Full Transcript for Cisco President & CPO on Rebuilding Products So Agents Can Use Them, Running a $460B+ Company in Founder Mode & the Rule of Thirds | Jeetu Patel

I actually run my company in founder mode, which is not easy, but it seems like you are also encouraging this type of founder mindset at a much larger scale. So I'm curious about how you actually make it happen. What I found is there's two kinds of people. There are people who are functional managers that are very, very good at going out and making sure that they can run their functional well.

And then there are people who are like owner mentality that run their business. By the way, you need both. It's not like one is better than the other. If you're an owner mentality, you can tend to piss off a lot of functional people. So you have to make sure that if you make the cultural change, you make it very explicit that we value an owner mindset.

We're seeing more agents using the internet than actual humans. So what does that mean to Cisco and the products from a security and safety perspective? Assume that you're a retailer and you have enjoyed a direct relationship with your consumer. They come to your website, you've got some advertising, you've got some upsell, you've got some cross-sell.

All of a sudden, you've delegated now, as a human, the job to an agent. So it's the agent that's going to talk to the website, which means that the agent is not going to be succumbing to cross-sell and upsell right the way that you are. How do you ensure that you stay in tune with what's coming next so you can catch that next wave and not be like a doctor? So I have a rule.

Hey, this is Carlos, CEO at Product School and your host on the Product Podcast. My guest today is Jeetu Patel, President and Chief Product Officer at Cisco, back for his fifth time on this show. He ran his own business for 17 years, then went inside a $63 billion company and calls himself a late-stage co-founder, 36 years late to the game.

He also says Cisco has no product managers, only feature managers, and he means that as a strategy. Here are the things we'll cover. Founder mode at Cisco scale and why it annoys functional people. Holding company to platform company after 250 acquisitions. Agents past humans in token use and what that breaks. When you can build anything, what you don't build is the job.

His 15-year-old daughter's hoodie startup and the pricing lesson in it. Let's get into it. AI helps us generate ideas, draft plans, and spin up work in minutes. But faster doesn't necessarily mean better. When everyone has their own AI tools, their own outputs, and their own context, you get fragmentation, fractured version control, scattered review processes, decisions happening in parallel, not in concert.

Rebuilding context between people and platforms costs you time and momentum. That's where Mural comes in. Mural gives teams a shared visual canvas to bring work together. You can generate structured starting points right on the canvas instead of moving outputs between tools. Keep all of your context in one place. And compatible AI agents can contribute alongside the team, bringing in context from the tools they already know.

The future of work is still human. Learn more at mural.co slash mural dash AI. G2, you are back by popular demand, man. I'm back.

How many times? Fourth? Fifth? Fifth time. Fifth time. Wow. Thank you. You've been supporting us since the very beginning. Even before you joined Cisco, you were still a CPO at Vox. But I tell you, your team lies. Because if G2 wouldn't be in this business, he'd be an actor. There's no freaking way someone would give me a job. Bald, fat, Indian guy? How many of those have you seen?

Quite a few, actually. But G2, one of the things that I want to cover with you is this concept of founder mode. As a startup founder, I actually run my company in founder mode, which is not easy. But it seems like you're also encouraging this type of founder mindset at a much larger scale. So I'm curious about how you actually make it happen.

Main page at 4:04

17 years of my career, I ran my own business. And then I went to a very large company. And what I found is there's two kinds of people. There are people who are functional managers that are very, very good at going out and making sure that they can run their functional well. And then there are people who are like owner mentality that run their business.

By the way, you need both. It's not like one is better than the other. I'm clearly the latter. I've always had an owner mentality, but the people, if you're an owner mentality, you can tend to piss off a lot of functional people. Because what the functional people want is they want to make sure that they're staying in the swim lane. You stay in your swim lane.

I stay in my swim lane. I'm going to run my function. You run your function. And then we'll just go at it that way. Owner mentality people poke their nose into everything. Like this is not working for the business. I don't really care whose function it is. I'm going to go tell them it's not working because we got to fix it so that we can make sure that we win in the market.

