91. AI in The Future of Work

Hello everyone and welcome to another episode of the operations room a podcast for CEO's I am Brandon mincing and joined by my lovely co host Bethany errors. How are things going this morning? You ask me the question every single time and it is a surprise half the time. I'm like, how am I doing?

The shocking, how are you question. Yeah. It's like the softball. I'm like, nope, can't answer it today.

I am fine. We do these on a Friday. I have had a true Silicon Valley, the TV show week of highs and lows. I don't even know you can have that many highs and low in one week.

Scale up world, of course it can happen. So we had our board meeting on Tuesday. I was presenting some new ideas and a new strategy. It's a little bit nervous, but it went down well.

And I'm like, oh, you sorted. Then one of our top engineers, I'm trying to decide whether or not I can share it. I think it's resigned, and so then I'm like plummeted down, like, no. Yeah, your team is small, key developer out of the picture, that's gonna, yeah, hurt your progress.

Well, I mean, some of it is hurting progress, and he is a very good developer and technically very able, but it's also like, I really like working with him, embodies the values, a little bit older, you know, I think I've mentioned this before that I just feel so old and so uncool at the company. And so it's nice to have a couple allies who are on the other side of 40 with me. Yeah, so I can't go into all of the highs and lows, but It was just like that the whole week, and then ended on a relative.

Hi, and our show and tell had a demo of the new expansion product we're looking at. In general, the product's moving quickly, really cool advances. And then get home to remember that I had a different, I was a guest on a podcast between seven and eight last night. And so I just had to dig deep.

I know seven to eight doesn't sound that late, but after the week I had, it was like the last thing in the world I wanted to do. I would have paid money to not go on a pod cast. You're like, why did I book this? I do that all the time in my head.

I'm like, do I really wanna have this conversation with this person I'm having right now? I booked it, so it's my fault, but yeah. Yeah, so I think how I'm doing is just, I'm so happy it's the weekend. Okay, so Silicon Valley TV show, any absurdities this week?

Sometimes in this scale up, kind of ups and downs, there's just silly things that happen that are either tragically bad and really they're so bad, they're kind of humorous in a way, or they're so weirdly awesome. You're like, how, this is ridiculous. Any of that. This is the problem, I was thinking about this last night, is we used to be able to share so much more when we weren't working, and now we're working.

We can't be as open. Like, ask me that question in three years' time, and I'll be able answer it, but I can't answer it today. Yeah, it's funny because this whole idea of the building out loud, building in front of people where people are doing blogs and talks around the machinations of their company, realistically I don't know how you can really do that because at the end of the day, the internal stuff, there's a fair amount of sensitivities just in terms of either people, twos and fros or challenges that are directly related to the individuals that you work with that if you share it, I mean, it will come back to bite you, there is like no question of that.

Yeah, and it's all just very real time. I think sometimes you need a bit of perspective as well. Yeah. And not definitely not live blog deep in the fields.

I had a lot of fields this week, but I I'm back on a even keel. I spend most of my time as a CEO using all of the nervous system regulation skills I have built and developed over the years. All the reservoir. Cheers!

Really don't know how I would have done this pre those skills. I think I'd be an alcoholic. I think as soon as you put three people in a room together, you have people issues. And I do find people issues more stressful than business issues.

And I react to them more. And so there's just like a huge amount of time that I'm spent like breathing deeply, going for walks, shaking it off. Yeah, yeah, yeah. Because you can get consumed by it time-wise and mind-share-wise.

I mean, this is almost the trick of the trade, I think, for scale-ups because the business focus of getting stuff done and executing and winning new business and progress with the product and all that, the self-inflicted wounds in all these companies by far that consume so much time and energy and diverts yourself is all that stuff. All the people problems. And if you get sucked into it, issues with people or you take it too personally, You lose hours which mountain to days and You know, if you're losing time, I guarantee you the others are losing similar capacity in terms of their mind share.

So it's just like a huge distraction that hopefully with experience you can kind of like ensure for the most part for yourself and hopefully for others around you, somehow you can make sure people are focused on the right to actual things here. Just don't know if it's possible. I think it's possible to insulate yourself from outside of the leadership team, but even the leadership team dynamics, they just come in. So I was talking to Ryan, one of our previous guests, about some of the stuff that was going on, and he was a previous CEO of the company that was sold to Microsoft.

And he was saying that he was talking to one of his advisors, who's super senior, deals with like top 10. CEOs of the top 10 businesses in the world kind of level. And then we talked to Ryan and Ryan be like, oh, the leadership team, this, that, this person's resigned, that's happened, oh, this stressful. And like, I would just laugh and be like the last call I had with, you know, whoever massive company was CEO was exactly the same.

Like, it always comes down to personalities, the best leadership team. People who are growing, people on the ascent, people of the decline. And just balancing it all the time. Yeah, I suppose in that way, it's very much undervalued, the intangible of the chemistry in particular of the leadership team.

Do you actually have chemistry where there really is like a complementary synergistic effect and there's a minimum amount of politicking or issues or problem related, people related problems and so on where the team works well together, they enjoy working together and they know how to work together. You know, and I think that intentional piece. And I was just thinking about this in terms of our company, we've got a bit of a leadership reset because one of the co-founders is kind of spinning off in a slightly different direction.

