#26 – Executive Assistants and Why They Are a Force Multiplier for Our Team | Ben Grynol, Sam Corcos & Michael Mizrahi

Executive Assistants (EAs) act as a force multiplier for a company’s productivity. They take on necessary repetitive tasks so that leadership can spend more time on higher priority work. In this episode, Levels Head of Growth Ben Grynol sat down with co-founder and CEO Sam Corcos and Head of Operations Michael Mizrahi to discuss how to best outsource work to EAs, why it’s important, and how to decide what to delegate and what to keep on your plate.

06:47 – Reluctance to delegate

Sam said that many people struggle with giving up tasks because it feels like scrimping on work.

The problem is there are some default things that you just assume an EA does or oh, they handle scheduling, but I’ve actually found those to be the least effective tasks for an EA. Those are much better handled with something like Calendly or automated tooling. What people really struggle with is I think it’s the ambiguity and the discomfort. It’s on a muscle that most people flex on a regular basis. It can feel like it doesn’t feel like are doing the work to get something done. It’s sort of like a lot of programmers that I know, when I watch their workflow, they don’t have any snippets. They don’t use hot keys. And they know how to make something work. But when I say to them like, “Hey, I noticed you’re doing all this stuff in React, you could create a snippet in VS Code that pre-generates this template every time you do it. I noticed you’ve already done it 10 times today for a component. You can just make this a snippet. It’ll take you probably three to five minutes to set it up. And every time it saves you five minutes.” And there’s a reluctance to do that for a lot of people, because it doesn’t feel like doing the work.

08:52 – Make time for bigger work

Michael said that removing repetitive tasks can feel uncomfortable at first, but it allows you to start spending more time on the heavier tasks.

There’s an interesting similarity here to some of the sensation that people get when you take away Slack and you take away meetings, and all of a sudden, your day is wide open. And you have that moment where you hit the pause and you’re like, “Okay, what do I do now?” And it’s like, well, you do the work. This is the part where you get to think and strategize and do the heavy lifting. And so there’s that blink space. And there’s something similar about delegating work out, you get this feeling like, “This is what I’m supposed to be doing. If I delegate this out, then what do I do? This is my job. Why would I give this to someone else? Sure. Someone else could potentially do it, but I’ve got the knowledge, I’ve got the context, and so I’m the best person to keep connecting all these dots.” But only when you delegate it out and you get it off your plate and you create all this extra space, can you move on to the other higher priority or different priorities and just kind of keep moving and keep adding value across the company. And so there’s something interesting there about that feeling of leading into the discomfort in your day, in your work, in delegation so that you can make room for other things.

13:32 – Have automation and delegation

Sam said anything that is repetitive but can’t be automated is a great thing to hand to EAs.

It’s really important to recognize the value of automation. This is maybe a semantic distinction, but I’m talking about automation versus delegation. There’s a lot of what the EAs do for me. Certainly the highest leverage things for me would really fall into the automation camp. They are things that it would be better if a computer could just do it, but there is no software or API that solves this problem. And so they can fill that gap. I posted a video on Twitter a year ago, maybe more than that. It was during the seed round process of how I send emails the day before for every call, how I put them all in Notion so I keep track how I do all the follow ups. That was one of the first tasks that I handed off to an EA. Is, “Figure out who these people are in my calendar, who I have calls with in the future, send them this email, create a Notion template so that when I jump in for the day, I have my list of calls. I see who they are. I see the context and I can quickly get up to speed on it.” So those are the kinds of things where they’re all happening in the background right now. And in the past, I was doing them. It took me 30 to 60 minutes every day just to prepare for my calls the following day. And now it takes me zero time.

15:29 – De-risk delegation

Sam said there are ways to reduce the risk of handing off tasks. Documenting your process or doing it concurrently with EA the first couple of times can do that.

But really the key to getting better at delegating is to figure out how to de-risk the request. So reducing the scope of what it is you need done and increasing visibility where you can now build confidence that they’re doing it consistently and they’re doing it correctly. Or doing it completely in parallel, where you’re still doing the thing you were doing anyway, but you have the EA tried to do it as well. And if it doesn’t work, then you didn’t lose anything. Just like, “Okay, well, I guess this is too complicated. They can’t do it.” So figuring out how to reduce the stakes to just try to get those cycles in to get an EA to do it, I think is really compelling. I think another thing to tie into that is people often may too many assumptions about how much effort it takes. I have found for me the overwhelming majority of the time, if I’m doing a thing that I want to hand off to an EA, all I have to do is record a Loom of my workflow doing the thing. And then I just say, all right, you just watch me do this for 30 minutes. Can you repeat back to me what you think the process should be? And then if that works, we’ll go from there.

16:55 – Document everything

Michael said that having thorough process documentation allows for easier transfer of knowledge and responsibility.

Loom is one of the most undervalued business tools of the last decade. It is infinitely more powerful than Slack or some of these other tools. Looking through, we’ve got this database of all of our tasks, our process database in our Wiki and Notion. I didn’t write any of these process docs for a few dozen of these tasks. I recorded the Loom doing the thing, maybe said some things along the way. And then Vanessa, who I work with has built all of these into templates and processes. And so the redundancy there is also great, because if someone else needs to step in, if Vanessa’s out on vacation, the knowledge doesn’t live with me and it doesn’t live with Vanessa. We’ve got plenty of documentation and process and Loom’s recorded for each and every time this task gets done. And so that’s helpful for debugging when something goes wrong, it’s helpful for training moving forward if someone else needs to pick up the tasks.

22:49 – Hire support staff early

Sam said that bringing on support staff is always the right decision and should be done earlier rather than later.

Yeah, I think the hiring support staff is something that people often have a struggle justifying, but it is almost always the right decision. If you can build a structure to be able to use that time effectively, when you calculate the difference in time, if it takes Mike Haney, our editorial director, 15 minutes a week to put together some slides to update people. If that 15 minutes could be done automatically by the EA team. That 15 minutes, we also forget how much of a load, just that cognitive burden of knowing that you have to do it costs. It’s much more than 15 minutes. It’s probably an hour of cognitive load to be able to finish that. So I think hiring a support staff as early as humanly possible is almost always the right call.

24:12 – Add in accountability

Sam said that having a system of keeping track of the work is essential for trust. Levels asks EAs to record themselves on Loom doing their tasks.

We ask the EAs to record themselves on a Loom doing the work so that if we need to debug something later, that’s part of it. But it’s also just to build the confidence, to just know that things are being done and that you can close the loop. There have been many times we have a database of every task that has been completed and a walkthrough of that task. And every once in a while, I’ll ask myself, “Hey, I wonder if that thing is still being done that I asked for.” And I go to the database and it’s like, “Oh yeah, they did it yesterday.” And I can’t even watch the recording of her doing it. And it’s like, “Yeah. Perfect. Okay, great.” I now have that trust and confidence there. So it’s super, super important to find mechanisms for building that trust. Because when people are in other countries, you barely interact with them at all. You need to really know that things are being followed through on.

30:01 – Divide process work and critical thinking work

Michael said that being aware of when you’re doing process-oriented work can help you to think ahead and document what you’re doing for others.

