tech
Delegation
Transferring authority and accountability, not just tasks.
Definition
Most founders 'delegate' by handing off tasks while keeping decision authority - which isn't delegation, it's outsourcing execution while remaining the bottleneck. Real delegation transfers the decision: define the rules, define the goal, let the person make the call within them. The standard test: if the team still has to ask you, you didn't delegate. The hardest part is accepting 85% quality for 100% leverage - founders who insist on 100% quality stay bottlenecked forever.
The five levels of delegation
Adapted from Jim Collins and used widely in US executive coaching, the five levels of delegation form a useful self-diagnostic. Level 1: do exactly what I say (task execution, no judgment). Level 2: research and report back so I decide (information gathering). Level 3: recommend, then I decide (analysis plus options). Level 4: decide and inform me before acting (autonomous with notification). Level 5: decide and act, report only on exceptions (full authority). Most US founders default to levels 1 and 2 on everything, then complain that the team cannot operate without them. The path to leverage is moving each role progressively to level 4 or 5 within defined decision rights. Map each direct report by responsibility area and current level; identify two areas per quarter to move up one level.
Defining decision rights in writing
Verbal delegation fails because both parties remember different things three weeks later. Written decision-rights documents (sometimes called RACI charts or decision logs in US practice) eliminate ambiguity. Format: list 15 to 30 recurring decisions in the business, name the person who decides each, the threshold above which it escalates, and the people who must be informed. Examples: 'refunds under 500 dollars - support lead decides, no approval needed.' 'Marketing spend under 5K per month - marketing lead decides.' 'New hires under 100K salary - hiring manager plus founder approval.' Build this document during a 90-minute leadership offsite, review it quarterly. The act of writing it forces clarity that conversation never achieves.
Coaching decision-making versus task execution
Founders who delegate by reviewing every output stay bottlenecked. Founders who delegate by reviewing decision-making process scale. The shift: instead of editing the deliverable, ask 'what tradeoffs did you consider' and 'what would change your recommendation.' This reveals whether the person reasoned well, independent of whether the final output matches what you would have done. Over 6 to 12 months of this coaching pattern, the team's judgment compounds. The hardest part for high-performing founders is accepting that 85 percent of their judgment, applied by someone else, beats 100 percent of their judgment applied to a fraction of the decisions. The math is overwhelming but emotionally hard.
The reverse-delegation trap
Reverse delegation is when work you delegated comes back to you. It happens through three mechanisms. One, the team member asks for input on a decision they own ('what would you do here?'), founder answers, decision is now founder's. Two, the team member presents a problem instead of a recommendation; founder solves it; pattern repeats. Three, founder reviews work, finds an issue, fixes it themselves instead of returning it. The disciplines that prevent reverse delegation: never accept a problem without a recommendation, never edit work yourself if it can be sent back, and answer 'what would you do?' with 'what would you do, and why?' US executive coaches call this 'no-monkey discipline' from the 1974 HBR article 'Who's Got the Monkey?' - still the cleanest framing.
FAQ
I tried to delegate but the result was bad. Should I take it back?
Almost never. Taking work back teaches the team that delegation is temporary and ends at the first failure, which means future delegation never sticks. Better path: do a debrief on what went wrong (was the goal unclear, were the decision rights ambiguous, did they lack a skill), fix the root cause, and re-delegate. The exception: if the failure is catastrophic (financial, legal, customer-facing emergency), pull the work back temporarily but recommit to delegating it again within 30 days. Permanent reclamation kills delegation culture across the whole team.
How do I delegate without losing quality?
Define quality explicitly before delegating. Most quality problems in delegation come from the founder having an unstated standard the team cannot see. Document: what does great look like (with examples), what is acceptable, what is unacceptable, what is the minimum quality bar to ship. Use checklists, templates, and reviewable samples. Then accept that the team will land at 80 to 90 percent of your standard initially, improving to 95 percent over 6 to 12 months. The 5 percent gap is the cost of leverage; trying to close it eliminates the leverage entirely.
How do I know if I am over-delegating versus under-delegating?
Under-delegating signs: you work more than 50 hours per week, you are involved in tactical decisions, your calendar is full of approvals and reviews, the team waits on you to move. Over-delegating signs: you cannot answer basic questions about how the business is performing, customers report inconsistent experience, decisions get made you would not have approved, accountability is unclear. The middle path: delegate execution and decision authority, retain strategic direction and final exception handling. Most US founders under 5M revenue are 70 percent under-delegated; founders over 10M revenue start over-delegating and lose feel for the business.
What are the highest-leverage things to delegate first?
Order by leverage. One, recurring administrative tasks (scheduling, expense reports, email triage) - delegate to an EA or virtual assistant for 30 to 60 dollars per hour, frees 5 to 10 founder hours per week. Two, deliverable production where you are not the differentiator - delegate to junior or contractor team. Three, hiring screening - delegate to a hiring manager with clear rubrics. Four, vendor management. Five, marketing execution. Keep for last: pricing decisions, strategic direction, top-10 customer relationships, and brand voice. The 80/20 of leverage usually comes from the first two categories.
Should I delegate to employees or contractors first?
Depends on the work. Project-based or specialized work delegates well to US 1099 contractors or agencies (design, accounting, legal, specialized marketing). Recurring operational work that requires institutional knowledge delegates better to W-2 employees who stay long-term. Start with contractors to test the function and identify the right structure; convert to employee role once the work is steady and the right hire emerges. Many US small businesses use a fractional executive model (fractional CFO, CMO, COO) as the bridge: 1099 contractor doing executive-level work 5 to 20 hours per week until volume justifies a full-time employee.
In your business
- →Define decision rules in advance (e.g., 'refunds under $500 auto-approved')
- →Accept 85% quality for 100% leverage
- →Coach decision-making, not task execution