It annoys the shit out of the people that are on the functional side of the house. So you have to make sure that if you make the cultural change, you make it very explicit. that we value an owner mindset. And so at Cisco, we've actually done that. We've actually said we value an owner mindset. So when you think about this, don't think from your function and your silo.

I think of myself as a late stage co-founder of Cisco. I was 36 years late to the game, but it's my company. If anyone tries to fuck with my company, it's not going to be good. You know, back off. Especially at a time when you're in... wartime mode and you're at a time like this when things are moving from you're having a platform shift and everything's changing and you're refactoring your entire business it's very hard to be in a functional mode and actually win in the market so what you want to be is in complete owner mentality founder mode and so when you make a decision to go to a company

try to assess whether or not they're in which mode they're operating in. If it's in the mode that you don't actually value, don't go there, you know, because you will find yourself superbly frustrated all the time, unless you have enough agency to change the mode, you know. Yeah, I want to connect that to what you just said, war mode versus peace mode.

I mean, it seems like a couple of years ago, you went through an important org design, right? Instead of having the traditional business unit design, you went into a more platform, cross-functional platform design. Yeah. You know, that's a big change, right? Like, how do you do it? How do you do it? What are some of those operating principles that can articulate this type of change? I think in large companies, majority of the things you cannot be doing top-down dictation on.

Main page at 6:53

But there's a few things that you have to find religion in. And those things, you cannot actually look for bottom-up buy-in. You have to just go drive it. And so in this particular case, we used to look like a holding company. We had done 250 acquisitions. Everyone wanted to be a GM. They thought that the way that they would progress in their life is by being a GM.

The problem with the GM is every GM wants their own sales, their own marketing, their own product, their own engineering. They don't want to talk to everyone else. They just want to keep growing their pie. And so before you know it, you're not a $63 billion business. You're a collection of $100 million businesses. And it's incomprehensible for the customer when you take $63 billion and you divide it by $100 million because you've got way too many businesses that you can't really have your go-to-market distribution channels even make any sense of what the hell you're trying to sell.

And so what we had to do was we had to move from being a holding company to being a platform company. And a platform company means that your priority has to shift from building a feature in a single product to making sure that you actually become part of the larger platform. So the way I tell people is we don't have any product managers at Cisco.

We have feature managers. We build features for the platform. Right? And the platform is what we sell. We don't sell individual products. And if you can do that right, you can actually shift the entire dynamic on how customers consume your product. Because one of the biggest reasons that startups fail is not because they didn't build the right product.

It's because they didn't really understand how to go out and drive it through a distribution channel. And startups will fail typically, and large companies fail at this as well. If you start a company,

You might never get your go-to-market going, but if you get your go-to-market going, let's say you build a company, you built a product, and that's great, and you've actually had an extreme success. Now you say, I want to become a multi-product company. What do you do? You build a second product and a third product. If the second product doesn't leverage the route to market that you've built with the first product, you're going to fail in the second product.

You won't become a multi-product company. You'll stay a single product company. It's very important that you don't actually create multiple routes to market because you can only go out and attack so many go-to-market motions at one time. And so that's pretty important. When you go to a large company, they already have established routes to market.

And when you have an established route to market, if you say I'm going to read one too many TechCrunch articles and start to go build products in different routes to market that you don't have an attack vector into, you'll fail there. And so it's extremely important that you actually drive your entire organization to making sure that they're leveraging the distribution channel that you have.

And just because you have a sales team of 20,000 people doesn't mean that you can sell anything with those 20,000 people. You can only sell through those 20,000 people buyers that they have access to.

And for the buyers that they don't have access to, if you sell it, you are at a disadvantage compared to a startup. Because you've got all of the lethargy of a large company, but you don't have any of the agility and scale that comes from a small company. And scale that comes from a big company. So that's what you have to keep in mind as you go through it.

Because right now you are at over $460 billion market cap. Whatever today's number is. And like over 63 billion in revenue. 63 in revenue last year. This year, the guide is we'll be north of 70. So as you look at the breakdown, you said you make a lot of acquisitions. So how do you make sure that that integration goes as smoothly as possible so they actually become Team Cisco? So we have dialed down on the number of acquisitions we make because one of the challenges we had at Cisco was most people would tell me, hey, Jitu, this seems like a very important market.