I’m thinking back to when I onboarded and was joining, Sam, you were handling a lot of the operations tasks at the time in the months preceding. And you had recorded Looms of a lot of these without any foresight of what they might be used for. But when I came on board and started watching those videos, my onboarding was supercharged because I didn’t need to ask anyone for help or have context on how a task was done. And even if there was a 30-minute Loom, I could watch that and get up to speed on what I needed to do to now own that task without even you knowing or having to get involved. And so there’s a lot of lessons to apply, even if you’re not working with exec assistant, when it just comes to being really, really aware of the work that you’re doing and mindful of when you’re actually doing process-oriented work, versus when you’re doing critical thinking and putting down a memo and strategy. There’s a big difference between those two, those categories of work.

50:38 – See what they can take off your plate

Sam said that one way to figure out what to delegate is to show an EA what all you do and see what they think they can take over.

I would say another is, like I mentioned, just reducing the assumptions around how much work it takes to hand something off. Something you very realistically could do is just open up a Loom when you wake up in the morning and just record your entire workflow for the entire day on a big Loom. And just give that to your EA and say, “Is there anything in here that you think you could take off my plate?” And just see what they come back with. Worst case, they come back with nothing. Best case, they say, “All right, here are 25 things that I propose I can start doing for you that you won’t have to do anymore.” Allow yourself to be surprised by just how capable people are.

54:37 – Audit communications

Sam said that one of the major things Levels has EAs do is make sure that important documents or conversations aren’t staying in private threads.

I very much view the Athena team as they’re my biggest allies in fighting organizational entropy. I use them all the time. One of the things that we do, which it’s a subtle thing, but it’s so helpful, is once a week, they audit our leadership channel in Threads. Which is just for people in leadership and it’s not shared externally to the rest of the team. And they check once a week and there’s only Usually, I don’t know, 5 or 10 Threads in there. They will reach out to every person and they’ll say, “All right, these are the 10 things that came up in leadership this week. Can we declassify this and share it with the rest of the team?” And almost every week, we find two or three things where we say, “Oh yeah, let’s move this one to general, let’s move this one to marketing, let’s move this one to product.” And we find ways of just fighting that over-classification problem, which is the end state where everything ends up if you’re not paying attention. Is everything is classified and only two or three people get to see any piece of information. They also do this for all of our Notion docs as well.

Michael Mizrahi (00:00:06):

EA is in the traditional co-located office environment, where traditionally is assigned to an exec or an exec team. And they book travel, the conference rooms, the schedule jockey. And so this exec assistant was kind of like your backs up to handle all those things, but not involved in the work. And so I think the way that we use it in the Levels context, where really delegating processes and tasks that we would otherwise spend our own time doing. And that’s where it adds a ton of value in our context. And so I think if we look at the tasks across most of the team, I think we all have the time and the bandwidth and the patients to work on our own personal tasks. It’s for really the work related minutia that gets taken off the plate. And that makes a ton of space.

Ben Grynol (00:00:52):

I’m Ben Grynol, part of the early startup team here at Levels. We’re building tech that helps people to understand their metabolic health. And this is your front row seat to everything we do. This is a whole new level. There is no better way of scaling your time than working with other people, working with team members that you can delegate tasks to. And these are not just recurring tasks, not things that are prescriptive, that there’s only one way of doing. The way of really scaling your time is having people who can help to make decisions and use discretion where needed. In some cases, having somebody to scale your time is as simple as delegation and trust.

Ben Grynol (00:01:43):

Well, for the Levels team, we’ve been working a lot with EA, Executive Assistants, and we’ve got this team from a company called Athena, a huge network of EA that we work with to scale our time. And they’re very much a force multiplier in how we can continue to be more productive. So Sam Corcos, Co-founder and CEO of Levels meets Michael Mizrahi, Head of Operations and I. The three of us sat down and we talked about this idea of working with EAs. How do you flip the heuristic on its head, that heuristic being well, I’m going to outsource things that are transactional, that are recurring, that are very prescriptive and easy to follow two EAs.

Ben Grynol (00:02:19):

How do you flip that on its head and say, “What does it look like when you start to work with EAs and them some discretion, allow them to do work that’s a lot harder, or maybe non-traditional from a typical EA perspective? Does it actually work? Are there tasks that they can take on that help to scale your time to be more effective as a team member and as an individual contributor?” So we dug into this idea of working with the EAs and went through all these different scenarios, the way that we could currently work with EAs, why they’re important and how other people can think about decision making when it comes to outsourcing their time and tasks? It was a great conversation. Here’s where we kick things off.

Michael Mizrahi (00:03:01):

So we’re talking about using Executive Assistants at our company, and even starting with that, even the framing of Executive Assistants is already wrong. I don’t think that’s the way we use this group of people and kind of have added them to our team. So what we’re describing here is adding contractors to our team who can help us with tasks related to our day to day jobs. This started with Sam reaching out to this group, Athena and this company, Athena that has a bunch of executive assistants that they pair up with leaders and tech execs across different companies. But what we’ve done over time has really made this a resource available to the entire team. And so anyone and everyone is to regularly delegate their tasks to one of the Athena EAs that are on our team.

Michael Mizrahi (00:03:43):

And they’re really embedded. They are in our tools, they work with us, they understand the cast of characters, the context, and they’re able to really take a bunch of the work that we’ve got on our plates and help us with it. And what this does, really just going right to it, what this does is it really helps us as individuals focus on where we can add meaningful value and really do work that creates leverage for the company. Instead of get caught up in a lot of day to day tasks and minutia that ends up slowing us down.

Michael Mizrahi (00:04:12):

And so we can set up processes, think through what we want to achieve, and then hand it off to someone that can execute it and run it on a regular basis. And so takes care of recurring tasks, takes care of a lot of nitty gritty admin tasks. And in fact, we’ve been quite surprised at the level of kind of work product and quality that we can get with this team. And so I think we can spend some time talking about all of that, explore what it’s done for us and kind of where we’re going, because this team is growing. I think it’s a team of about 10 right now that supports our team of 50. So it’s been interesting to see that growth.

Ben Grynol (00:04:43):

Yeah. One of the things that there are all these principles, like how to use EAs effectively, but I think there’s a foundation. It might even be worth rewinding all the way to Sam’s told the story before of how he first started working with Lori as an executive assistant. But the question to pose and Sam can dig into it, the question to pose is, when people find out that we’re working with EAs or they ask, they’re like, “Well, I don’t really know what to delegate,” and there’s this, like, I don’t know if it’s a mental hurdle, but the question is like, why do people in general find it so hard to find something to delegate when it’s actually pretty easy once you get going? So I don’t know. Sam, maybe rewinding of the story, because it’s such a good story of how you just started working with Lori and she’s been around for a decade now.

Sam Corcos (00:05:30):

Yeah. Lori’s been my assistant for, I think about nine years now. It’s kind of hard to believe. And I think this was after reading a book, I think it was after reading the four-hour work week and Tim Ferris in the book mentions how delegation takes practice and it’s really important. And so I posted an ad on Craigslist where I was looking for somebody to delegate things to, and I didn’t actually have anything for her to do. I gave myself the challenge of finding something to delegate. So basically once I hired her, I had to figure out how do I take some of these things that I’m working on now and hand them off to somebody else? And it took a lot of practice. It did not come naturally to me. And over the course of working with Lori, realizing when she would fail at something, it was almost always, because I didn’t give her enough context or I made too many assumptions.