We've had a maternity leave for the CRO, we have some other individuals that are now in play from a seniority perspective that need to be I think part of leadership in some capacity. So that what I'm cognizant of is we had good chemistry, good momentum, we understood how to work together quite well. So this reset, I'm just vaguely concerned that somehow... What we had built is going to, I mean, it's gonna be different regardless, but we're not gonna be able to reinvigorate it and have the same feeling.

I'm not going through a reset, but we also, on top of all these other things this week, it's the second week of our sales and marketing leaders, and so bigger leadership team settling everybody in. I have not spent enough time with them because I was prepping for the board meeting and then dealing with this other stuff, so I already know that I'm no following any of my own advice behind the curve. I was on the podcast last night feeling like an utter hypocrite, which is like, when you hire somebody who is experienced and you think they're great.

Don't just bring them into the business and then walk away and like, okay, you're in charge now, like spend time, really align, be clear on your joint objectives and understandings and really make sure your values are aligned. I have done none of that. I brought them into the business, pointed them at problems and said, off you pop. And I have to get back to being a good leader.

Also, they just, what I have managed to do is like the ways that I like to work that I hadn't articulated to them so then they don't know. And then I have to like, oh, by the way, like they were doing a lot of DMs. So like me plus two people or me plus one person or particularly because they're all like different combinations of the leadership team. So I had to finally just say, just send it to the leadership Team.

It's good for everybody have context, even if they're not interested and they might not read it but they can scan it and let's have as much open communication as possible. Like there weren't things that were sensitive. But if I think about as a new person, you don't really know, you don't know what the culture is. You don't want to blast people with spam.

It's just so much better than these like little tiny fiefdoms or like pockets of information and then you're trying to remember who knows what. Assume you know it. Oh, for whatever reason you weren't in that chat. Yeah, I hate that.

It just slows this down. That's how I'm doing for question number one of today. Well, I will say that so you look like you're in good shape because some of the calls we've had recently, you did not look your best. Just in terms of not physically, I could tell you were tired and a little angsty, whereas today you look and feel kind of back to the Bethany that I'm used to.

Yeah, I think it's just so we ended up other we ended up with a roller coaster on an upward trajectory on Thursday. So really, other than just the exhaustion of doing the podcast, like I think I've I'm in a good place right now. So let's see what today brings. It could be very different by the end of the day.

Yeah, you never know. Well, it is Friday and you're working from home today as I am as well. So I feel like today's like a solid catch-up day on things that I need to be attending to that I've ignored for the past 24, 48 hours. Just before we got on the podcast, I was just mentioning two things.

So one is yesterday, I had a massive headache by midday and it just got worse over the course of the afternoon and we have leadership in the afternoon. And I kind of, I had not. Pulled my stuff together quite yet for pre-reads, so it was quite late, so I'm kind of getting the pre-reads done for two things related to option grants and something else, eventually distributed it by noon. At that point, I was not feeling great, and the heading was killing me at that point.

Distributed the docs, went into leadership, and I kind of own leadership in terms of running it effectively. So it wasn't my best showcase, I would say, in terms running that hour 45 block. And at the outset, there were some angst related to some recent issues related to execution on. Of OKRs and so on.

So it was a bit angsty just overall at the beginning and I don't think I managed it as well as I probably should have. And then it was compounded by after that we had due diligence report that had come back from the investor for the latest round. And it's fascinating the observations that a third party has on you where they've done kind of like a McKinsey style job to kind of dissect the business from an external point of view in terms of, you know, how are we doing? How do they perceive us?

What's our score So that was fascinating to go through with the group. We're all sitting around the table, throwing in our viewpoints and so on, and I can feel myself. I had a couple good ones, I thought, in terms of observations on the due diligence thing, but in terms to being able to articulate it, you're just like, oh my God, what am I trying to say here? So it was not my best showcase, so I'm just thinking at the end of it, as you have pointed out many times in the past on this podcast, I need to catch myself in these situations earlier and just say, look, it's not helpful to anyone to have and go into a session.

And just like ramble on about stuff that is not clear basically, whether it's a leadership or in this kind of other session and just pack it in, admit defeat, go home. So that's my lesson learned I think from yesterday. So, actually something else of interest here, we had our strategy days, team days as we call them, last week on Thursday and Friday. So we ran our survey this week in terms of results, good report card, printers across the board, everyone enjoyed it and just good feedback all around I would say.

The interesting bet, two interesting bets, one is the survey results, we'd asked which piece of content, if you had a stack rank the content over day one, day two, what was the most valuable piece of content. And what came back may not be surprising to you, but it was sort of surprising to me, but the most valuable session was the AI upscaling on day two, and we had three tracks, as I mentioned before, Charlie Kamen on ChatGBT, we had two others on Cursor, more for the engineers and in Cloud Code, and one other for more of like non-developers on the livable side of things.

The quality of those tracks, as we know Charlie's fantastic at what he does. The other two were not perceived as high quality. Despite the concerns around the quality levels on tracks two and three, it was still considered to be the most valuable thing for the company. I think the macro point is people are really keen to get better at this stuff, basically.

And then we asked another question in the survey, you know, what is your interest level and more AI up-skilling and more challenges and so on. And right across the board, it's either extremely interested or very interested. I think there is. A lot of enthusiasm and desire and appetite to do more of this stuff and have the time and space to actually do it.

Ha ha! Awesome. I have like a little I told you so dance going on in my head. Yeah, but I appreciate it because I think all our previous conversations kind of led me to doing this in the way that I did it tonight.

