwhile, but it does not produce a decision.

Measure three kinds of return

A capacity experiment should track three forms of value:

1. Work transferred

Which tasks were moved to the new person? How many hours did they require? What error rate or rework occurred?

2. Capacity recovered

Did the overloaded employee actually stop doing the transferred work? Recovered time has no value if the leader continues checking every detail or fills the space with more low-value activity.

3. Value created

What happened because capacity was recovered? Faster invoicing, improved collections, fewer expedited shipments, better scheduling, more client meetings, documented processes, or a developed team member may all count.

Some returns are immediate and financial. Others are leading indicators. A manager who finally trains a successor may not produce revenue within 90 days, but the company has reduced a material dependency.

Agree on the evidence before the experiment begins.

Design the role around outcomes, not leftovers

A common mistake is to give the new person a miscellaneous pile of tasks nobody else wants. The role becomes an organizational junk drawer.

Instead, define a coherent service.

For administrative accounting support, the service might include matching invoices to purchase orders, preparing the weekly payables list, maintaining collection notes, and reconciling customer payments. The operations leader retains supplier priorities and exceptions. Finance retains cash decisions and controls.

This makes delegation easier because decision rights remain clear.

It also reveals whether the company truly needs a role or merely needs to repair a process. During the experiment, the new person may discover duplicate approvals, inconsistent invoice numbers, or missing information. Eliminating those problems can reduce the permanent workload.

The best hiring test may prove that no full-time hire is required.

Protect the experiment from its sponsor

Capable leaders are often terrible at receiving support.

They delegate a task, redo it themselves, check every step, and conclude that delegation takes longer than doing the work. In the first week, they may be correct. Training is an investment. The return arrives later.

The experiment needs explicit rules:

• Which tasks will be transferred completely?
• Which decisions still require approval?
• How will quality be checked?
• How quickly should questions be answered?
• When will the process be reviewed?
• What will the leader do with the recovered time?

That last question is essential. If the senior employee cannot name the higher-value work they will perform, the business case is incomplete.

Decide before the test ends

Set the decision date at the beginning.

At 30 days, review the handoff. At 60 days, assess the measures. At 90 days, choose:

• Stop because the hypothesis was disproven.
• Modify the role or process and run a second test.
• Continue the part-time arrangement.
• Convert the work into a permanent role.
• Expand the role because the return is stronger than expected.

Without a decision date, pilots become permanent by inertia. The company pays for a solution it never formally evaluates.

Treat capacity as an investment portfolio

Owners are right to worry about overhead. Payroll commitments can sink a business when revenue falls.

They should worry just as much about constrained capacity. A company can miss its future while protecting its present. Senior leaders spend their days processing transactions, opportunities go unfollowed, and weak systems remain weak because nobody has time to improve them.

The answer isn’t indiscriminate hiring. It’s disciplined experimentation.

Ask managers to propose capacity investments the way they would propose any other business investment: the problem, the hypothesis, the smallest useful test, the measures, the budget, and the decision date.

Then stop arguing about whether someone “feels busy.”

Approve the experiment and let the evidence do some of the talking.