Sam Corcos (00:06:33):

So when you work with an assistant, an EAs, a personal assistant, however you want to define the term, the challenge for most people, I’ve talked to a lot of friends who have tried working with assistants. And the problem is there are some default things that you just assume an EA does or oh, they handle scheduling, but I’ve actually found those to be the least effective tasks for an EA. Those are much better handled with something like Calendly or automated tooling. What people really struggle with is I think it’s the ambiguity and the discomfort. It’s on a muscle that most people flex on a regular basis. It can feel like it doesn’t feel like are doing the work to get something done. It’s sort of like a lot of programmers that I know, when I watch their workflow, they don’t have any snippets.

Sam Corcos (00:07:33):

They don’t use hot keys. And they know how to make something work. But when I say to them like, “Hey, I noticed you’re doing all this stuff and react, you could create a snippet in VS Code that pre-generates this template every time you do it. I noticed you’ve already done it 10 times today for a component. You can just make this a snippet. It’ll take you probably three to five minutes to set it up. And every time it saves you five minutes.” And there’s a reluctance to do that for a lot of people, because it doesn’t feel like doing the work. It’s sort of like work that enables work. I think, there’s some similar feeling in the reluctance there where spending 10 minutes defining the process to hand it off to somebody else. And they might not get the right the first time.

Sam Corcos (00:08:21):

I think it’s because there’s some ambiguity that it might take another iteration. It might take 20 minutes. It might take 30 minutes to hand it off. I think that’s what leads people to not delegate more things. It’s also, the practice just comes from knowing what you can delegate. I don’t think there’s a shortcut other than just keep trying different things. And sometimes, the result comes back. It’s like, “Okay, I guess I have to keep doing that.” Or, “Somebody else has to do that.” So there’s no shortcut, but just practicing as much as you can.

Michael Mizrahi (00:08:52):

There’s an interesting similarity here to some of the sensation that people get when you take away Slack and you take away meetings, and all of a sudden, your day is wide open. And you have that moment where you hit the pause and you’re like, “Okay, what do I do now?” And it’s like, well, you do the work. This is the part where you get to think and strategize and do the heavy lifting. And so there’s that blink space. And there’s something similar about delegating work out, you get this feeling like, “This is what I’m supposed to be doing. If I delegate this out, then what do I do? This is my job. Why would I give this to someone else? Sure. Someone else could potentially do it, but I’ve got the knowledge, I’ve got the context, and so I’m the best person to keep connecting all these dots.”

Michael Mizrahi (00:09:32):

But only when you delegate it out and you get it off your plate and you create all this extra space, can you move on to the other higher priority or different priorities and just kind of keep moving and keep adding value across the company. And so there’s something interesting there about that feeling of leading into the discomfort in your day, in your work, in delegation so that you can make room for other things.

Ben Grynol (00:09:54):

In general, I think one of the misconceptions is, because we’re human in every way and we default to this hyperbolic discounting of thinking these short term, like three-minute tasks are not that expensive. We’re like, “Well, it’s only three minutes,” but it’s not the three minutes. If the task comes up, like let’s make an assumption right now. You’ve got a task. It comes up in your brain. You put it wherever it goes, whether it’s on your calendar on some notepad, it doesn’t matter how you categorize that. And then you think, “Okay, I have to do this thing.” Carrying the cognitive load of like at 4:00 PM or the next day, or sometime, this three-minute thing needs to get done. It doesn’t take many three-minute things of nevermind the actual end of three minutes. It’s like, that is probably 29 minutes of total work.

Ben Grynol (00:10:49):

there’s probably some like 10 X Factor applied to this thing to get done that is like, oh, upload welcome video for someone. It’s just something so transactional. Download a Zoom video, put it in Loom and redistribute on Threads. These are all things that executive assistants become a supercharger, to use the car analogy. They really supercharge the engine to scale your time. And it doesn’t take that many of these small tasks. Nevermind meaty and meaningful and valuable tasks. It doesn’t take that many of those to go, “Cool. You’re actually not doing valuable work right now. You’re wasting time.” And your time’s worth an engineer will call it $10,000 an hour. It’s worth so much money that three minutes extrapolated over the course of a week is so expensive for a company. And that’s something that can really help anybody. It doesn’t matter whether it’s personal or work-related task. It’s just EAs are a superpower.

Michael Mizrahi (00:11:43):

I’ve personally failed with EAs in the past. And so this is my first time actually being successful with it. And I have two examples. One was in some of my busiest days at Uber, my personal life was kind of getting set to the side. And so a friend of mine who was way more important than I was, had an EA who had some extra hours to share. And so I jumped on and took half of this time for a month. And I figured I could move along a lot of my personal tasks. This person didn’t have access to my work email or my work calendar. And so it was very much a personal list thing. And very quickly realized these projects didn’t add up and they were way more difficult to delegate than I thought, because there was no form output that I needed. There were just things that I needed to move along.

Michael Mizrahi (00:12:27):

And then anytime it got to travel or calendar management, that was so intertwined with work, that it just didn’t make sense. And so quickly fizzled out. And then the other example is I think EA is in the traditional co-located office environment where traditionally is assigned to an exec or an exec team. And they book the travel, the conference rooms, the schedule jockey, even move things around and running meetings meeting and got to get lunch and need a room, because rooms are the hot commodity in offices. At least, they used to be.

Michael Mizrahi (00:12:58):

And so this executive assistant was kind of like your backs up to handle all those things, but not involved in the work. And so I think the way that we use it in the Levels context where we’re really delegating processes and tasks that we would otherwise spend our own time doing. And that’s where it adds a ton of value in our context. And so I think if we look at the tasks across most of the team, I think we all have the time and the bandwidth and the patience to work on our own personal tasks. It’s really the work related minutia that gets taken off the plate. And that makes a ton of space.

Sam Corcos (00:13:32):

Yeah, it’s really important to recognize the value of automation. This is maybe a semantic distinction, but I’m talking about automation versus delegation. There’s a lot of what the EAs do for me. Certainly the highest leverage things for me would really fall into the automation camp. They are things that it would be better if a computer could just do it, but there is no software or API that solves this problem. And so they can fill that gap. I posted a video on Twitter a year ago, maybe more than that. It was during the seed round process of how I send emails the day before for every call, how I put them all in Notion so I keep track how I do all the follow ups. That was one of the first tasks that I handed off to an EAs. Is, “Figure out who these people are in my calendar, who I have calls with in the future, send them this email, create a Notion template.”

Sam Corcos (00:14:30):

So that when I jump in for the day, I have my list of calls. I see who they are. I see the context and I can quickly get up to speed on it. So those are the kinds of things where they’re all happening in the background right now. And in the past, I was doing them. It took me 30 to 60 minutes every day just to prepare for my calls the following day. And now it takes me zero time. Occasionally, I have some feedback where they created the wrong template, because they didn’t have enough context to know who the person was. And then I would say, “Oh, you couldn’t have known this. Next time, if you can’t find them, make sure you check my Twitter DMs and LinkedIn, because it’s possible that they reached out to me there. And if they did, then you’ll have the context.” So just having those feedback cycles, there’s something about, I think people often feel like handing things off to an EA is a one way door where once you’ve handed it off, you lose control over it.