And for sure, we're going to have some level of ongoing activity, I just can't, I'm not quite clear right now as to what exactly I'm going to do. Well, so the thing that I really need to do, know I need to do, and again, in terms of like, not following my own advice, but have to really find the time to follow my own advise, is management training for everybody, particularly the engineers, because at this point, I think there's like little bits of new technology, new techniques, you can always learn somebody else has figured out a cool way of using the tech, but that's not going to make the biggest difference for the engineers.

At this point, the biggest difference would be leadership skills, not leadership skills management skills. So what do you delegate? What, how do you parse up work in bite size chunks that an agent is equivalent to a human? How do you provide enough context, getting comfortable with slowing down to speed up, holding the agent to account rather than just being like, oh, whatever, I'll do it myself.

And I know I need to teach these skills. But I have just not gotten to the point of like building the framework. I have 20 odd years of management experience and it just becomes second nature. And so I'm gonna have to think about which books did I read?

Which frameworks do I use? What is the process that I actually go through? Which is now intuitive in order to break it down for the teams. And I just haven't had time to do it yet but I absolutely have to.

Yeah. And do you think there's a special twist on this whereby we're not just talking about managing other people now, we're talking about managing agents. So therefore, because I mean, we describe is every company's problem on the planet, which is how do we make sure managers are highly effective. So do you think there is something fresher to be done in terms of the agent twist?

I don't know if there's a fresher to be done for the agent twist, but it just means you have to go down a layer on management. And also you're asking people who are newer in their careers and have not been managed themselves very much. Because I think part of it is that by the time you become a manager, you've seen what works and doesn't work with your own. Whereas if you've only been in your career for two to five years or whatever, you're not always thinking about how have I been managed and how do I do it.

And it's also really separating the management skills from the leadership skills. Like you don't need to motivate the agents. Come on, agentic person A, get moving. But what you do need to have, which I guess is leadership, is the ability to communicate a vision and what your output you're looking for is.

And I don't know, is that a management skill? Is that a leadership skill? I put a bit more in leadership. So that's what we need because everybody is a manager.

And the more effective manager you are, the more affective you are using AI. And I'm just hobbling the team by not really spending a lot of time with them to have that mind shift on how you manage. Yeah. So I think this is a twist because what we're saying is every single person in the company, regardless of your role, whether you're a product marketing manager or anybody else, you are a manager, flat out.

So whether you have one year's experience, 20 years experience, you're manager. So therefore we need to level set across everyone in terms of how to manage. Well, I think the parts, going back to earlier in our conversation, the hardest parts of management and leadership are the people and the egos and the feelings. And you actually start to deal with that with agents.

All give what I think are good instructions, lots of context, go for a plan. And then it just like starts doing something weird. And with a human, I would have to say, that's an interesting take. Click.

Walk me through your thinking. How did we get here? I'm not quite sure this is where, you know, whereas with Claude, I'm just like, stop. Why are you doing this?

I wanted you to do X. You're doing Y. Don't do Y. Do X.

These reasons this way. It's quite freeing. I was thinking last night, I wonder how all this interaction with AI is going to affect the way that we interact with each other as humans. And I wonder if there is gonna be a level of bluntness that comes out because you spend all your day being so blunt with a machine that it starts to become a habit.

Yeah, for sure. Or I could go the reverse, where we just lose our skills entirely of how to talk to people properly, and it ends up being softly, softly. Yeah, so good to hear that the training went well, it was well received. Charlie is he's just a gifted trainer at the enthusiasm level.

And also just very clear, he's a great storyteller and he knows how to balance the variables of storytelling versus like trying to get us to understand key bits we need to understand within a compact two and a half hour time frame because that's always the balance a little bit between too much fluff versus like the concrete piece of it. Two things of interest actually, so the one showcase that we did at the end of the day was fabulous. So for our company as we're scaling up. We are going to be hiring quite a few engineers.

And when you do engineering interviewing, there's always the task phase. And the task face requires a senior developer to look at the task as part of that interview process and assess that task and whether that task has been well done or it's not been well done. And it's a key criteria by which we discard candidates. So the time consumption for a senior developer due to this activity is enormous.

And then compound that by the number of engineers, the number candidates are gonna be flowing through our pipeline over the next 12 months. Huge time consumption, huge suck. Very concerning for us as a company just in terms of our kind of time allocation. The AI output that we produced and that we showcased was automating that, whereby the task comes in and it's assessed, it's either a pass or it's a fail.

That there's a bit of a gray zone right now that we've left in. Whereby a developer needs to look at that if it goes into that gray zone in the middle band where the developer needs to just do the finite assessment. But on balance, the production of whatever we produce that day basically works. So now the thought process is we'll run it across all the tasks, we'll get the developer at least for the initial time period to do the same thing they would normally do, compare the outputs and just make sure the feedback is going into the agent to optimize it to make it on balance kind of where it needs to be.

Once we're in that state, then we can use it wholesale across all the tasks that are coming through for the engineering. So you've built yourself a engineering recruiter. Or at least automating one component of the process. What I also love with AI is there are many more ship-it days and many fewer hackathons, so things can get into production.

And I was actually thinking about this because every person that I spoke to called that day something different. We're like, we're having a hackathon. We're having, you know what I mean? There was various phrasings used around the day itself and I was thinking about this afterwards.

We probably need to like position this a little more clearly as to what we're doing here because to your point, I don't think this is really a hack-a-thon and we're like squashing bugs over the day or whatever. We're really producing artifacts or outputs that are usable in some form, if not a form like I just described where it's legit and we can actually use it going forward. That's really the purpose of these days. So making that more clear somehow to the group seemed like a useful thing to do.

