What to Delegate to an Executive Assistant: The Delegation Matrix

Delegation ·Published October 13, 2026 ·By Kasim Aslam

Ask a hundred founders what they delegate to their executive assistant and you'll hear the same three words: inbox, calendar, travel. Good. Do that. I do it too. I closed Slack on my computer entirely. My Right Hand, Gina, runs it for me and sends me an end-of-day update, unless something is urgent, in which case she texts me.

But if that's where your list ends, you've hired a very expensive filter. You protected your time and then spent it on the same things you were doing before.

You already know you can't delegate well. That's why you searched. So I'm not going to spend this guide convincing you that you're the bottleneck. I've written that one: how to stop being the bottleneck. This guide answers the narrower question: once someone good is sitting next to you, what do you actually hand them?

My answer is a worksheet I teach in the Delegation Mastermind called the Big 3. It takes about 90 minutes, and it ends with three pieces of work that matter more than anything in your inbox.

01 · The Matrix

What should you delegate first? The delegation matrix

Delegate first whatever is low stakes and eats your hours, then anything that waits on you only because the rule lives in your head. High-stakes work with no owner goes to your assistant as an experiment. The only thing you keep is the small set of choices that are truly yours.

Sort every task on your plate by two questions: how much is riding on it, and does it really need you? That gives four boxes.

Most lists stop at box 1. Boxes 2 and 3 are where an assistant stops being a filter and starts moving the business. The green, yellow and red rules for box 2 are in how to stop being the bottleneck. The rest of this guide is about box 3: how to find the work, score it and hand it over.

02 · The Floor

Why isn't "inbox and calendar" enough to delegate?

Because inbox and calendar buy back hours but don't tell anyone what those hours are for. Here's the thesis of everything I teach: get more done without doing it yourself. You're not allowed to do it yourself. The little voice that says this is hard, I'll just do it? That's the thing we're correcting.

But I define "more" as qualitative, not quantitative. More things done is not the goal. Better things done is the goal.

I'll tell you how I learned that. When I ask AI for something, I get too much good stuff. I get so overwhelmed that I stop. And I noticed the same thing happening with my team. Everybody is so well equipped now that they almost do too much, and I can't contend with any of it.

So the question changes. It stops being how do I get more tasks off my plate? It becomes what are the few things that, if someone else moved them, would change the business?

The inbox doesn't answer that question. Inbox triage gives you hours back. It doesn't tell anyone what those hours are for.

03 · The Big 3

Start with your Big 3: rocks, risks and reaches

The Big 3 is the simplest planning tool I own, and I come back to it constantly. Everybody's business is different, but we all have the same three kinds of big thing.

Big rocks. I'm borrowing rocks from Stephen Covey: your biggest, most important things that need to be done. I don't like the word goal here. A goal is a finish line. A rock is the thing that has to move.

Big risks. What could break the business? What would go first if something went badly wrong? Risks are my favorite of the three, and I'll explain why in a minute.

Big reaches. If you actually had the time, the money and the inclination, what would you go after? The moonshot you keep not starting.

Write three of each. That gives you a grid of nine. Don't overthink the wording. A rock can be "close the Series A" or "get the new location profitable." A risk can be "we lose our biggest client" or "a competitor gives away what we charge for." A reach can be "launch a software product" or "open in a second country."

04 · Who Sees It

Put the Big 3 in front of your whole team, not only your assistant

Once you have the nine, show them to your assistant first. Then show them to everyone.

Here's why. When I built Solutions 8, we had 100 employees when we sold. In due diligence I found out how often people were doing work that no longer had any value.

"It was shameful how often SOPs were running that didn't need to be ran, or people were doing initiatives that were superfluous to the exercise. And it's because I didn't do a really good job in all stages at communicating what was most important."

The Big 3 fixes that. When people work on your behalf, they know what you really want. They don't have to guess, and they don't have to wait for you to tell them.

It also gives your assistant a filter they can use against you. More on that near the end.

05 · The Exercise

Turn each of the nine into an experiment, not a solution

This is the most important rule of the whole worksheet, so I'll put it in bold. You are not solving these problems on the sheet. If you could solve them, you would have done it already. You're defining the experiment: what would you play with if you had the time, the money, the people, the data and the access?

For every one of your nine, run it through a short set of questions. These are mine. You might find thirty more that fit your business better, but don't dismiss these out of hand.

For each rock

• What's the constraint preventing this rock from moving faster? Why is it a rock and not a pebble? Why can't you buy your way there, or do your way there, right now?
• What assumptions are you making that might be false? "I know for a fact it's hard and expensive to build an iOS app." That used to be true. Is it still?
• What experiment could bypass that assumption in 14 days? Not solve it. Test it.
• What would progress look like without full completion? How do you move the ball down the field without scoring a goal?

For each risk

• If something really failed, what would go first? What's the first domino, the canary in the coal mine?
• What proof would convince you this risk is under control? What would somebody have to show you?
• What's the cheapest way to generate that proof? One of my favorite questions, because it's fun.
• What experiment reduces dependence on you, or on any single point of failure?

