Saying yes to everything: How customer obsession built Samsara | Kiren Sekar (CPO)

Brett: Well, thank you so much for joining.

Kiren: Great to be here. Thank you for having me.

Brett: Yeah. I thought we could kick it off and you could sort of share a little bit of context, in terms of what was going on like 12 months before you all started. Samsara.

Kiren: Yeah. well, so it's actually easiest to, to fast forward a little bit even, before, rewind a little before that and then work our way up. So I started working with our co-founder, Sanjit and John at Meraki, which was a IT networking company, back in 2009. the two of them had stepped, founded the company out of grad school at MIT.

It was 35 employees, couple million dollars in sales when I joined. fast forward to, late 2012 and, we'd scaled the business to about a hundred million in revenue. Cisco acquired the company for a little over $1.2 billion, and then. Really the next phase was how do we make sure that this product and this team that we really care about is, thriving within Cisco.

And, as you know, a lot of acquisitions, they fail. Like we wanted this one to work.

Brett: So just, just for a sense of pride or there was some other reason.

Kiren: Yeah, a sense of pride, a sense of you, you care about the customers, you care about the product that you work hard to build, you care about the team that you built, and think that it can do a lot of good in the world. there was no financial, incentives or anything tied to it. It was really just you want things to succeed.

and so that first year was about how do we actually get this business established within Cisco, prevent the, you know, prevent it from getting squashed. and that was often what happens when, you know, large companies acquire small companies. And we got to the point where. The product Was selling really well within Cisco.

Brett: Was their sales team selling it in? Yeah. Or, or was more like kept

Kiren: No.

and it was actually what we said. The way this works is you take their tens of thousands of, of, of sales reps and hundreds of thousands of channel partners and get them excited about this product. And, you know, we had to overcome, I think a lot of assumptions about, oh, this is for SMBs, it's not for enterprise, et cetera.

But, you know, we found those evangelists, we got it working, we got, we showed them customers getting excited and it showed them how it was good for their business and, showed them that the, the team was a really strong team as well. And so when that got to the point that there was real momentum, you know, I think for, for myself and, and for, for Sanjit and John, we wanted to be building a company.

so we kinda made the deliberate as kind of like a probably year long, process of. Handing over the business to next generation of leaders. Some of them are still there, many of them stayed for, you know, five, eight years. and, kind of transitioned that was in year two.

So we would call it the process of evaporating, right? So how do we make us leaving a non-event so that the, the business can keep growing? And then, you know, towards the end of that, I started thinking about what, what I wanted to do next. And I ended up talking to lots of different startups and VCs and founders, and talked to early stage companies, late stage companies, companies that were just being formed.

And, you know, for me personally, I realized that there was really a couple of things that, that I was optimizing for. One was I realized how lucky I had been with the people I got to work with at Meraki. And I realized that that caliber of team was actually not, not common. And then also I saw that, you know, we had built something that.

Was really having impact at scale. And I said, you know, whatever I do next, I want there to be the potential for be, for there to be even more impacted and even larger scale. and that kind of filtered out a lot of things that maybe were interesting products or point solutions, but you couldn't see them getting big.

and then in parallel started talking to Sanjit and John about what they wanted to do next. And, we all somehow started talking about, iot, the internet of things and, and what was becoming possible from a, a technology perspective. And we can talk more about some of those details. But then we also started talking about markets and, you know, at, at Meraki, we were building primarily it was a wireless networking products.

And it, it evolved over time into broad set of networking suite. but we were selling wifi systems to everyone. Everyone was pulling, putting in wifi networks often for the first time. And you know, what we saw was that. What people were connecting to their networks and what they were doing with their networks was so wildly different.

You go to a, a bank or a hospital or a carpeted office space or a tech company and they've got great software and great tools and at the time, you know, they're starting to really rely on mobile and tablets and data and cloud. And then we'd look at our customers in construction or food and beverage production or the logistics networks and these kind of physical operations, companies.

And we saw a lot of mainframes, a lot of pen and paper, a lot of, green screens that you type on for 10 minutes to make a small change. and we started kind of connecting the dots that these are really big important industries. They didn't have access to data, they didn't have access to software, but there was this combination of technologies that we thought could really change that we said, Hey, this is something that could actually make a.

An impact at a, a, a global level over multiple decades. And, and he decided this was what we wanted to, to spend the, the next chapter of our life on.

Brett: When you reflecting on that on the opportunity set, was it in an attempt to find an opportunity to start a startup, or was it just. You were having lunch and you had noticed when you were talking to this customer, wow, that's pretty different than a Silicon Valley company,

Kiren: I, think it was a little, from column A, a little from column B. we were definitely thinking about what we wanted to work on and build next, but I think that there was also a process of just kind of playing and tinkering and looking at what was possible. And, you know, I remember, you know, John, our CTO, hacking together prototypes of Bluetooth sensors and saying, Hey, like, do you realize how cheap this is?