Sam Corcos (00:15:29):

But really the key to getting better at delegating is to figure out how to de-risk the request. So reducing the scope of what it is you need done and increasing visibility where you can now build confidence that they’re doing it consistently and they’re doing it correctly. Or doing it completely and parallel, where you’re still doing the thing you were doing anyway, but you have the EA tried to do it as well. And if it doesn’t work, then you didn’t lose anything. Just like, “Okay, well, I guess this is too complicated. They can’t do it.” So figuring out how to reduce the stakes to just try to get those cycles in to get an EA to do it, I think is really compelling. I think another thing to tie into that is people often may too many assumptions about how much effort it takes.

Sam Corcos (00:16:17):

I have found for me the overwhelming majority of the time, if I’m doing a thing that I want to hand off to an EA, all I have to do is record a Loom of my workflow doing the thing. And then I just say, all right, you just watch me do this for 30 minutes. Can you repeat by to me what you think the process should be? And then if that works, we’ll go from there. It’s like you don’t, you don’t necessarily have to be the person who defines the process. You can just do your workflow and then they can, you can delegate the definition of the process if you’re able to record your workflow. So just figuring out how to reduce that lift in order to start getting things delegated

Michael Mizrahi (00:16:55):

Loom is one of the most undervalued business tools of the last decade. It is infinitely more powerful than Slack or some of these other tools. Yeah. Looking through, we’ve got this database of all of our tasks, our process database in our Wiki and Notion. I didn’t write any of these process stocks for a few dozen of these asks. I recorded the Loom doing the thing, maybe said some things along the way. And then Vanessa, who I work with has built all of these into templates and processes. And so the redundancy there is also great, because if someone else needs to step in, if Vanessa’s out on vacation, the knowledge doesn’t live with me and it doesn’t live with Vanessa. We’ve got plenty of documentation and process and Loom’s recorded for each and every time this task gets done. And so that’s helpful for debugging when something goes wrong, it’s helpful for training moving forward if someone else needs to pick up the tasks.

Michael Mizrahi (00:17:45):

One of my favorite uses, and I don’t know that either of you have seen this was when Zach took paternity leave. He was going to be out for four weeks. There’s a lot going on in the legal front, but we’re in the middle of a fundraise, there’s a bunch of equity issuance. There’s a lot of contracts in flight and he needed some way to stay loosely in touch. And so he leaned on Joyce, who he works with to augment and support him while he was out. And when she did was daily-review all of his emails and tag a few of them into a high priority folder that he can check in on every few days. It was just a really creative use case. And he outlined the process for which emails are important, which domain names to watch out for. And then she ran this process. And from what I heard, it looks like it worked.

Michael Mizrahi (00:18:26):

But that was a really good use of any EAs to reinforce our kind of cultural principles of really disconnecting if you’re going to be offline, but building in systems so that you can stay in touch the way you need to with some piece of mind. And so Zach created a global out of office memo and then specific ones for the people he was working with. So me and him had a Miz and Zach out of office and then he had one with Joyce as well. Like, “I’m going to be out, here’s what you need to cover while I’m out.” Or, “Here’s what to know. Here’s how to reach me. Here’s where these documents go.” And so that was a really good use case by Zach.

Sam Corcos (00:18:57):

I have something somewhat similar, which is my EA goes through my email twice a day, right around the time the mailman does the batch then. And they, instead of flagging as important, they do the opposite, which is marking is not important. So they filter out a lot of the things that I don’t really need to get back to. It cuts down on my email load by at least 50%, maybe 75%, just so I can stay focused on what needs to be done now.

Ben Grynol (00:19:29):

One of the interesting things with EAs is this heuristic that I think we can challenge our societal heuristic of how to create and capture value from EAs. And that is EAs are great at transactional work that doesn’t require a lot of company context or discretion. We have this mental model. And so we’ve started challenging that internally a little a bit and saying, “Is that actually right?” Just because we think that’s right, what happens if, and this has all been experimental, but what happens if you bring an EA into the fold and you take it as far as you can? And we haven’t quite done this yet, but let’s say, “Vanessa go right a gross strategy memo and spend five hours on it and give four bullet points.” So you record a Loom and it might be five minutes and you start a Notion document, it’s literally five bullet points.

Ben Grynol (00:20:26):

And the point is to come back and be like, “What is this going to look like?” And so we’ve done this a couple of times where Vanessa, she started the foundation for one memo. It was around crossing the chasm. And it was just research. Research that might have taken two or three hours. And she came back with an 8,000 word memo that she did in two hours because it was a lot of aggregation of information. It wasn’t saying this is how Levels will cross a chasm, but it’s providing a foundation so that there are enough rocks there and you can start to fill in around it. And then she did the same thing with some research around big, so Big Food, Big Pharma, Big Sugar. It was this idea of what’s it going to come back? What’s it going to look like?

Ben Grynol (00:21:09):

And so she put together all of this information, you go, “Wow, if you can start to outsource the things that you think like, no, no, I’ve got discretion. I’ve got company contacts.” All the things that feed our egos as humans, like thinking why we’re so self important at being able to. The only person in the world who can do this task as me, as soon as you remove that, and you say like, “Could Vanessa do this?” And she comes back and something’s not just good. It’s like really good. Then you go, “What does it look like when we start to scale up?” So we’ve taken it to the extreme with not just memos, but even we’ve got, Tony has nine people. So we’ve been working to scale his time. We’ve got nine people working on multimedia.

Ben Grynol (00:21:50):

He’s the only multimedia producer and Irwin has taught himself how to use Premiere Pro for video editing. Vanessa has taught herself audio editing into script. We still will send Loom, but we’re doing all these things that are value ad work. Instead of just like, “Hey Vanessa, can you book me a haircut?” And there’s still some value in that, but it’s more along the lines of like, “What does it look like when you treat them like a team member?” That’s where you get a ton of value. And it’s been a very fun experiment to run, to keep, like, “How far can we push this thing until it’s like basically the EAs do all of our work and we’re just pawns on the chess board?”

Michael Mizrahi (00:22:28):

Then the next question I think comes up, which is when do you hire internally? And what defines the line? If folks are getting involved in starting to have discretion and applying thought that has judgment that’s relevant, there’s some line where someone might say, “It sounds like you just have employees who are contractors, how do you make the distinction?”

Sam Corcos (00:22:49):

Yeah, I think the hiring support staff is something that people often have a struggle justifying, but it is almost always the right decision. If you can build a structure to be able to use that time effectively. When you calculate the difference in time, if it takes Mike Haney, our editorial director, 15 minutes a week to put together some slides to update people. If that 15 minutes could be done automatically by the EA team. That 15 minutes, we also forget how much of a load, just that cognitive burden of knowing that you have to do it costs. It’s much more than 15 minutes. It’s probably an hour of cognitive load to be able to finish that. So I think hiring a support staff as early as humanly possible is almost always the right call. One of the challenges with EA is that, especially when they’re virtual, is that… This is the struggle that I’ve seen with a lot of people.

Sam Corcos (00:23:59):

Is because they’re so physically removed, people often don’t know what they’re doing. And so it’s extra important to have a feedback mechanism where you are… We do. We ask the EAs to record themselves on a Loom doing the work so that if we need to debug something later, that’s part of it. But it’s also just to build the confidence, to just know that things are being done and that you can close the loop. There have been many times we have a database of every task that has been completed and a walkthrough of that task. And every once in a while, I’ll ask myself, “Hey, I wonder if that thing is still being done that I asked for.” And I go to the database and it’s like, “Oh yeah, they did it yesterday.” And I can’t even watch the recording of her doing it. And it’s like, “Yeah. Perfect. Okay, great.”