Yeah. So we're doing our Charlie days next week. There's definitely some themes in today's podcast. So there's another theme in that after we did the training last time, one of the clear takeaways was it's too much to do general training in the morning and specific training for engineers in the afternoon.

Guess what we're in this time. Oh dear, what? 100% the same thing. So we're general training in the morning, specific training for the engineers in the afternoon.

And then that's all Monday. And then Tuesday is a building day. And so on the non-technical teams will be automating as much as they can. We have been compiling a notion list of all of our use case requests.

So Charlie has access to it as well. He can see what we want to automate. Every time we have an idea, we stick it on that paper. And then on Tuesday, whatever we haven't done on Monday as part of the training, the team will do on Tuesday.

And then for engineering, it's a slow down to speed up day. And so building a lot of the foundations that we know we need to put in place. Almost exactly like the management stuff. So like, this is the way we code.

These are the, you know, just giving way more context and rules so that we can move to asynchronous coding. And that's the goal is at the end of the second day, we have entered into the world of asynchronous coding Okay, yeah, that's amazing. So coding that is happening when people are not there. People are asleep who wake up in the morning to new new product.

Yeah, yeah, that's super cool. The other interesting little anecdote coming out of our one was we had one of our, I think it was the VP of AI, so one of out senior, kind of like executable of persons, highly talented in the AI space, he took on the GTM challenge of enrichment of our accounts and put together a really fascinating demo that did some amazing things. However, his recommendation at the end of that enrichment exercise was go license clay. So it's just interesting how you can get to a certain level with some of these artifacts, but there's a bit of a recognition sometimes.

There is a reason why companies exist on top of open AI, which is to provide verticalized solutions that are easy to use, that provide a comprehensive set of outputs whereby it solves the problem wholesale, which in this case is enrichment. So he felt we had done a good job on a very specific siloed strand of enrichment, But was a broader problem to be solved with enrichment and also just purely from a usability standpoint having a tool that is just more integrated with stuff effectively.

So that was an interesting outcome. All right, so we've got a great topic for today, which is AI in the Future of Work. We have an amazing guest for this, which Agatha Nowika. She's the Managing Director of AI Visionaries in partnership with Google Cloud.

So before we get to Agatha, a couple of questions for you. The first one is, what are the skills that matter for the next generation of employees? And I was thinking about this specifically in the context of, for myself, we are looking at our job descriptions. There's a clear recognition in our job descriptions that people will be using AI to respond to our job description in terms of prepping for their interviews, for the screening calls, for sure, for the tasks that we're going to provide and so on.

So then it got me and our head of talent thinking in terms how to set up the tasks and the questions that we were asking to really talk about much more of these skills for the next generation of human qualities of trade-offs, judgment and so on and kind of the more this like system level thinking and the analysis part of it and walking through crisply like how they think about solving problems to gear up our talent function to address that more directly. So my question to you is skills that matter for the next generation, what's your take on this?

I've been thinking a lot about it. There was an episode of the AI Daily Brief podcast that I listen to every day, and it was one of these where they were looking at KPMG surveys and this survey and that survey. It was a survey day of insights coming out. One of the things - it wasn't actually so much the surveys now, but how rapidly things are moving and in X number of years, one year, maybe up to five years.

Execution will no longer be the bottleneck. And to think about that, like you can say it, but then you actually have to think and just imagine like, you know, if we do get to this point of asynchronous coding, coding agents are really good. Anybody can build whatever product as quickly as they want. What is the role of humans?

And what is your competitive advantage? And how do you still grow a business? Because our entire lives have been about leaders. It's ultimately.

Allocation of resource and getting shit done. In some ways, we're going to have unlimited resource. Allocation of resource is going to be a whole lot less important because you don't have to hire a bunch of people. You can just do so much more.

This is a bit like, I can't remember her name, the woman that we interviewed from NYU around the coaching. She was talking about this a bit. I had a first glimmer of what's it going to list on. And therefore, what are the skills that you need?

And I think what you're talking about is right, Brandon. It's taste, it's judgment, it it's harder problem solving. It's having vision and understanding where you want to go and why. And then rallying your agents to do that either within an organization or as a single person.

Some of the people issues I had this week, I was like, Oh, fuck it. Like, why can't I just do my own business of one? And like, are we all just going to be businesses of one and then we go and hang out in the pub, you know, with people in the evenings have nothing to do with our businesses. But yeah, I think it's that I'm worried about what are the impacts to humanity, because people who have those skills and can rise to the top, it's going to be unbelievable.

But for people who don't, or you know that's not their passion, or they were quite happy. Entering data all day long and then going home and running the marathons or whatever like it's going to be. Hard if you don't care i don't think this would be any quiet quitting the other point that i thought when you're talking about everybody's gonna use a for their tasks is good. What's you don t be worried about but what it gives you an opportunity is to see how good they are day i so if it sounds good in the first reading and then you realize it says nothing it's slow.

It's polished nothingness. They've helped you filter them out. So I wouldn't worry so much about if they're using AI or not, I would worry about the quality of the work that they're putting in front of you. Yes, yes.

I think this is exactly on point. I mean, you can see it. I can see it all the time now in terms of documents that are created using chat gbt or you know the gemini or whatever. It's usually pretty clear.