Like you could start putting these little chips on everything and you know, that process of pretty organic technology or explore exploration. Also thinking about the markets is kind of what I think gave birth to what we ended up building.

Brett: What were the other things that you were taking from the Meraki experience? One of the things that you mentioned was that, you know, seeing real world impacting customers, so that was a sorting criteria. Were there other things in those first handful of months that that drove sort of the, here's the customer, here's where we're gonna start, here's the culture, here are the people, sort of those type of things.

Kiren: Yeah. You know, there was, at Meraki, we were doing it all for the first time and we were doing it by intuition and, and trial and error. And when we started Samsara, there was the opportunity to look backwards and say, okay, what were the things that were really keys to the success of the company that also aligned with just how we organically enjoyed working.

And, and we wanted to recreate those, those elements. and we knew that there were a lot of things that are, are very contextual, like about the time, the, the business, the the customer. But there's certain things that felt a bit more, fundamental. And I think there's a couple of things.

that we got to be deliberate about bringing over.

One was really being customer centric and starting with what does the customer need to be successful and how do we work backwards from that? And yeah, it sounds kind of obvious, like who doesn't wanna be customer centric? But if you look at, there are a lot of really successful companies and they have different true Norths, right?

If you think about Apple at its heydays, a design first company or. Google and its incredibly high growth period, I would say was like a technology first company. You know, we have banks that are profit for, right? There are different kind of true north. I think for us it was, it was the customer.

Brett: And was that the same as Meraki?

Kiren: Yeah. Yeah.

That's something that I think we did organically at Meraki and we said, okay, that's something we wanna replicate. I'd say another was really being long-term oriented, like Meraki?

ended up being acquired. We never planned on on selling the company. We always kind of made decisions assuming we were gonna live with them forever.

And I think that actually worked really well. it's that, that long-term orientation, but then combining it with how do you optimize for speed and iterate. So I think that idea of long-term orientation, knowing you're gonna live with decisions for a very, very long period of time, but also saying how do you work towards that, that long-term vision.

With a very high pace and kind of constantly tacking and correct and, and correcting us as as we went. and then I think the th third is we really focused on quality of team and kind of consistency of, of culture. Again, we did that very organically, but said, Hey, that that worked and, and that is timeless.

And that, that was definitely a priority from, from day one at Samsara as well.

Brett: What does quality mean for you when you think about in the context of talent or people.

Kiren: I think people who are intrinsically motivated, and people who are doing something because they believe in it and they wanna make it successful as opposed to purely chasing some external reward, which everyone needs those too.

But people.

who are fundamentally, want to be building something and, and you see them actually doing that in their spare time when no one's looking.

Brett: Talk more about that, like in the context of an interview, how do you figure out if somebody's intrinsically or extrinsically motivated?

Kiren: It's great to learn about what people do outside of work, right? And, you know, are on the product or engineering side, are they trying new things and do they have, you know, novel opinions about them? Are they. Building side projects. And you know, I think at some point everyone realized, Hey, to interview, well I, I need side projects.

But then can you go deep and can you ask them more and more questions? And is it clear they didn't just do it to have something on their resume? but they wanted to create something and they get excited when they talk about it, right? And they, they can talk about what they learned and what they would wanna do differently next time.

And you know, I think just the ability to kind of keep peeling back layers and how deep someone goes is a great way to understand. were they intrinsically motivated?

or were they trying to kind of complete an assignment?

Brett: On the first point about customer centricity, what, when do you feel the most trade-offs with that? or when is it hardest to be customer centric? 'cause it is one of those things that's a little bit like apple pie.

Kiren: Yeah, I mean, it, it, it only counts when it costs you something, right? I think, you know, a a a lot of the best examples are when it feels like there's a, a trade off. We can solve a problem quickly or we can do it right. We can make the, the customer experience good, or we can hit our gross margin targets. You know what, whatever it may be, having people say, okay, let me see if I can find a clever way to.

Simplify the problem or optimize the solution and actually bridge the gap, and find a way to do both. Right? That's the, the biggest challenge. 'cause ultimately if you build something that customers love, but the economics are terrible, you're not gonna have a business. If you've got great internet economics and, the product doesn't work, you're not gonna have a business when you can actually find a way to, to thread the needle.

that's kind of where the, the magic happens. I think also, you know, finding ways to, to say yes and win, you know, especially I think once there is, once there is some momentum, there's a lot of people who say you need to say no to as many things as possible. And focus, and not saying you should never focus, but also you should say, well, for a customer, how could I sit, say yes to as much as possible?

and sometimes that feels unnatural, right? You're kind of exposing yourself to risk, right? Might. you might fall short and you know, a lot of people don't want to do that for internal audiences, for customers, but by kind of stretching yourself to say, how can we actually solve as much of your problem as, as humanly possible?