Sam Corcos (00:24:47):

I now have that trust and confidence there. So it’s super, super important to find mechanisms for building that trust. Because when people are in other countries, you barely interact with them at all. You need to really know that things are being followed through on. So in terms of when do you bring them on in-house or full-time, I’m not totally sure that it matters. It’s more just a scope and context question. So I was talking to somebody the other day, this was like a few months ago and his EA ended up doing such a great job that when he left his role, he actually recommended his EA fill his executive position.

Sam Corcos (00:25:29):

And yeah, that’s not a common thing, but she moved from the Philippines to take on this major leadership role at the company, which is pretty wild.

Ben Grynol (00:25:41):

Wow.

Michael Mizrahi (00:25:41):

Yeah.

Sam Corcos (00:25:42):

Yeah. And that just shows you if you can build this confidence, it’s very possible that as they take on more and more of responsibility, they can end up filling that position. So I think in general, figuring out a really good way of having that long term trusting relationship and just building confidence over time, because that physical separation, the lack of regular oversight can lead to a slow erosion of trust.

Ben Grynol (00:26:13):

Yeah. And removing the bias, that’s the key. Is removing the bias that an EA is only capable of transactional work that I lay out perfectly. That’s where it’s really easy to fall into a trap or to go wrong. Is like, “Oh, no, there’s no one that could possibly do this memo.” Sure. It might be challenging to the question of when to hire somebody internally. Well, it’s really hard for an EA to have enough company context and to spend time thinking… Let’s use community as an example. It would be really challenging for somebody who is overseeing many tasks that have a wide scope to go really deep on one category for months at a time. That’s just not possible. And maybe this EA becomes very good at that one thing, community, it’s still harder to say, “Think really deeply about it. Have enough company context, have breadth and skillset.”

Ben Grynol (00:27:08):

Let’s say this community person is technical. They’ve got a business lens and I don’t know, a design degree. They’ve got all of these cross-functional skills. It just makes them really good at that one thing. And it’s still a matter of knowing that you are giving someone discretion, but knowing that it’s also you’re going to need people who can be thought leaders and advocates within your company to work across product and community and business, and name all of these functional areas. So it’s an ongoing challenge to figure that out. But if you can get the balance right, you really can find that EAs can do more than just giving them these recurring tasks of, “Check my emails once a day.”

Sam Corcos (00:27:54):

One of the things I’ve already emphasized this, but I’ll emphasize it again, is finding ways to just reduce the threshold for trying to hem something off. Oftentimes, you could just record yourself. So like when I mention how I have my EA go through my email and filter things out, the entire process for that was just me doing my normal routine, which is going through email. And just while I’m doing it, having a Loom recording it and just out loud, it’s like, “Oh yeah, these kinds of LinkedIn messages always go to not important. These GitHub notifications always get moved. This kind of thing always gets moved. This one, sometimes, this one, sometimes, this one always.”

Sam Corcos (00:28:39):

And it just went through my normal workflow at just the normal speed that I would do it. And then I handed that off. There’s no extra work there. And the worst case scenario is that nothing happens and it turns out to be too complicated and then they don’t do it. And then fine. Handing things off, there was one project that I wanted to do an analysis of one of our channels. And I asked one of the EAs to spend 20 hours giving me some information related to these channels. She spent 20 hours, the deliverable that’s exactly what I was looking for. And it turns out the information just wasn’t very useful. So it wasn’t on her.

Sam Corcos (00:29:21):

It was just like, “Oh, it turns out that there’s not a lot of good information here.” And then I just threw it out. And that’s fine. It’s not a, given the relative time cost, if I had spent 30 minutes working on that and then came to the same conclusion, instead of her spending a half a week, her job is to free up my time. And she did that very, very effectively. So this is a comparative advantage question. So it took her more chronological time to get to the same outcome, but it saved me a lot of time and that’s super, super high leverage.

Michael Mizrahi (00:30:01):

I’m thinking back to when I onboarded was joining, Sam, you were handling a lot of the operations tasks at the time in the months preceding. And you had recorded Looms of a lot of these without any foresight of what they might be used for. But when I came on board and started watching those videos, my onboarding was supercharged because I didn’t need to ask anyone for help or have context on how a task was done. And even if there was a 30-minute Loom, I could watch that and get up to speed on what I needed to do to now own that task without even you knowing or having to get involved.

Michael Mizrahi (00:30:35):

And so there’s a lot of lessons to apply, even if you’re not working with exec assistant, when it just comes to being really, really aware of the work that you’re doing and mindful of when you’re actually doing process-oriented work, versus when you’re doing critical thinking and putting down a memo and strategy. There’s a big difference between those two, those categories of work. And so I’m just thinking about folks who say like, “We can in 40 days,” or, “We won’t work with them,” or whatever it is, there’s a lot of lessons still in just thinking about working this way.

Sam Corcos (00:31:13):

Yeah. I mean, I would say you can’t afford not to have them. If you have your most expensive people at the company spending their time doing tedious work, one of the things that I’ve really been pushing our engineering team to do more is to do use the EA team. We have some of our engineers, this is a frustration that I have had in engineering teams in the past. Is there’s so many different places for documentation that you have to like, “All right, I have to write out this long description in the poll request. Then I basically have to copy over that information into Slack so that people know that it’s there. And then I have to put it in the documentation and confluence.”

Sam Corcos (00:31:53):

And I’m just duplicating information in a bunch of different places and they inevitably fall out of sync. I just don’t like doing it. And me as an engineer, it’s a very bad use of my time and I hate it. Or you can have a single source of truth where engineers agree to put the information on, and then if you need to replicate it, you just have the EA team automatically just as part of their job every day, check every poll request and then duplicate that information and all the places where it’s relevant. And they can totally do that. And it saves the engineers a ton of time and ton of overhead. And it probably happens a lot more consistently too, because the juniors really don’t like doing it. So they’re going find of reasons not to, if they don’t want to.

Michael Mizrahi (00:32:38):

There’s something nice about the async time scale that we work on, where most of these tasks can get completed within a day or so. And they don’t need to be done within a few hours and the work still gets done and everything ends up in the right place. And so having a natural buffer built in allows for this kind of work to happen as well.

Ben Grynol (00:32:57):

We did a podcast and it was about eng documentation, and so we were riffing on all these things. And I half jokingly said, “Well, why don’t we get to a point where the EAs are doing the reviews?” Because it’s actually a transactional thing, says the guy who’s not an eng. But it’s transactional in the sense of if you teach somebody what to look for and how to look for it, that’s actually what they’re really good at. Is pulling apart detail and go, “Does this recipe match exactly? Is it exactly laid out as it should be?” Because if you think about the amount of time and context switching involved in doing reviews and maintaining velocity, good engineers are there too.