What are we going to buy? And that seemed like it was the wrong question because you should never have your strategy based on what your acquisition strategy is. So now what we do is when they come to me, they say, Jitu, this seems like an important market. What are we going to build? Now, that doesn't mean that if you have a true north that you've defined and a company comes along and they can accelerate you in achieving your vision, that we would be shy to deploy our balance sheet.

We wouldn't be at all. We made an acquisition of one of the largest companies. acquisitions we've made under Cisco was under my watch with Splunk. But the strategy is not to be acquisitive. The strategy is to be very clear on what we're going to build. And if something comes along that accelerates our end state vision, then we won't be shy to deploy the balance sheet.

But otherwise, we're just going to keep building. And that also gives you a lot more like mental peace because you're not just desperate about, oh my goodness, like we've missed this market. Now we need to go out and desperately buy someone else. And the only times that I look to buy someone is if the founders are committed to taking on a larger vision where they're actually going to join me in solving the problem that we have encountered at Cisco.

When we acquire a company, the first thing we do is we kill their brand. You know, we try not to keep the brand. No more sweatshirts of your brand that come in. I mean, there's exceptions like Splunk. We've kept the Splunk brand because Splunk has such a good brand. But most small companies, I prefer buying small companies rather than large companies.

We'll kill the brand. We'll integrate it into the platform. And I take the founders and I give them a larger mission. And the mission typically would be not just go out and take your company and do something with it. I always tell the founders, don't try to make your company a $100 million company within Cisco. Try to make Cisco a $5 trillion company.

And what are you going to do to make that happen? And that's going to give you a larger aperture. Then they asked me, why should I stay at Cisco? One of the big reasons to stay at Cisco is if you stay for a four or five year period, my commitment to them is I'll make you a better operator than you could have ever been if you were not working at Cisco, you will understand at scale motions, you will understand how to go out and run some things in a way that's, you know, you're running like the world's largest startup, that the next time you become CEO, you will be far more successful than what you would have been had you not worked at Cisco.

Main page at 13:14

And that's the commitment that we make to them. And so far, I've made personally about 28 acquisitions. There was one

company where the CEO was like a jerk and didn't work out well.

27 of them worked out pretty well. And what we do right now in my team is the way I've configured my team is it's a rule of thirds. One third of the people are very good at navigating Cisco. They're consummate operators. They know how to navigate Cisco. The other third we get from the industry that are systems thinkers that know the industry and the markets that we're in pretty well.

And the third third are CEOs of startups. that I bring in to say, you know, work with the other two constituencies, but break glass and go out and do things that we would not be able to do and operate in complete founder mode. If I look at one of my key guys, DJ Sampath, who used to run a company called Armor Blocks. We got him, we killed that product, we killed the brand, but he's actually running our biggest platform in AI initiatives right now.

He's running all AI and all platform for all of Cisco. The guy was running a 20-person outfit, but he is getting to be such a great operator that hopefully he'll never leave. But when he does leave, he'll be a much better CEO than what he would have been had he not been at Cisco. And that's the promise you want to deliver. As a fun fact, I met someone today from Cisco and the leadership team that said, I had an offer on the table and I came to ProdCon last year and I heard you two speak and she ended up joining your company because of you, that type of dragon energy that is so hard to find in an arch enterprise.

No, I think that's great. Hopefully that person is still there. Otherwise it'd be an awkward moment. Okay. Oh, you're here. Okay, yay. Thank God. My God, that was stressful. But no, I think we're lucky because we've got a fantastic mission. Look, every once in a while, you find a company where the body of work that you do is going to shape the lives of billions of people and is actually going to shape the way that the world operates.

And Cisco happens to be at that point right now. We've got the right amount of scale, but we've got a tremendous amount of ceiling to grow. And I feel like this is... I'm hoping that the people that are there at Cisco right now in the next five to 10 years are going to do their life's best work at Cisco. And I'm pretty sure that they will because we will be doing some crazy stuff in the next few years.

