Should You Delegate This Task? Or Just Do It Yourself?

Why Delegation ROI Is Almost Never Calculated

Most managers make delegation decisions based on feeling, not math. The feeling usually goes: "I could teach this, but it would take longer than just doing it myself. So I will just do it." That reasoning is short-term correct and long-term catastrophic. A task done by you once is an hour lost. A task done by you 52 times per year for 3 years is 156 hours lost, plus all the second-order costs: you become a bottleneck, the person you should have trained never grows, and your whole team plateaus at the ceiling of what you personally can produce.

The Math Behind the Calculator

We use loaded hourly cost (salary ÷ 2,080 × 1.3) to approximate what the company actually pays per hour of someone's time. The 1.3x factor covers benefits (health insurance, 401k match), payroll taxes (FICA, unemployment), and standard overhead. This is the accepted approximation used by most corporate finance teams when building labor cost models.

For the delegation scenario, we add four costs: (1) a one-time training investment equal to your hourly cost × training hours, (2) review overhead during the ramp-up period equal to your hourly cost × review hours per instance × number of ramp instances, (3) the employee's time to do the task at their hourly rate, and (4) a quality correction cost equal to the quality drop percentage × ramp instances × task value.

The break-even calculation compares cumulative cost of both paths instance by instance. The moment the delegation path's cumulative cost drops below the manager-does-it path, that is your break-even point. For most recurring tasks, this happens within 3-7 instances. For weekly tasks, that is the first 4-8 weeks. After that, every instance is net savings. This is why the "I'll just do it myself" instinct is so dangerous: the cost is invisible in week 2 and enormous in month 12.

Three Scenarios Where You Should NOT Delegate

1. Tasks that happen fewer than 3 times per year. Training cost is fixed. Frequency amortizes it. If you teach someone a process that runs twice annually, the amortization never catches up. Keep it, or batch-train someone as part of a broader cross-training effort (where the cost spreads across multiple tasks).
2. Tasks that require managerial judgment or authority. Performance reviews, salary decisions, firing, representing the team to senior stakeholders, strategic direction. These are not delegation candidates because the authority itself is the value, not the task. Delegating these undermines your role and confuses accountability. See our guide on how to delegate as a new manager for the full decision framework.
3. Tasks where a mistake has outsized downside. Legal filings, board deck numbers, customer escalations for tier-one accounts, anything regulated. The bounded quality cost assumption breaks down when the downside of one mistake is a lawsuit or a churned customer worth $500K. For these, the math is not the question, the risk profile is.

For everything else, which is typically 60-80% of what lives on a manager's calendar, the math is usually decisive and usually in favor of delegating. The instinct to keep things is the #1 cause of new-manager burnout. Our article on time management for new managers goes deeper on this trap.

How to Actually Delegate (Once the Math Is Clear)

Knowing you should delegate and actually doing it are different things. Most managers fail at the handoff, not the decision. The three moves that matter:

1. Teach once, deliberately. Do not explain as you go over the course of three tasks. Block one dedicated training session, walk through the why, the how, the edge cases, the failure modes. Record it if possible. Training as byproduct of doing is 3-5x slower than training as an event.
2. Define the authority level explicitly. "Do it and tell me" is different from "do it and I will check before it ships" which is different from "propose it and I will decide." Most failed delegations failed here. Be specific about where the decision lives and what quality bar triggers escalation.
3. Review against outcomes, not style. The fastest way to kill delegation is to re-do the work yourself "because they did it slightly differently." If the outcome meets the bar, leave it. Style is your preference. Outcomes are the deal.

If you want a complete system for the 12 delegation decisions new managers face most often (decision matrix, authority levels framework, task briefing template, review checkpoints, and a growth-tracking tool), the Delegation Matrix Kit gives you all of it in one PDF.