That last one is personal. At Solutions 8, if I lost my head strategist, I lost that business. So I hired a bench of strategists who were really, really smart. If I'd ever lost him, I'd have spent my life convincing clients the others were smarter. That was the experiment that reduced the risk.

For each reach

• If it worked instantly, what steps would have been needed?
• Which of those steps can be skipped, simulated or delegated?
• What's the compressed version? The smaller, littler version that isn't as big a reach.
• Where's the leverage inside the experiment?

When you're done you have your innovation experiment for rock one, rock two, rock three, and the same for each risk and each reach. Nine experiments.

06 · The Inversion

The risk is often the reach

I said risks are my favorite. Here's why: if something is a risk to you, it's a risk to every competitor you have. Figure out how to mitigate it and, in most cases, you've also figured out how to exploit it.

So try this one. Ask yourself:

"What if a competitor in your industry took everything that you're currently charging for and put it in a Google Drive and gave it to everybody for free? How much would that hurt your business? First question. Second question, why don't you do it?"

I've done this more than once. In the Montessori space, we gave schools a library of resources (marketing templates, attendance sheets, health and safety protocols, admissions material), more than they'd expect from buying into a franchise. Free. People in that industry were livid with us.

I'm doing it again with Bully.org, a business I co-own. We hired a Right Hand through Pareto, and he built a tool that generates resources for schools dealing with bullying: playbooks for teachers, parent communication kits, re-entry plans after a suspension. There are thousands of consultants who charge schools for exactly that. Everybody in that industry is going to be livid too. And then guess who everybody signs up with.

The point isn't "give everything away." The point is that one question turned a risk into a reach. And every part of it, from the research to the resource library to the tool itself, was delegated.

07 · The Score

Score the nine: impact, confidence, ease, cost and access

Some of your nine are going to suck. That's fine. That's why we score them.

If you know the ICE model that growth teams use, you'll recognise this. I like ICE a lot, but I think it misses things, so I modified it. Score each experiment from 0 to 5 on five levels.

Access is the one I added, and it's the one most people skip. When I've done ICE scoring in the past, I kept discovering that my scores assumed access I didn't actually have: a partner I hadn't asked, data I couldn't get, a person I hadn't hired. Scoring access sometimes needs a little homework. Do it.

Add the five numbers. The maximum is 25. Circle your top three totals. Park the other six. You now have three active innovations, already in the order you should approach them.

08 · The Handoff

Delegate the experiment, not the solution

Now the strong recommendation, from someone with a vested interest in your success: don't run these yourself. Delegate the innovation to your assistant.

That doesn't mean your assistant has to know how to build the software, launch the community or rebuild the tech stack. What an experiment needs is a quarterback. Your assistant researches what already exists and what doesn't, where the risks are, and what kind of people the project would need. They play with it on your behalf, and they bring you a report card.

Innovation is the best thing you can delegate for a reason most founders miss. There's no SOP yet. No department owns it. You're not stepping on anyone's toes. All the things that make delegation hard inside an established company don't apply.

Here's what it looks like in practice. On a live call, a member asked whether their assistant could rebuild the resource-library process I'd just shown. So I dictated the brief to Gina in front of everyone, the way I'd send it by text on any normal day. Define the audience and get my approval. Define the categories and get my approval. Build the template and get my approval. Then build the thing that creates resources. And find a name for the process, because if you name it, you own it.

"That was 1 minute and 21 seconds. That's as much time as I'll spend on that task."

That's the whole investment. One minute and twenty-one seconds of messy voice brief, and an experiment starts moving. If you want the mechanics of briefing like that, it's in how to delegate without micromanaging: define the inputs, define the outputs, and stay out of the middle.

Gina does this for me all the time. One night I sent her a software application that might help one of my agencies: play with this and tell me how hard or easy it is, and whether it's worth pursuing. When she brings back her report, I'll say "this is really helpful," or "I don't agree with this at all," or "dig deeper here." That back-and-forth, repeated, is how I came to trust her instincts.

09 · The Volume

Hand over more than they can finish, on purpose

Most founders do this backwards. A new assistant starts, and the founder says: take it easy, read the manual for two weeks, watch some videos. That sets terrible expectations and builds terrible habits.

Nobody runs faster than they do in their first 90 days. So on day one, I overwhelm them, and I tell them exactly what I'm doing:

"Your job is to manage me. I do not expect your evenings or your weekends. I don't want to stress you out, I don't want to kill you. And I'm gonna send you as many things as I can send you, and it's up to you to tell me when you're overwhelmed."

The rule is twelve things for a ten-unit day. If I send twelve, there's no expectation that all twelve get done. All I want is: "These are the ten I think are highest priority. I'll get to the other two as soon as I can. Is that okay?" And I say yes.

That's Parkinson's Law turned to your advantage. Work expands into the time you allot, so allot almost none and keep the queue full. Here's everything you're going to do, here's your back burner, and here's your back burner's back burner. Your Big 3 experiments live at the top of that queue. The inbox lives underneath.

One warning. Don't send busywork. I think it diminishes the person's value in their own mind. That's exactly why the rocks, risks and reaches matter: they're really important work, and your assistant can tell.