Ben Grynol (00:33:40):

Sure, good engineers are there to fix bugs and to do reviews and to say, “Hey, this could be better. Why was this done in this certain way?” But if you can find a way to use another team member as that strategic resource, that being an EA, it’s a really interesting mental model. It gets back to the idea of well, how do you test for this? Because it’s so easy to go like, “That’s not possible. Let’s ignore it. That’s sort of like the default, I think. No, no, there’s no way anyone could ever do this, so then people don’t explore it.” Well, a way of working really well within EA is to be prescriptive.

Ben Grynol (00:34:17):

If you want things done, like write out snippets and go hit semicolon, whatever your snippet is, send this email and that’s prescriptive and they’ll follow it. The way to stretch for discretion is to actually give, or at least the way we’ve been trying, it is to give as little direction as possible, which sounds counterintuitive. But the reason you’re doing it is to see how little information you can give them and how good what they come back with is so that you go, “Wow. If I actually gave them 15 minutes of direction instead of 3, and did this 10 times, they would be exponentially better at doing that one task.” And so you go, “Is this possible?”

Ben Grynol (00:34:58):

And so these are the things that we’re trying to test with the idea of one-time task versus recurring high value, versus transactional work that you’re prescriptive in the approach. But it’s a neat thought exercise to go through, to go, “Is it possible?” And it’s intuitively, and again, this is probably being totally irrational and overly optimistic, intuitively I think it is possible. Especially if you’re talking about relatively straightforward builds to have someone like an EA do reviews on code. I really do. Says the guy who has never written a line, but conceptually, I think it’s possible. So I don’t know, challenge it. What do you think, Sam?

Sam Corcos (00:35:39):

Yeah, I think that it’s not as crazy as you think it is. There are things that EAs should be able to do in this process that are currently being done by engineers, but don’t need to be. So for example, the EAs can function as a first pass on poll request formatting. So the person fill out the template? Did they record a Loom? Are they following the basic guidelines? In my mind, EAs are the entropy fighters in organizations, where as you add new people, processes start to get done a little bit less consistently. And then eventually, it all just breaks because people stop doing them. At a minimum, you need to know that that’s happening, but better would be you have somebody who’s actively fighting against it.

Sam Corcos (00:36:36):

And so some companies have the documentations are where they review every poll request. And if you didn’t update the documentation, they will decline your poll request until you update it. So they have final veto on everything that goes into the code base. They’re fighting the entropy of technological debt and bad documentation. So EAs can do things like that. That is a good role for them. Is making sure that things are being consistently followed. They can watch every Loom, and they can say, “Hey, we request that everyone does a walkthrough and does a code walkthrough.” And you only did the application walkthrough and you didn’t walk through the code. “Please redo this Loom and also go through the code that you changed.”

Sam Corcos (00:37:24):

Another thing is sort of the QA function. If we have good specs upfront, they could potentially get it up and running and they could say, “All right, here’s the spec, here’s what’s in this poll request, do these things match?” And they can just give a thumbs up or thumbs down. And that’s something that Loom actually helps a ton with this. Which is when you have a Loom recording of what is in the full request and you see the person with their emulator open walking through it. One of the most tedious things for me doing code review when I was spending a lot of my time doing net managing eng teams was, it’s like all right, we’ve got the new poll request open. I’ve got to pull it locally on my machine. I have to reinstall all the dependencies because it’s going to potentially be a different version.

Sam Corcos (00:38:13):

I have to get it up and running on the emulator. I have to do all the signing stuff. I have to… It often takes 15 to 30 minutes just to get the environment into a circumstance where I can even see if it does anything close to what it’s supposed to do. So being able to do this in video or being able to have somebody else jump into that step, that saves you 30 minutes on every poll request, it can save a massive amount of time. Especially for people like senior engineers, which are some of your most expensive resources in the company. So, if you can find a way for an EA to make the code review process for senior engineers 25% more effective, that’s a huge, huge leverage point for the company.

Ben Grynol (00:39:01):

One of the analogies to jump into is, let’s make the assumption that you can teach. I mean, not assumption, it’s true. Just you can teach someone a certain language and they pick this up and then they’re doing a review and maybe doing a Loom walkthrough as there’s a review. So if there’s a second set of eyes, somebody’s watching that on 2X. Interesting. Well, think about the analogy of a movie set. Those are essentially EAs to, there are people that have technical skillset, but they’re EAs to the lead actor in a movie. The lead actor walks in and everything is set up so that they can do their thing, do their best work.

Ben Grynol (00:39:42):

Imagine the scenario where it’s like, you had to go and you had to wait. And there was you watch the lights get set up. And because you are the star, you’re saying, “I want that light there. And I don’t know about the lens on that camera.” And all these things. It would be the most absurd thing. There is a reason why people that are, it doesn’t matter, professional, anything. When you get to an elite level, you’re a professional hockey player. you go on the ice and the Zamboni, he’s already done it. You’re outsourcing to all these EAs. And so it’s the same mental model of saying, “What am I here to do? I’m here to do what I can and no one else can. And what I’m really good at.”

Ben Grynol (00:40:21):

Whether you’re an actor filming a movie, whether you’re a hockey player that is going to face off on the ice, or whether you’re writing code, it’s all the same mental model. And so if you start to think of the EAs as a strategic resource that can do things, if you trust them and have technical skill. These are people, these are humans, they want to do meaningful work. Nobody wants to be a box folder in a factory. There’s a reason why there are machines that do that. Now, because it’s just repetitive transaction work, it’s not fulfilling. Get them to figure out how to build the new box machine. Someone’s into that for sure.

Sam Corcos (00:40:59):

I might actually dis agree with that last point. The lead actor is really a perfect analogy for how the support staff really helps. In Cal Newport’s latest book, A World Without Email, he talks about how incredibly undervalued support staff is and how most companies wait way too long. And almost everyone should have more support staff. It is a net increase in company output when you add support almost always. Where I might disagree is that one of my biggest learnings from working with Lori for as many years as I have, is that it’s maybe this saying different strokes for different folks.

Sam Corcos (00:41:38):

Which is that she loves doing a lot of the things that I really cannot do. I just don’t have the patience for it. She loves wrapping Christmas presents and having them already and coordinating that. When I see something at a store or have an idea, I will regularly send Lori a message in like February. And I’ll say, “Hey, make sure you send one of these to my brother for Christmas this year.” And usually around November, December, I’m scrambling thinking like, “Oh man, what Christmas presents do I get people?”

Sam Corcos (00:42:12):

And Lori says, “Oh, I already have these eight gifts wrapped and ready to go, and I’m shipping them on December 20th.” It’s like, “Oh, I totally forgot that I told you to do that. So great take care of that.” The kind of like rapid Christmas presents, the routine of that is something that I couldn’t do because I’m relentlessly novelty seeking. But I think that there are many people. I imagine a lot of accountants are this way. It’s like being on top of things, having order, bringing order to things is something that is actually is something.

Ben Grynol (00:42:47):

Let’s debate it though. The Lori example’s a good one, because that is still requires some discretion. I think where we go wrong as a society is assuming that if Lori only for eight hours a day had to wrap the exact same size box with the exact wrapping and there was no other type of work, that’s the thing that feels draining. That’s what you hear about when people go, “My job doesn’t feel stimulating.” They don’t feel good about the tasks they’re doing. And if EAs get treated that same way where it’s like, “I only ever want you to schedule calendar meetings.” That’s all they do. And there’s not a ton of discretion, “Send this calendar link,” that feels deflating, I think just for anyone. We’re humans.