Well, let's talk about some of the AI crazy stuff that is going on right now. I want to get your take on this. I think we are at the beginning of another super cycle. We're seeing more agents using the internet than actual humans. Yeah. And all these new AI assistants like Instinct, Muse, Crockbot, Dots. So what does that mean to

Cisco and the products from a security and safety perspective?

So let's take a look at the stats right now first. So sub 2% of the world is using agents in any kind of power user capacity today. My prediction is that in the next 12 to 15 months, you'll have about 30 to 40% of humans using AI in a power user capacity. Agents in a power user capacity. Not AI as in chatbots, but just delegating work, long-running horizon tasks to agents.

So that's number one. Number two, in February of this year, agents surpassed humans in token consumption. By August, agents' token consumption had grown 14x and now is five times that of humans. So when you start to think about the exponential curve that agents are on, these agents work seven by 24, they're going to keep working. 60% of the world's compute is today, by the end of 26, going to be consumed for inferencing capability, not for training.

But scaling laws and training still apply, and people are going to spend much more on training than they've ever done before, which means the growth rate for training has not slowed down, but the composite ratio between training and inference, inference has way surpassed training. And what does that tell you? That means that agents and AI is actually working.

People are using it. The biggest challenge that you're going to have is going to be one of safety and security. People are going to be concerned with it. The second big challenge you have is we're going to be supply constrained with not enough compute, not enough power, not enough network, not enough data center capacity for the foreseeable future, next five, seven, 10 years.

We happen to be in both those businesses.

And I think with agents, the personal agents are going to be like the next chat GPT moment. that you see in the market. But here's what's gonna happen with agents that's gonna be really interesting is

Main page at 18:18

you will have an enormous amount of pressure as a business with understanding the business model shifts that are gonna happen as a result of the disruption that agents will cause. And let me give you a very simple example. Assume that you're a retailer, okay? And you have enjoyed a direct relationship with your consumer. You have a website.

They come to your website. You've got some advertising. You've got some upsell. You've got some cross-sell. You can sell them something. You have a store. They can come to your store. Everything's great. All of a sudden, you've delegated now, as a human, the job to an agent. So it's the agent that's going to talk to the website, which means that the agent is not going to be succumbing to cross-sell and upsell quite the way that you are.

And even if it does, even if the brand gets smart enough to do that, you now have an intermediary between the brand and the consumer who is the agent. So I think like there's gonna be profound shifts that'll happen in business models that if you're, how many over here are in a company that's sub 100 people?

Okay, how many over here in a company sub 1,000 people?

How many in a company sub 10,000 people?

What do the rest of you guys do? My point over here is you will have a profound shift in business models. And whenever there's a big shift in business models, you know, when you go from direct to indirect, when there's an intermediary, when there's disintermediation of the middlemen, that creates a lot of opportunity for wealth creation. but it also makes sure that there's a lot of disruption that happens.

And you will actually start to see that be very, very prominent now in the next wave. And by the way, the third phase, which is physical AI of robotics and humanoids is already here. Like if you go to any of the semiconductor fabs, you go to Taiwan and you talk to TSMC and look at where, like there's a ton of robotics work going on all around the world.

And so what you're gonna start to see is,

We all, as product people, have to get very good in figuring out what kind of product management vocation evolution needs to happen in the world of AI that was different from the world pre-AI, and especially in the world of agents. And what you can't do is get intellectually lazy and just delegate work to an agent. What you have to get is intellectually superior by using an agent to boost your thinking.

And right now, I think there's a lot of slop And if you can do one thing when you walk out of here, don't contribute to the slop. Contribute to the distillation and clarity that people have. Because in a world where you can code anything, what you choose not to build is far more important than what you choose to build. But what you choose not to build requires judgment, it requires taste, it requires instinct, it requires a visceral understanding of the market and the customers.

