Your best engineer just quit. Not for money — she told you in the exit conversation that she spent six months chasing you for decisions you never made. You were too busy approving Slack messages about things that, honestly, you shouldn't have been approving at all.
That sentence stings because it's common. Most founders I talk to don't have a delegation problem — they have a throughput problem disguised as one. They think the answer is working harder, or hiring better, or buying another tool. It isn't. The answer is a set of thresholds and rules that let you let go without losing the plot.
Below is the framework I actually use, built from the mistakes I made over several years of running a small team, plus the rules that finally made delegating effectively as a founder feel less like a leap of faith.
Key takeaways
- Delegation fails when the brief is vague, not when the person is wrong.
- The 70% rule: if someone can do a task at 70% of your quality, hand it over and coach the rest.
- You keep three things: vision, capital allocation, and hiring for the first five roles. Everything else is negotiable.
- A decision without a deadline, an owner, and a "come back if X" trigger is not a delegation — it's a wish.
- The 80/20 rule of delegation means 20% of the tasks you hand off will free up 80% of your thinking time. Find those 20% first.
Why founders struggle with delegation (hint: it's not control)
Ask a founder why they don't delegate and you'll hear "trust." I don't buy it. In my experience, the real blocker is unclear standards. You can't let go of a task whose definition of "done" only exists in your head.
I made this mistake with our first content hire. I handed over a weekly newsletter with the brief "make it good." Three weeks later I was rewriting everything at 11pm, convinced I'd hired badly. I hadn't. I'd given her a task without a spec, then punished her for not reading my mind.
The real cost of keeping everything
Here's the part nobody warns you about: the cost isn't your time. It's your team's decision-making muscle. Every time you override a call, you teach your team to stop making calls. Two quarters of that and you'll be the bottleneck for a task as small as picking a font for the deck.
Look at it this way. A founder who personally handles 40 decisions a week isn't more productive — they're just doing the work of four people badly while being unavailable for the one thing only they can do: deciding where the company goes next.
What is the 70% rule of delegation?
The 70% rule says: if someone on your team can complete a task at around 70% of the quality you would deliver, delegate it — and coach the remaining 30% over time. Waiting for 100% is a trap, because it never arrives.
Why 70%? Because the gap between 70% and 100% is almost never the person's ceiling. It's information you haven't shared, context you haven't written down, or a standard you haven't articulated. Those are fixable. What isn't fixable is a founder who becomes the permanent quality control layer for every piece of work.
When 70% is not enough
Not everything qualifies. I'd never apply the 70% rule to:
- Legal contracts, especially anything touching equity
- Anything that leaves the building under your name, like a press statement
- The first version of your pricing — this one sets the whole commercial logic
Everything else? Ship at 70% and iterate. A blog post at 70% that goes live this week beats a perfect one sitting in drafts for a month.
What are the 5 principles for effective delegation?
Strip away the theory and five principles do most of the work. They're boring. They work.
1. Define the outcome, not the method
Tell people what "done" looks like and let them choose the path. If you specify both, you haven't delegated — you've outsourced your hands while keeping your brain attached to the task.
2. Match authority to responsibility
If someone owns a decision, they need the budget and access to make it. Responsibility without authority is how you create a team of people who ask permission for everything.
3. Write the brief
Every delegation I've botched had a verbal brief. Every one that worked had two paragraphs in writing: what success looks like, and what "come back to me" means.
4. Agree on the check-in cadence upfront
Weekly sync? Async update on Fridays? Whatever it is, define it before work starts. Ad-hoc check-ins signal distrust and quietly destroy morale.
5. Review the decision, not the work
When something goes wrong, debrief the reasoning that led there — not the output. People learn from decisions. They only get defensive about output.
What are the 7 steps of effective delegation?
This is the sequence I follow, in order. Skipping steps 2 or 3 is what turns delegation into rework.
- Name the task in a single sentence. If you can't, you don't understand it well enough to hand it off.
- Estimate the time it should take. This becomes your benchmark, not your whip.
- Pick the person — not the one who's free, but the one who'll grow the most from doing it.
- Write the brief: the outcome, the constraints, the deadline, and the escalation trigger.
- Hand over the context, not just the task. Give them the history, the previous attempts, and the reason it matters.
- Set the check-in and stick to it. Do not check in early "just to see."
- Debrief once it's done. What worked, what didn't, what changes for next time.
Step 5 is where most people shortchange it. Context takes ten minutes to write and saves you a week of back-and-forth. I started keeping a one-page "context doc" for recurring tasks, and it cut the number of clarification messages per delegation roughly in half.
What does the 80/20 rule of delegation mean?
The 80/20 rule of delegation means roughly 20% of the tasks you hand off will free up about 80% of your thinking time. The rest are administrative — nice to hand off, but they don't buy you back the hours you need for strategy.
The trick is identifying that 20% before you start handing things around randomly.
How to find your twenty percent
Track your time for one week. At the end, sort every task into two columns: only I can do this and I do this because I always have. The second column is your list. In my own tracking, three items showed up again and again: reviewing marketing copy, running weekly reports, and answering partner emails. Those three were easily a third of my week — and none of them actually needed me.
What to keep and what to let go
There's no universal answer here, but a rough map helps. Here's how I split it:
| Keep | Delegate | Judgement call |
|---|---|---|
| Vision and narrative | Reporting and admin | Hiring the first five roles |
| Fundraising and investor relations | Vendor negotiations under a set threshold | Pricing changes |
| Final call on team structure | Day-to-day customer support | Content strategy |
| Anything touching the founding story | Scheduling and logistics | Partnerships under a certain size |
The "judgement call" column is where most founders get stuck — and where a written threshold does the most good. For example: "Delegate anything under €5,000 per month. Come to me above that." A number beats a vibe every time.
The failure modes nobody talks about
You'll get a lot of the usual advice about delegation — trust your team, let go, etc. That's fine, but here are the failure modes that actually cost me a quarter of momentum:
- Delegating and then re-doing it. Guaranteed to teach your team that their work will be overwritten anyway.
- Delegating without a feedback loop. Without a check-in, you learn about problems only when it's too late to fix them.
- Delegating the wrong task. Handing off a task that requires strategic judgment, then being surprised the output is tactical.
- Delegating everything at once. Too much at once and you lose visibility into what's actually happening.
My own worst version of this was handing off our entire customer onboarding to a junior hire during a growth spike. Six weeks later, churn was up. Nobody told me because nobody had asked for a check-in. That's on me, not on them.
The mindset shift that makes it stick
Delegation isn't a productivity hack. It's how you build a company that doesn't collapse the day you take a week off. The founders I know who delegate well aren't more trusting by nature — they've just made a decision, in writing, about what they keep and what they hand off, and they revisit it every six months.
If you take one thing from this article, make it this: pick one recurring task this week that someone else could do at 70%, write a two-paragraph brief, and hand it over with a deadline. Then resist the urge to fix it.
The next time your best engineer considers quitting, you'll want to be available for the conversation that actually matters — not stuck approving Slack messages about fonts.