Ben Grynol (00:43:31):

We need to feel like we are helping to do things that create value. And so it’s like finding the way so there’s still transactional work, finding the way to use discretion when needed. I agree that some people don’t want that. Some people feel fulfilled by doing certain tasks, but finding the right people doing the right things and then stretching them in the same way we all want to be stretched as professionals. We want to be just on the, what’s that saying? You’re in the flow state when you’re just on the edge of your skillset and challenge.

Sam Corcos (00:44:03):

Yeah.

Ben Grynol (00:44:03):

That’s where everyone wants to be, because it feels really good. It really feels good. And for, let’s say the wrapping presence, Lori might be on the edge of finding the perfect present and the perfect wrap, that’s putting her in flow. So it’s an interesting way of thinking, using that mental model for how to be really effective with EAs.

Sam Corcos (00:44:23):

Yeah. I wonder. I do think that different people find joy in different things. There are people who love doing books and accounting and I find it mind-numbingly boring. Zach, our head of legal really doesn’t like doing contract fall, and I don’t blame him. I also don’t like reviewing contracts. We’ve been interviewing a number of legal associate candidates. And one of the people I interviewed, I asked him, “What’s your ideal day look like?” He’s like, “Oh man, I love contract law. If I could have my ideal day, it would just be reviewing contracts all day.” And I remember thinking during this interview like, “What is wrong with you?”

Sam Corcos (00:45:08):

But also that’s what we’re looking for. So, he is the kind of person who just finding the nuance and the language of contracts is the thing that brings him joy. I think that’s great. I wonder if we’re projecting too much on the way that we operate or different people operate. Is for me, a lot of these things would not be fun. There are people who are for the term use of they’re gardeners. Gardeners, often really like their work and they’re not finding the optimal garden. They’re just keeping things in order. And I think that is a personality type. That is a preference. You can be in flow state as a gardener or keeping your Zen garden in a particular form. And I don’t think there’s anything wrong with that. There’s no judgment associated with this. I think it’s just a matter of preference.

Ben Grynol (00:46:02):

Man, pulling weeds, you can get into a flow seat with that. That’s not a joke either.

Michael Mizrahi (00:46:06):

I think the key is just an overarching theme of respect and appreciation and understanding of what the task involves. And so it’s not sending someone to a blackboard to write the same thing 500 times, but this work is really valuable, it’s moving the company forward in a meaningful way, it’s empowering all of us to do really great work and helping us build towards this mission. And so all of this is necessary and helps us get it there faster. So, so long as everyone has a rightful and honorable place in that entire picture, all as well.

Sam Corcos (00:46:41):

That’s a really good framing for it, because you have to believe that what you’re doing is contributing to the larger goal. If our EA stopped working and they stopped doing a lot of these routine tasks, the company would not function anymore. They really are an integral part of a lot of the processes that we have every day, multiple times a day. So yeah. I think that’s a really good framing for it.

Ben Grynol (00:47:09):

Being totally honest, I think it was one or two weeks ago, I had sort of this moment of reflection where I go, “I almost messaged you.” I go, “Uh-oh! If Vanessa’s gone. the whole multi media effort’s can crumble.” I thought it was that existential and it is sort of true we could ramp somebody up, but the amount of work she is doing to allow us to put out all the content that we do from a podcast standpoint, from a video standpoint, it was one of those things where I was like, “Uh-oh! We cannot lose Vanessa as part of this process.” It would come to a halt. We would figure it out, but it would come to a halt for the period of at least one day, probably more like a few until we got the process going again. Because everything is a part of the machine, it’s lubricant and it’s making it work.

Ben Grynol (00:48:06):

So it’s one of those things where as soon as you start to bring them into the fold, they really become levers of your company in scaling your time. One of the things we should dive into though is we’ve been talking philosophically about all these things, but what are some takeaways that people can do as far as Mario… So Mario from the generalist, he and I were talking about EAs and he was thinking of working more with them. And it’s one of these ideas of like, “What exactly can I use them for? I’m trying to figure out.” Same with Wiz, who’s one of our investors. I want to use them more, I don’t know what I can use them for. So what are some things that if you were to give, if each of you were to give recommendations, what can people do to start delegating? Where to start and what are good tasks to start with?

Sam Corcos (00:48:59):

I think the most obvious one is if there are any tedious, recurring tasks that you’re currently doing, try to find a way to hand them off. There’s automating these things in the background. It would be hard for me to explain to somebody just how much of leverage I personally get from our EA team right now. I think the last I looked at the numbers, I have enough recurring tasks to occupy about five full-time EAs. Just with things that I would’ve otherwise had to do myself, there is now more than if I had to do myself I could never possibly do that much work because it’s so fully automated. So those types of things. Often, very small things like weekly check-ins on something, just have your EAs do that. Or I added another, just very simple monthly task, which is I have a spreadsheet where I keep track of how I’m spending my time.

Sam Corcos (00:49:59):

And every month, the chart gets out of date because it’s only up until the current month. And so every month, I have to go in and then I manually change it, which only takes a couple minutes. Sometimes, I forget how to do it, because the calculation’s a little confusing. So I just recorded a Loom and said, “All right, on the first of every month do this.” And it took me two or three that’s to do that, and now I never have to do that again. And they send it to me on Threads, so I get a message when it’s updated now. So it’s just adding those types of things they start to really add out. So finding those little tiny tasks.

Sam Corcos (00:50:38):

I would say another is like I mentioned just reducing the assumptions around how much work it takes to hand something off. Something you very realistically could do is just open up a Loom when you wake up in the morning and just record your entire workflow for the entire day on a big Loom. And just give that to your EA and say, “Is there anything in here that you think you could take off my plate?” And just see what they come back with. Worst case, they come back with nothing. Best case, they say, “All right, here are 25 things that I propose I can start doing for you that you won’t have to do anymore.” Allow yourself to be surprised by just how capable people are.

Michael Mizrahi (00:51:29):

Yeah. Plus unto all of that. Looking through my tasks, there’s a healthy amount of them that are recurring. And then there are a lot of tasks that are ad hoc, but that come up on a non-recurring basis that I’ve had to do or handle. And I think a lot of these honestly have kept us from having to hire additional full-time staff. And so some examples, things that are likely to go out of date if they’re not regularly revisited. So we have a lot of team directory tables in Notion and everyone’s got these fields filled out on their profiles as new hires join. We put in the onboarding to tell them to fill out these fields, but they don’t always get to it. It gets left over whatever it is.

Michael Mizrahi (00:52:13):

So on a regular basis, not after each new hire, but let’s say every two weeks, Vanessa goes into that team directory does an audit of all the fields and then emails those people individually to fill that out. That’s something that there’s no API Notion or maybe there is now, but we would need engineering time to build this internal process. It takes just a few minutes. And also something that’s worth mentioning is that there are probably things that could have notification set up or something like this. But being reached out to by a person makes you much more likely to respond and handle that task. It’s very easy to archive a notification or an alert from some automated system.

Michael Mizrahi (00:52:53):

It’s much harder to do that. When you get a Thread from Vanessa with me on CC saying, “Hey, you’re missing this field, can you please fill it out to keep us up to date?” And so those are the kinds of things that we’re talking about here that on the whole keep a really healthy system. Everything is up to date all the information’s in the right place. We’ve got the phone numbers copied over and then there’s examples of ad hoc processes. When someone starts ordering a laptop is something that I previously did. I really don’t need to do that.