I think again, that is what kind of creates those, that outsized impact, right?

Brett: Are you often willing to deviate from your roadmap to do so, or you're constantly. Trying to balance your own roadmap with any given customer and all the little things that they're asking for.

Kiren: I think that we, we try to have a long-term vision that it changes over time, but it's, it's a little bit, it is, it's, it's more stable, right? And it's rooted in data and experience and customer feedback. That specifics of how we build that vision over time ends up being very fluid and very dynamic. Based on that feedback from customers and, and what we're hearing on, on the street.

and so I'd say the, the roadmap details can be very dynamic and very loosely held as we get more and more, signal from customers. But that vision is something that we kind of keep mo wa moving towards unless, unless or until something, you know, very fundamental changes like a new technology, like ai, for.

example, or a major shift in the market.

Brett: You know, I'm sure when you're building the company and you're spending time at Coca-Cola, pick any. Large enterprise, they're like, we really need this, or this is a problem we have. Yeah. you know, and sometimes it comes in the form of, of forward deploy engineers or sales engineering or, you know, there's, if you went to the extreme of customer centricity, you could quickly just become a consulting shop for Pepsi or pick whatever company. You could also go in the other direction, which is like, we have a vision for what we're doing. And, and you know, the more classic jobsian thing,

Kiren: I think we have always focused on prioritizing the, the products and the capabilities that are gonna be, you know, cross-cutting and, and broadly appealing. and we try to make truly one-off work very, very rare. I think there's something that that happened both at, at Meraki and Samsara that was really helpful in the early days, which was we started off primarily focusing on mid-market customers and, you know, the only way that we were gonna be successful was by solving repeatable problems.

And we found that actually this was kind of a sweet spot for us in, in the early days, the, the very large enterprises, to your point, they would have a lot of. Maybe customer specific requirements. Some of them were not even customer specific, it was just a high hurdle to get a product to market in terms of reporting and scalability and security and, and customization and all those things.

On the flip side, there were, you know, SMBs where they might be able to deploy products very fast, but they didn't have the kind of depth of feedback that you get when someone is a professional user of a product, right? Who's kind of using that product all day, every day. And so these mid-market customers with maybe a few hundred, to a few thousand at the higher end employees, they were able to give us really high quality, deep feedback.

'cause you were selling to professional buyers and professional users of the product. they didn't expect 50 custom things because, you know, they weren't getting that from other vendors. but we could actually be engaged with a, a large number of customers, really find those common threads. And when we start hearing things from, you know, 50, 70, 80% of the customers, you go build it and turn it around quickly.

And then you get very tactical feedback around, Hey, I love this, but you know, these three things. I have to go visit three different pages. I do this 50 times a day, make this easier. Or, I love this and it's solving this problem, but I have this other problem that looks very, very similar. If you could make these changes, you could solve a bigger problem for me.

And, you know, having that kind of, customer feedback, but getting it across a large sample of customers, I think really helped us build these, these foundations that were broadly applicable. And then over time we kind of filled in the, the, the gaps to really be able to move up market. And today, you know, most of our business is, is enterprise.

We still have a large mid market base as well. Starting with mid-market worked really well for us both in, at, in Meraki and, and Samsara for, for those reasons.

Brett: I guess to the point you were making it also, it's much easier to go mid-market and kind of chew your way up to enterprise than it is to start way down market and then sort of try to make the hop to enterprise.

Kiren: Yeah,

I think that, that we found that mid-market customers use the products and buy the products like a simple enterprise user, whereas in some cases, the s and b customers, they're actually closer to a consumer.

and it was a more of a fundamental difference, whereas the, the bridge from mid-market to enterprise, we could do very incrementally and they start getting larger and larger and more complex and more complex.

And then the next thing you know, you're kind of working with the, the largest of the, the large enterprises.

Brett: so let's go back to the beginning. One of the things you were explaining is your, your co-founding team was noticing some things and was partially thinking about starting a company, partially reflecting on what was happening in Meraki, different things connected to wifi networks, and then new technology, sort of what you could do with Bluetooth at very low cost.

What happened from you all sort of chatting and saying, huh, that's interesting to, like the first thing you thought about building or, or actually shipping to someone.

Kiren: So it was a little bit unusual. We started with a very high level and honestly kind of, vague, Hey, this, there's these big set of operations industries that are underserved by technology. There's something interesting here and then also saying, Hey, there's this, this new technology stack where we could take, sensors and connect them to the internet.

connect data from those sensors into the cloud and build applications on top of them, and do that in a way that was, easy to use and cost effective. And we did some, you know, we did some research looking at who was building iot companies at the time, and it was names like, you know, GE and, and these IBM and these, these large, large conglomerates.