Already have an assistant and wish you'd started this way? It's not too late. I wrote a short reset email for exactly this: I want to change the way we're doing things. You're not doing anything wrong. A new year, a new quarter, or a Monday works fine as the excuse.

10 · The Return

When the work comes back: the 262 rule

Overwhelm someone properly and some of what comes back will be rough. That's the moment that matters.

Give anyone ten things to do and here is what happens. Two of them they'll be bad at. Six they'll do well. Check the box, well done. And two they'll do better than any human you've ever met.

The business world trains us to focus on the bad two. Performance improvement plans. Shadow Jerry. Instead, unless those two are a core part of the job, take them off the list. Then build the role around the brilliant two.

This is how my own assistants grew. My first became my director of social. My second became my director of automations. My third became my CTO, managed an eight-figure exit, and is now my business partner. None of them got there because I fixed their weak two. They got there because I fed their strong two.

And when the first imperfect thing lands, look for the part they did well and say so. If you lead with "this sucks," you've built a Pavlovian relationship where nothing ever makes you happy, and they stop trying.

11 · The Urge

Why you'll want to edit it, and why you shouldn't

On one of our calls, my business partner described a feeling I'd bet you know. Someone hands him a finished deck, and he feels he must ask for a change. Not because the deck needs it. Because if he doesn't comment, he feels like he's slacking off.

Psychology has a name for this, and it's actually two names. Jeffrey Pfeffer, Robert Cialdini and colleagues described them in a 1998 study of why managers don't empower workers:

• Self-enhancement bias: we rate a piece of work more highly the more involved we were in producing it.
• Faith in supervision: we believe work done under supervision is better than work done without it.

Both make you reach for the red pen. And the cost is bigger than wasted time. The minute you say "change the red to blue and the circle to a square," your assistant takes their hand off the wheel. If the thing fails, it was you driving. If it succeeds, it was never theirs. That's what I mean by authorship is ownership.

My partner's filter is the best one I've heard: put a dollar amount on every change. Will moving this image ten pixels to the right increase the company's bank account? Once you ask that, most of the changes stop mattering.

Critique what's objectively wrong. The studs have to be 18 inches apart because the inspector will fail the building otherwise? Say so. Anything that's taste, let go.

12 · The Pushback

Teach your assistant to push back, using the Big 3

Founders are shiny-object machines. I've started full businesses because of one conversation. I once spent far too much money and time trying to crack newsletters because someone I admire was talking about them.

The Big 3 gives your assistant a polite way to stop you. A smart assistant who knows your rocks, risks and reaches will say: can you help me understand how this new idea relates to the big three? Or better: here's everything I'm working on. Which of these important things do you want me to drop so I can chase this for you?

You have to reward that. When someone disagrees with you, it's easy to bristle and put the CEO hat back on. Train yourself to say "thank you for being contrarian" instead. One of my favorite lines: if two people agree on everything, one of them is unnecessary.

My third assistant, the one who's now my partner, lectured me more times than I can count. He was usually right, and I usually deferred to him. It's the reason he's my business partner today.

13 · The Short Version

So what should you delegate to an executive assistant first?

Here's the order I'd use, starting this week.

1. The admin floor. Inbox, calendar, Slack, scheduling, travel, follow-ups. Delegate it now, messy, without a perfect brief. This buys back hours.
2. Your Big 3. Write three rocks, three risks, three reaches. Share them with your assistant and your whole team.
3. Nine experiments. Run each item through the questions above, or hand your assistant the prompt and let them draft them.
4. The top three. Score impact, confidence, ease, cost and access from 0 to 5. Keep the three highest totals.
5. The quarterback brief. Send each of the three as a goal, by voice if you like. Ask, don't tell: when do you think you can have this? If they say next month, try "I'm thinking Friday. Don't kill yourself, just do the best you can by then."
6. The return. React to the report card. Find what's good, critique only what's objectively wrong, and index toward their brilliant two.

Do that and your assistant stops being a time-saver and starts being the person who runs the business's most important experiments. Which brings us back to the bigger question of how to delegate as a founder: it was never about the tasks. It was about who owns the thinking.

Frequently asked questions

Sources

Sources

• Kasim Aslam, Delegation Mastermind session "Big 3 Innovation Framework", January 9, 2026: the Big 3 worksheet, the five scoring levels, twelve things for a ten-unit day, the 262 rule, and the quotes on this page.
• Jeffrey Pfeffer, Robert Cialdini, Benjamin Hanna and Kathleen Knopoff, "Faith in Supervision and the Self-Enhancement Bias: Two Psychological Reasons Why Managers Don't Empower Workers", Stanford GSB Working Paper No. 1432, 1997 (published in Basic and Applied Social Psychology, 1998), opened October 12, 2026.
• Stephen R. Covey, A. Roger Merrill and Rebecca R. Merrill, First Things First (Simon & Schuster, 1994): the "big rocks" idea.
• Pareto Talent: 100+ founders served, 40+ hours of training per Right Hand, first candidate profiles within 24 hours (company records, October 2026).

Written by Kasim Aslam from his own experience and the Delegation Mastermind. Reviewed and updated quarterly.