Michael Mizrahi (00:53:17):

That’s a very easy task to outsource. There’s like very specific skews that we order. The shipping address is in the right place. So that one comes up on demand and there’s a process in place to do all these kinds of little things. And so even just looking at our company task database today, so tasks that got completed this morning until 130 Pacific, my time is about 60 tasks across the team have been automated this morning alone. So there’s a lot of work happening in the background just like updating track, doing research, cleaning up, calendars, cleaning up the recruiting pipeline and making sure everyone’s in the right place. Preparing bios for meetings, uploading to YouTube, downloading from the…

Michael Mizrahi (00:53:54):

All of these things that already take up a ton of time and we’re able to keep our team somewhat lean by leaning on the Athena team to run all of this. And so, yeah, ad hoc recurring and anything that you’re holding tight and precious can probably be outsourced at some point in time. And it’s just having the mindset that someone else can do this. And you can free up your time to let go of it. Like Sam mentioned earlier, “All of this, nothing here is permanent, can all be edited.” You can take a task back if you really like doing it, or it was important to you to have that context can adjust the process. Because each time it’s completed, you’re getting a notification from the team that this was done. And so you can make tweaks along the way. It’s very, very iterative.

Sam Corcos (00:54:37):

I very much view the Athena team as they’re my biggest allies in fighting organizational entropy. I use them all the time. One of the things that we do, which it’s a subtle thing, but it’s so helpful is once a week, they audit our leadership channel in Threads. Which is just for people in leadership and it’s not shared externally to the rest of the team. And they check once a week and there’s only Usually, I don’t know, 5 or 10 Threads in there. They will reach out to every person and they’ll say, “All right, these are the 10 things that came up in leadership this week. Can we declassify this and share it with the rest of the team?” And almost every week, we find two or three things where we say, “Oh yeah, let’s move this one to general, let’s move this one to marketing, let’s move this one to product.”

Sam Corcos (00:55:30):

And we find ways of just fighting that over classification problem, which is the end state where everything ends up. If you’re not paying attention is everything is classified and only two or three people get to see any piece of information. They also do this for all of our Notion docs as well. Where they go through and they just look for discrepancies, something that we classified it as public, but is not shared to the web. Or something that’s classified as internal so it’s not supposed to be shared publicly, but is shared to the web. And they just go through and they just find out like, “These are the 25 documents where there’s a discrepancy.”

Sam Corcos (00:56:11):

Or, “These are all drafts in our Notion database that are more than a month old. Why haven’t we shipped these?” It’s like, “Oh, I just forgot to change them. That’s good to know.” So you don’t have all these things that are out of date, where it’s like, “Oh yeah, I forgot about that. Let’s market as deprecated so there’s no confusion.” Or, “Oh yeah. I just need to put 10 more minutes in and then we can ship this.” So those are the kinds of things that they can really help with of just figuring out how to keep things more coherent as time goes on as the organization.

Ben Grynol (00:56:44):

The biggest thing that I’ve found as far as making a difference is, if trust so exist idea of defaulting to trust, like trust that they will do it, trust that they can do it and trust them with the account information. So, by giving Vanessa access to basically my entire life down to, I was about to give her access to my bank account to deal with everything, it allows anything that, I see tasks and by task, I mean something super transactional as work in my brain. And anything that holds up any mental space for work, feels like real estate that is occupied for free. So that might be like, “Okay, go download the lean startup for the leadership book club.”

Ben Grynol (00:57:29):

And to me, it doesn’t take many of those tasks until your brain is just filled with garbage. It’s just like a low value stuff. And it doesn’t mean like there are people that will enjoy it, but being able to say, “Hey. Can you go download this thing?” So it just pops up on your phone and just not having to like go through the checkout flow and all these things. Those are the things by giving people full trust, where you’re giving them your account info and your passwords it’s not going to be for everyone, that’s for sure. But I find that establishing a level of trust and giving somebody that account access so that you can really start to get them to do these things that are tasks, whether it is, “Can you format this memo? Can you take these notes?” There would be chicken scratch notes.

Ben Grynol (00:58:14):

“Can you put that into some documented Notion?” Here’s what it’s called, all of these things start to feel like a time multiplier and that’s where your ability to work with an EA starts to feel a lot more meaningful, they feel good about the work they’re doing, they know they’re helping you to scale your time. And yeah, it’s the idea of if there was one recommended takeaway it would be, once you find somebody you can trust, and if you give them access to your accounts and you get them to start doing these line item tasks, downloading eBooks, getting haircuts, creating strategy memos so you all have different complexity, it really does make a difference of your time and the way that you enjoy the work. Because you can start to focus on not being a contract lawyer, if you don’t want to be one.

Michael Mizrahi (00:59:06):

I’ll add a note on where we’ve drawn a line and that’s that the Athena team’s in place to help us with our tasks and our data, but we don’t give access to any of our customer data. So anything that involves the business side itself. But interacting with customer support, being able to access any internal information, we have controls in place there where even internally for employees, you have to have a reason to open up someone’s account and provide the help scout ticket number, whatever it might be. So as it relates to privacy and customer data and just respecting our privacy policy and commitment on that side, we’ve got some lines drawn. So the Athena team stops short of any tasks that involve that kind of information.

Sam Corcos (00:59:47):

Yeah, I think to continue to emphasize this is, it actually doesn’t require extra effort to hand something off to an EA. People make too many assumptions around that. Other people are generally quite capable, especially if you show them your own workflow. It’s pretty easy to get people to see how you do things and then try to just replicate that functionality. So, as a starting point for most people, I would say, just recording your daily workflows is probably the best place to start.

Sam Corcos (01:00:21):

And you’ll be surprised how many things an EA can take from that recording of how you’re doing things. As simple like it can be responding to emails, it can be writing memos, it can be coordinating between people, running those kinds of experiments is nearly costless. When you think about if EAs are 10 to $15 an hour, which is generally what the cost is for a virtual assistant. It is not very expensive to start to run these experiments and that kind of support staff is really a super power if you can use it effectively.

Ben Grynol (01:01:00):

To that point, that’s why we were running so many of those experiments because with some of them, let’s say some of the research ones where it was spend anywhere from, I think it was the task. Go spend anywhere from 15 minutes to no more than 2 hours. So then I was like, “Okay, this thing’s going to be 30 bucks at the most, and even if I spend like any ounce of effort on this, it’s just exponentially different as far as the cost, but that’s either here nor there. This experiment might actually be worth $500.” It’s worth some amount of money to find out what the outcome is. But when it’s 30 bucks, you’re like, “Run it, run that. Run 200 of those,” because that’s where you’re going to start to find all these unlocks. You’re using your laptop camera, right?

Sam Corcos (01:01:56):

Yup.

Ben Grynol (01:01:57):

Give it a wipe. Part of the 10X digital content memo, one of the leads we can pull on is quality. And one of the recommendations is that we all get cameras.

Sam Corcos (01:02:07):

It looks like that actually made a difference.

Ben Grynol (01:02:10):

Oh dude, 100%.

Sam Corcos (01:02:11):

Geez. That’s weird.

Ben Grynol (01:02:12):

Clearly you have not read all the quality things that have been written.