There's a lot of if you glance at it, it sometimes looks perfectly fine but we actually read it properly around our situation, our problems. It's missing a bunch of stuff and there's a genericness to it. And even for myself the other day and this is where there's a bit of like a weird balance happening now or rebalancing my mind around how to use this stuff effectively because As we all do, I use it very much to create what I would have done much, much faster where using my prompts and various kind of existing materials, dump it all into the mixer of chat GBT, add in a bunch of prompts to kind of restructure things and so on.

But when it does like its little reworking bit, sometimes it rewords things in ways that are like, you miss it. Just one example from yesterday where everything was like, great, the document was good, it was robust, this is the option grant thing. There's just one statement around. Exception handling for giving more options than our table defines for a given role type based on exceptions that occur and thereby it's escalated to the CEO and had some examples of escalation, things that would justify an escalation basically.

So something the wording was a verified competing offer constituted an ability to escalate. And I'm like, a verified compete offer, that's never going to happen. Like, so I missed it. And somebody in the meeting called it out, not in a bad way, but just like, well, Brandon, what do you mean by verify competing offer?

I'm like, oh, chat GPT slop. Sorry. That's my mistake. Yeah, and I do find that that's something you really have to fight with it.

It loves his adjectives and it loves his adverbs and it just throws them in unnecessarily and I'm constantly stripping them out. We don't need this. And then also likes to say the same thing three or four different times. And then I have to explain to it.

You've said it once. These next three sentences are pointless. Take them out. And then it goes, you're absolutely right.

That's a good catch. I was like, I don't to know that. Thank you. Like just take it out.

All right, love that. Question number two, are COOs treating AI as an incremental efficiency tool or properly using it as a survival level strategic shift? Well, so we recorded this a while ago where it was probably more a efficiency tool, but I think it's all moving so quickly that there's a recognition around the survival shift. I think we're just on that cusp of moving to survival.

Have you played with cloud code yet? I mean, I have, but not recently. I've been focusing on my Charlie Cowan training around GPT. So I'm just in the process of starting to play around with cloud code, mostly because ironically, co-work came out as like the non-technical version of cloud code, because I've seen Charlie was talking about cloud code and everything he was doing with it in October.

And I was like, Oh my God. So they've actually made an agent framework that seems to work better than any other framework, but then I was, like, oh, it's terminals and it's hard and blah, blah, I'm afraid of it. And for whatever reason, co-work. Everybody I know who's non-technical who uses Claude code, tried co-work is like, there's no point using it.

It's so limited. It's nothing like Claude Code. Just like put on your big girl pants and get on with it. And so, okay, fine.

I'm going to learn how to, how to do everything in Terminal. And I'm not suggesting everybody goes and does it, but it's the best experience of the future. And the question is how soon is this future going to come? And I'm answering your question in a roundabout way, but I feel like if you're running the transformation in your business and you're making sure that we're getting all of the best strategic changes, like the collapsing of systems, you don't wanna just automate the existing systems, your businesses are gonna restructure, processes are gonna fundamentally change.

As a COO. You need to be leading from the front and really understanding it intuitively yourself. And the only way to do that is to become a super user, regardless of company size. Because otherwise, if you don't understand the limit and how it works and the changes in your brain you have to make and how you're speeding up your life, how are you able to have the credibility and force the other people?

How do you know what to call bullshit? How do know when to inspire? And how do you know how to create your vision? So I agree with you a thousand percent on this front.

And it's a bit of like the jagged frontier that Charlie had talked about. You have to be on that line of using not just ChatJBT, but a variety of tools to see where this frontier lies a little bit and be using it aggressively yourself in ways that in the CEO capacity would be using it into your point leading from the front in this respect, I think matters quite a bit. But I think I've come to two thoughts now, which is, you know, previously, I'm gonna try to describe this exactly, but like, There's so much hyperbole around the AI stuff that's going to change our lives, change our jobs, and so on.

And the reality right now, for most companies and most people, is this more efficiency orientation or assistance orientation to do what we were just talking about before, which is create better documents faster for different purposes, and that definitely is great and we're all kind of there. This question of this next step a little bit really is this shifting of the mindset. Into survival and being much more aggressive around going from assistant to really more wholesale reworking of flows within the company, including yourself, and that we're on the precipice of this really starting to happen in a much more agressive form, I guess, with the tools that are getting a little more mature to enable and to allow it to occur, hence your terminal point in this case.

So I kind of feel this year is the year and there is a survival element to it and You know, this whole kickoff that you and I have both done in our companies right now, I can feel it within the companies, too. Everyone's kind of geared up a little bit. So I think this is the year I need to lead from the front. In my business, in the entire interview process, I talk about becoming an AI native company, what this means, fundamental changing process, embracing technology, getting there.

And then we're just rounding out our budgets and everybody's come to me with a budget, that's the old playbook. And I was just like, no, we are not in the old playbook, you're just rolling out what you know and you're rolling out what you're comfortable with, go again. Where your it requirements your where your requirements. What are the new skills that you need in your teams?

The only area where I have a little bit more leeway of maybe the old playbook is in sales because although I've been thinking about this, like with engineering, we're gonna be able to get 10X out of one engineer and one engineer will be able to do the equivalent of 10. In go-to-market, I don't like let's say your AEs right now have a million dollar target. I don't think efficiency is going to mean that they can have a $10 million target anytime soon. Even if they're not having to fill in HubSpot and things are automated for them and they don't have to do a lot of follow-up emails, they still can do a maximum of eight meetings a day.