And that is the job of the product manager to make sure that they can direct that. So make sure that you folks are thinking about that as you go through it. It's an exciting time to be alive in tech. I feel like if we do it right, this is the best. My daughter right now is 15 years old. If I could have one wish and what, you know, like you want her to do whatever she wants to do and whatever she does well.

Main page at 21:39

But if I had one secret wish is like, I wish she got into product. because there's so much opportunity. But she actually, it was interesting, she's 15. She has a boyfriend, which is like very awkward. You know, it's been going on for a year, but her and her boyfriend are now, he's a great guy and everything, but it's still very awkward. Two of them are starting a startup and they're like, well, we want to, she likes streetwear.

So she's going to start a startup for going out and doing streetwear. Hoodies and t-shirts and sweats and all that kind of stuff. So here's what she did. 15 years old. This blows my mind. Not particularly tech savvy. Doesn't like coding. Doesn't like any of the tech stuff. Very good with social media and creativity. Loves makeup, that kind of stuff.

Built a website for her hoodies. Designed the hoodies. Reached out to three companies. One in China, one in Vietnam, one in LA for the manufacturing process. And then was telling me about what she's going to do from a pricing strategy standpoint. I'm like, how much are you going to charge for these hoodies? It's like, I don't know, dad, like $325 between $275 and $325.

I'm like, that's expensive. It's like, dad, I'd rather sell few units at a higher cost because the profit margin is going to be higher. And sometimes if you charge higher, people think it's better. For 15, that's extremely sophisticated to understand your pricing strategy, right? That is only possible because she used Lovable. She was able to go out and build a website within 24 hours.

She was able to design the products. She was able to understand what's going on. And the fact that she had to go out and reach suppliers to figure out who's going to manufacture these came from ChatGPT. She just understood the entire ecosystem so well. And I think that's who we're going to be. That's the next generation that's actually growing up.

So we have to make sure that we can start to think about this in a very creative way, end-to-end as a business. Product people have to think about the entire business. You can't think about it functionally. If there's one function that should be in founder mode all the time, it should be product management. Because you have to think about sales.

You have to think about marketing. You have to think about engineering. You have to think about product management. You have to think about pricing. You have to think of packaging. You have to think about finance and accounting. Every single area has to be thought about as you go out and build your products. So it's a great time to actually do this with AI.

It also gives you a lot of disruptive ability to do things in a way that you might have never imagined before. And use AI not just as a productivity boost. Use AI to do things that could never happen before. And please, one thing I would ask you is every PowerPoint deck you see these days, every product you see these days looks the freaking same.

Like we don't want to become a world where AI has taken away all originality from people, right? You want to make sure that you inject a level of creativity into this thing and we can do that. So use AI to enhance our thinking. Don't use AI to be intellectually lazy. So, Gitu, one of the challenges I see in product is that your customer is not always a human.

Main page at 25:07

In fact, you just mentioned that many of your customers can be agents on behalf of humans, right? So as you design an experience, it seems like today agents are pretending to be humans and kind of hacking the system. And then sometimes they get these blocks, like Amazon recently decided not to allow Muse to go and purchase the stuff. How are you thinking about designing those type of products for both humans and agents? So this morning I was in Austin.

I just came from Austin. We actually had our WebEx event over there, and we just launched a bunch of agents and a personal concierge for every brand. So every brand can now have a personal concierge agent for every customer of theirs. Think about what agents are going to be able to do for you. It's going to be able to give us complete personalization.

If you work with Cisco, it'll give you safety and security. But the third thing it gives you is intense amount of scale. because you now are not bound by just having human caps of how many people you can go out and have a personal concierge with. Like historically, you would be able to say my top 20 customers, my top 100 customers are gonna have a named account rep, right? Over here, you're gonna have your, every single customer, no matter how big they are, how small they are, they'll have a personal concierge.

They will be able to do everything from support to sales, to consulting, to anything that you want them to do. And that's the way we have to kind of think about this. So I feel like there's a business model shift that we have to fully internalize by ourselves. And if we do that right, The amount of opportunity economically that'll exist right now is gonna be at epic proportions compared to what we've ever seen before.