I don't even know if you could do that because you need a little bit of thinking time in between them. You need to build rapport. You need think through opportunities. So, I think sales is like a fairly safe.

Bet for a new career, you know, a career where there's not going to see the same level of headcount cuts as other areas, because until agents buy from agents, people still buy from people. I think there will be a point where there is going to be an agent buying from agent process as well. That, for me, is going be a little bit slower than some of the other stuff. Yeah, for sure.

But do you not think all the administration that a sales rep does today, all that's going to go away? So let's say they spend, I don't know, 30% of their time filling in HubSpot and doing crap, basically, that the 30% will be time back, thereby giving them more capacity to sign more deals. Yeah, so maybe they can go from one million to one and a half million to two, but I don't see that their quote is going to go to 10. And so you're fundamentally still just going, you're not going to see the same, until agents start buying from agents, you are not going see the efficiency gains in actual quota carrying reps as you are the rest of the organization.

Yeah, I mean, you maybe just think of something. If we look at developers more specifically, the impact there to your point, I think is probably far greater in terms of our capacity and throughput in terms getting stuff done. And if that's the case, then the ratio of engineer to designer to product manager in a squad, I think that ratio changes quite dramatically, right? Because I'm just having this conversation the other day, what is our ratio right now for a user-facing squad that develops workflows that are directly interacting with an end user.

What do we need in terms of that ratio? Is it the classic one product manager, 0.5 designer, let's say three developers, something like that. So that's no longer as an example, the right ratio is a more of like one product editor to one engineer now, because the engineer can do three times the amount as an an example.

And if that's the case, then in our hiring plan right now, we're probably need a lot more product managers. But I think to your point, thinking this through a little more AI wise right now in our Hiring Plan probably makes sense. I also am a bit concerned about product as a discipline. I think the world of product and engineering are going to blur fundamentally.

And so it's either gonna be product engineers or technical product people, because really what we found in the more that AI is writing our code, the more our engineers are spent specking and prototyping. And so we've actually changed the process. So we have a PRD, but what we have is our engineers are gathering requirements, creating the spec, prototyping it before they write any code, so vibe coding it. To our commercial team, walking them through it as if they've shipped product, getting all of the feedback then, changing, iterating, and then writing code.

I mean, that should be also a product role, but what we're trying to do is reduce the number of handovers and get everybody closer together. We're a smaller team than you are, but my mantra right now is measure twice, cut once. So I don't want that our engineers are just going through and taking the next ticket and building code. I want them to be thinking through.

Much more of an end-to-end solution, and then building the code. And I suppose in a way, this maybe kind of dovetails slightly into OKRs, because you want to get more of this holistic view on what problem are we solving, what results do we actually need, what initiatives actually make sense. Because usually those developers that are in OKR as part of that team or that cross functional squad, they're there because A, they can communicate, they think holistically, they can do this kind of systems design thinking and architecture piece.

They've got the rest of their team to do more of the so-called coding as it in that case. So that individual that sits in those OKR teams is more, I suspect, the person that we're going to be wanting more of in the future. 100% agree. I don't think it's a question of what's the ratio of designer to product manager to engineer anymore, but it's what roles do we need, what skills do we need and how much can one person do.

So that was the thing for another group that I'm in. We had a master class in AI and for me the biggest takeaway was how many handovers have you eliminated? And so I'm constantly thinking about that now. So, why don't we wrap it here and get on to our conversation with Agata Nowicka.

Maybe what I could say is that some of the key aspects that I see that those really high velocity of growth founders are now focusing on are distribution. Distribution is something that literally you need to think about from day one, something that maybe was put for post-seed stage afterwards. The first was actually the first indication of the value of your proposition. Then you are thinking, okay, now how can I start?

Working on these partnerships and so on. Whereas distribution today is much more of a kind of day one problem, just because of the expectations of investors being you need to deliver a high ROI and high velocity. And therefore you cannot have that high velocity of growth if you don't have a pipeline already. And if you don't know how to how to convert those conversations quite rapidly into revenue.

So that's the first one. The second one that is also important for early stage founders. And this is, I will actually write about it in my next newsletter. I just got this idea because I spoke with a few founders about it and I, and I can see how many founders don't really take this topic seriously is monetization.

So actually these really successful high velocity founders right now really take monetization as a day one topic to address. And again, this is not what we used to see two, three, three years ago. We would be thinking, okay, first seed your first customer group and then think of, you know, go through that freemium model was very popular. Let's go with free.

Let's see, seize the market and then let's think about monetization next. Whereas what we are seeing now and all I'm seeing definitely has a trend. Some of the lowest hanging fruits when it comes to investment for VCs are those very vertical solutions, addressing problems that are very vertical, very niche and deploying AI and that's very often just means automation. So let's say agents for healthcare workers or front workers.

For example, I spoke with a company yesterday. Compliance for manufacturing, workflow compliance, you know, agents. So what I'm trying to say there when it comes to monetization, once you raise your seed round or kind of end round. Your expectation is to deliver that ROI that basically is going to show us your revenue within the next 12 months.

That's it. You know, we've had this breakout success with Lovable that managed to become a $2 billion company within eight months. And fortunately and unfortunately, they set a very clear precedent when it comes to the expectation for growth. And that's, you know, they set the precedent for what's possible.

And now kind of all these AI-first companies are working off that benchmark. And the third one I would say is agility. Again, and this is probably the one because as much as monetization, you know, probably as a series B company, you've already, you are generating revenue, you have monetized your products, but then it's also a question kind of what else can you do? What sort of value can you create now to rapidly kind of expand that revenue opportunities of how can you monetize extra value proposition.

But the third that I wanted to mention is agility to innovate. And that's something that I would say series B plus companies might very often struggle. Because I'm talking about companies that, again, originated from this 1.0 or 2.

0, however you look at it, era, if we think that this is a 3.0 era. And Originally, you had to build a team and you had a sales team, marketing team, partnerships team, however you name it, in order to go to market. And these roles can be, not necessarily they have been taken out, but they are hugely augmented.

So you end up with one person being a marketing person, one person doing sales, or a couple of people doing sales wherever you are. And therefore, that agility is maintained for these AI-first companies. I spoke with a few CEOs of those Playlabs and they are actually struggling because obviously on one hand you want to maintain culture, you want to make people feel still that you're on the top of the world, even though you might not be. I've been a founder myself and you're always trying to drive this vision.

But on the other hand, you need to be reshaping your workforce in order to be agile. So these would be the three things that I see. So your answer was very much looking outside and go-to-market strategy. We tend to have an audience of COOs that worry about go- to-market strategy, but also worry about how to embed AI internally with both within the product, but more to get those efficiencies.

Cause this is something I'm struggling with at work. And so I'd be curious is I can see how we can have one person do an amazing marketing job. I can see I have one person doing an amazing finance job because there's like so much that you can automate. But what I am not sure about is sales productivity, because ultimately, you still need to talk to people.

You still need that human touch. Even if they're not filling in CRMs the way they used to and feedback and follow up is way faster, there's a limit to how much a salesperson can do. As I'm wondering, what's the like? It used to be, let's say, you would have a mid-market rep would do $750, an enterprise rep would be $1.

2, $1,500. What do you see for productivity once things are augmented? Cracks in this is you need to deliver not only incremental value but really significant value to your customers. If you're able to show that your product can deliver that significant value, the whole onboarding or sales process can be a little bit compromised because eventually sales is you should be able to demonstrate that value.

So you can play with the product for free. Obviously, there are some caps, but basically that's one. And then if you like the product, you can onboard yourself for that, you know, the first tier, let's say pricing yourself. So actually that role of a kind of sales rep is getting smaller and smaller and it's only allocated and I can see, I know because I spoken with founders as well that evaluate because obviously all this data in terms of signups is coming through their pipes and they are evaluating who is actually their high-value customer.

And who they will definitely want to make sure that they want to convert. So it's more strategic as opposed to you have every rep speaking to every, every person. That's what I definitely see on the enterprise sales, by the way, AI obviously is very useful for anything to do with preparation of that meeting with origination, evaluating on a much more granular level, those prospects. Do you have a PLG motion, Brandon?

It's one of those things I've been thinking about, but we have other stuff we need to get done first. It requires quite a bit of investment, but it is worth thinking about. Yeah, there's definitely a lot of mechanics, I would say. And the mechanics are very ripe, I would say, for AI to do some interesting things to make it way faster and way easier.

Because we spent a lot time and effort crafting all the mechanics around PLG automation-wise to make it all fluid in terms of users seamlessly coming on board, interacting in ways that were interesting. And eventually, we would bubble things up from an individual user to a couple different users to a team to two teams. And eventually, based on intent and signals and usage, we'd have, what do we call them, a customer advocate person contact them to help them with the underlying subtlety of trying to figure out, you know, are they ready for an enterprise contract?

And if so, how to service that to a sales rep in that case. So I have a question for you. So I've a friend of mine that is fundraising, and in their fundraise right now, they're trying to look at their plan for headcount growth over the next 18 months based on the fundraise. And the question that's rising in this person's mind is, which roles in this model should I cut?

That usually I would have stuck in that don't make sense anymore, or I don't think makes sense anymore. Which ones in that 10-count plan need to go? What's out? Okay, what stage are they?

Series B. Oh, seriously? Okay. All of our listeners, we pretend are Series B, so.

Anything that can be literally automated, that is a kind of analyst job, I wouldn't be having an analyst, basically, putting it this way. I would be definitely replacing the kind of junior jobs. And we are already seeing it, by the way. I'm not saying anything new.

We're already seeing there hasn't been a worst time for graduates right now to get jobs for that reason, because all that kind of... Administrative layer is being replaced by AI. So for me, I would be definitely, what I would be assessing my workforce would be fluency when it comes to using AI. I feel this is a kind of must.

So everybody in that company should be now retrained. And those people who are resilient, resistant, just replacing them with people that are not. I think that we've had enough time, and maybe I'm a little bit sounding harsh, but I think we've had enough over two and a half years, and especially if you're building a tech company, you want to have people who are embracing technology internally as much as building, obviously being part of the tech business. So giving them an opportunity, but every employee in the company today should have a clear view in terms of what's needed or what problem they're facing that they would like to at least be solved with AI.

So have that have that view. Mother of children who will be entering the workforce within the next five to seven years, how do I not think that they're going to be living with me? Like, who are the young people who will get hired? What are the skills that they should be training for?

I think that that topic is definitely going to be going through evolution. And as an engineer, the difference is you're not going to be required to code, absolutely not. That's like that. We're already seeing it.

This part is going to definitely be executed by AI, but it's about system thinking. So those engineers that were software engineers, now you need to be thinking more as an architect or product architect, system architect. And, you know, I cannot definitely see. Areas such as psychology, anything that basically enhances our humanity or empathy.