But it requires an entirely different way of thinking. And it requires that you don't become part of the herd. You become contrarian in your thinking. When everyone says go this direction, start thinking about going someplace else. And be contrarian in your thinking, but don't be contrarian just for the sake of being contrarian. Start to question why everyone is going that way.

And typically, when you start to see herd mentality,

there's not enough thought going into the herd mentality. That's when you want to make sure that you create counterintuitive thinking. It really helps. And in product management, that is the only way you can get an insertion point when there's an incumbent. Otherwise, the incumbent has an advantage. But the way that you get an insertion point is because you've created counterintuitive thinking and you've created asymmetry where there wasn't any otherwise.

And that's what actually creates a level of tempo and momentum. By the way, tempo is contagious, right? So once you start winning, like at Cisco, no one thought six years ago, they thought that was freaking nuts going to Cisco. Like, what are you doing? Are you crazy? Right?

Now, there's a spring and a step in the employee base. And now people call me from the outside saying, hey, can we come join you? You want to make sure that you're starting to build things that actually get your employee base proud. And the team that does that most effectively is a product management team that gets the entire company to feel the tempo.

The brand and the momentum gets felt for the better or worse by how the product management team operates. because they create the long running true north. And when they create the true north, if it's gonna be a shitty true north, no one's gonna follow you and everyone's gonna be in a malaise. But if you create something that's exciting, everyone's gonna love it.

Now, the excuse a lot of people have is, well, I don't have budget.

Budget does not precede a vision. A vision precedes budget. And so you always wanna start with something that's a vision. It's so easy these days to build product because you can actually build a product with one or two people. and a bunch of tokens. So don't let that hold you back. Make sure that you actually have extremely discontinuous, asymmetric creative thinking that gets injected.

And just to try to bring the point home here, you talk about judgment, operating in founder mode, try to stay in touch with what's really going on, hopefully build instead of buy. But there's also a lot of legacy in a way, right? You have a core business that you have to run and keep the lights on. How do you ensure that you stay in tune with what's coming next so you can catch that next wave and not be a late adopter? So I have a rule.

Every week I have dinner with one person outside of Cisco. And the reason for that, I only have one objective in that dinner. I want to learn one thing that I didn't know from that one person. I have enough questions for that person that I can get one insight that I did not have prior to entering dinner, right? And I need to make sure that I can teach them one thing so it's valuable for them as well.

I think large companies, especially, even small companies, we get very insular in our thinking. You become the sum average of the five people you spend the most amount of time with. So keep changing the five people and make sure that all those five people are always at all times smarter than you. on some dimension, you know, because it's important that you actually learn from them.

It energizes people when you learn. It depletes your energy when you don't learn, you know, because then you start worrying about the stupid stuff that doesn't matter when you're not learning. When you're learning, you're just focused on learning. And so that's what I typically tend to do is make sure that I'm always in a learn-it-all mode, like, you know, Satya says.

And I think it's a great way to be because if someone gives you feedback about something that's not going well, just think about it this way. They care enough about you to tell you things that are deeply uncomfortable to them because it's much easier to say that everything's great. If they come and tell you that something sucks and that you're not doing a great job, it's only because they care.

And so make sure that if someone gives you feedback, that's negative, you preserve those people in your life and you actually make sure that you bring them into your life more often. You welcome them because that means they care. And if you can, and seek them out. What I tend to do is the people that are the closest in my inner circle, don't give me false flattery.

They'll tell me when something's working. They tell me when something's not working so that you can constantly make sure you're improving. Because all you have to do is be 1.27% better today than you were yesterday. And if you do it for 365 days in a year, you're 100x better by the end of the year because of sheer power of compounding. It's just math.

If you do that for two years, imagine what happens. If you do that for three years, imagine what happens. Man, if you do it for a decade, imagine the disproportionate kind of advantage you would have over someone who is not thinking in that growth mindset. It's tremendously creative. So that's the way that I would think about it. To many more years learning from you.

And thank you again for being part of this. What a way to wrap up the day today here at ProductCon. And we are going to invite you all to join us in the after party at the rooftop. Thank you, G2.