I think those skills, when it comes to even leadership or actually being expert and helping other people are going to become even more important. One of the topics that actually interests me, and I don't think that it's actually covered that much, it's also ethics. Because when you think about the fact that actually if things go in the same direction as they're going. A human in a loop is going to be put in the position of making much more significant decisions.

If there is a reason to have a human in the loop, this means that you will need to make an important decision. And therefore, ethical decision making is something that I'm passionate about, but I don't think that I haven't found many conversations about this topic, and I think this is going become even more important, because eventually that's what will give us power versus machine. The machine not necessarily will be ethical. Of course, maybe at some point, but I think this is something that we should be caring about.

So how do you get an organization activated to become AI first? So as within all organizations, everyone's using ChatGBT as individuals. Some folks are doing a couple of notebook LMs for various things. You have developers using a couple tools here and there.

And that's generally what's happening in most companies. But the question of becoming really AI first and using it in a more transformational way, how do start doing that? Yes, so this question kind of builds on what I said earlier, which is that AI-first projects shouldn't be locked inside the IT department. This is like a completely wrong way of doing it, and I've seen it already fail in a couple of companies that I've spoken to.

It should be a responsibility of every single, not only department, but also every single individual. And that starts with asking a question, exactly identifying your bottlenecks. And, you know, everybody, and again, depending on how big organization is, therefore, maybe it is more department when you get to a certain scale, but asking your employees to have a very clear view where their bottlenecks are. Without even knowing what's possible with AI, that would be a starting point.

Everybody has got pains. You know, Bethany, you've just talked about slides. Maybe this is enough for sales. It's like, I hate doing slides.

Okay, let's gather these ideas. Maybe, you know, once you speak with, I don't know, legal sales and maybe some other guys, you're going to discover that actually this bottleneck is... There across multiple departments or multiple people. Maybe there's an opportunity to invest into a tool or try a few tools that will be able to address that bottleneck.

Because one of the key things that, again, I'm repeating myself, is that velocity of growth. So anything that limits you to either deliver a product or grow your product or close those sales should be priority. But now... As a CEO, you don't probably know what these things are, and very often it's down to very specific and awkward almost parts of your workflow.

So first of all, it should be almost unlocked fast on the individual level. And secondly, it shouldn't be addressed on obviously strategic level, but still thinking about the top line and also about velocity of growth to make yourself competitive again with that new wave of AI-first startups as they scale. And then going back, Brandon, a bit to your question on, you asked which roles are out here for your friend. And I also think which roles are in and more in than they've ever been, which are all the BizOps, RevOps, ProductOps roles.

If I were a young person, that's where I'd be going, because I think you can learn how a business works and you can use the tools quite well. We aren't at a point yet, we're getting there, but we're not to make true agents and true automation easy. There's still a level of both system thinking and I think by the end of the year there might be quite good, because I know like NAN just released natural language workflows, but I don't know how good it is in reality, is you still need to think through your workflow and you still have to have the plumbing.

That hasn't gone away yet and I don't think it will because businesses are fundamentally complicated. A couple of things on that actually, just because I reminded myself, a couple of things about product actually and marketing. So firstly, definitely I'm seeing a trend of product people doing sales and marketing, again, through hackathons, demos, and building those communities around their products, actually it is product people are supposed to marketing people because they know so much about integration.

Actually value and also they are the most knowledgeable in terms of where you are when it comes to product because you know product cycles have become much shorter because you can do things so much faster and therefore it is down to very often not even engineers because engineers cannot very often talk about products it's product people actually who are the best sales people right now and on marketing market is again unless you can program it unless it becomes programmatic and automated.

That role of marketers is also shrinking or actually it is being reshaped. Again, because of the cycles of shipping your new features, new versions of your products are so fast that unless you have a tool that is able to release and update your customers rapidly on a weekly basis, you're going to be always playing that catch-up. And I definitely have spoken with multiple marketers over the past few months who can't really catch up. They don't know what's happening and therefore making sure that if you want to, if you see value of brand is very important, those aspects that potentially product people are not going to be able to unnecessarily add value to, you definitely should be bringing those two teams together.

Marketing has to be up to date when it comes to your product right now in order to be useful. Product marketing has become more important than it's ever been. I mean, it's always been important, but product marketing is the most valuable right now because marketing is about the repeatability in the systems and product marketing is about content and connecting it, right? We are completely running out of time.

I just noticed the time here, but nobody gets out without answering the final question, which is from everything that we spoke about today or just listening to the episode in general. What is the one takeaway for our listeners? Don't spend time rethinking, just start doing. Like literally, the time for series B companies, I think that is the most challenging right now time.

And unless you start implementing some of the ideas that we've addressed today, or actually yesterday, you're just not going to survive. The chances are that you're not going to survive because the level of ambition, level of vision, the level of expectations of investors has been reasoned so much Thanks for watching. Unless you're truly embracing AI and seeing it as a huge opportunity for your growth and for driving that velocity, you're just not going to survive. Sorry for ending it at the negative, but that's the reality, that's reality.

So, you know, the future is going to be, you know it belongs to visionary founders who are very good at fast execution and iteration. And that can translate that vision quite rapidly into execution, into revenue and growth. And there's nowhere in between. All right.

So thank you, Agatha, for joining us on the operations room. If you like what you hear, please subscribe or leave us a comment and we will see you next week.