And then there were some startups building kind of developer platforms for iot. And no one had built a product where it was plug and play and easy to use. So you had these, these companies who would go to a, you know, a, a very, very, very large customer and say, we're gonna bring in sensors from one company and connectivity from a another, and then stand up a cloud and then build custom software.

And it would be, you know, a $50 million project that meant that it was, it was very, very limited. Then there was these developer platforms where they said, here's a way you can build iot products, but the end customer was non-technical. Right? And so they weren't really able to, to engage with them. So we said, there seems like there's an opening here to build something that gives customers visibility and data, and it does it in a way that is easy to use, where kind of an everyday operations company could, could, could get it up and running in an afternoon.

Right. And that was, that was kind of the, the, the, the level of specificity as you know, there, as you can, as you can imagine, there was a lot of gaps to fill in. So we actually started by building a prototype very quickly.

Brett: And at this point did you say, we all wanna start a company together, we're gonna start prototyping and exploring, or it wasn't even that

Kiren: No, at this point, I think there, there was kind of the playing with stuff in, in san's basement. Once we started kind of building prototypes and getting them customer's hands at that point, we, we, we'd started the company and said, we're, we're gonna do this. we had a small initial team and, the first, product that we built that, again, really kind of a way to get working product into customer's hands and start that feedback loop.

It was, it was very simple, but it showed the vision of what we were doing and allowed us to get real feedback. And so what it was practically was, was a, a little box and it had a cellular radio in it, so you can kind of think about it like a cell phone without a screen, and it connected to power and then it could connect to different inputs.

And so we said, how do we get, how do we start by just getting data from the physical world into software, into the customer's hands?

Brett: Yeah. And at this point, had you spent time talking to customers or you weren't at that

point yet?

Kiren: We were, talking to customers, but we got this, this prototype done like eight weeks, for, you know, the hardware piece, the software piece, kind of getting it all working end to end.

And then we just started going to prospective customers and saying, Hey, we've built this technology, we have this vision. Can we try it out? Will we give, will we give you, will you give us feedback?

Brett: And were these like friendly relationships that you had from Meraki or like where did these mid market customers that would let you plug this stuff in come from?

Kiren: Yeah, exactly. So we had some friendly customers from Meraki that happened to be in these operations industries. and then we did a little bit of lightweight PR around starting the company. And, you know, this was a, a, a set of industries that had really not been focused on by Silicon Valley. And, you know, there's a small fraction of the, the folks in these industries who are early adopters and were reading, you know, the tech blogs and, you know, some of them knew Meraki, from setting up networks, et cetera.

And they actually were inbounding to us saying, Hey, we're excited about what you're doing, keep us posted. And so, you know, we set, I remember we set a goal of saying, how do we get this prototype installed in a hundred real customers environments? And how do we do that quickly? How do we do it across a set of, industries and environments?

So we're kind of getting, seeing the broad market and you know, I remember that we thought that temperature sensing was the thing that was gonna, was gonna work. That was gonna be the first product we said. Think about all the industrial refrigerators. If one of these things fails, they've got pharmaceuticals.

They might have, you know, lobster and caviar in them. Like of course you'd wanna be able to monitor those. And, you know, some of the early customers, they were trying this out. They weren't paying us any money. They were just curious and wanted to help us out. They thought it was cool. And I remember them saying like, Hey, this is really cool.

I can, I can see it from my phone, it can text me when the temperature changes, but you guys know, these things don't really fail, right? Like, this is not a real problem. but then, you know, one of them, it was, a cowgirl creamery a cheese producer in Northern California. Said, you know, where we lose product all the time is, is in trucks, in, in, in transit.

is there a way we could put one of these on a truck? And so we found a way to, you know, connect it to a, a vehicle and get power from?

the vehicle. And, that kind of led the way to our first successful product, which is saying, how do we track a vehicle? And then how do you track its, its cargo, it's, and, and starting with kind of refrigerated, refrigerated transportation and, and then along the way finding all these other problems that, that customers had, had no idea where these things were.

they had no idea if they had engine faults, et cetera. All of these kind of became clear by actually getting product in the customer's hands and, hearing about what their real problems were and then trying to solve them quickly.

Brett: So before you had the eight week prototype, you didn't do more classic sitting down with end buyers or customers doing needs finding. How big of a problem is this? Where's the urgency? You more sort of had a sense that there's some opportunity in this noom form factor of technology. Let's go build this prototype and let's learn by bringing it to the customer.

Kiren: Exactly. Exactly.

And you know, as I was talking about earlier, the, the vision that I think we had a lot of conviction in was when you bring data and visibility to an environment where there previously was no data and visibility, it is just useful, right? We've seen that everywhere. You know, retail, financial services, healthcare, pick your, pick your industry.

And so that, we felt pretty, pretty clear about what it looked like. And is it, you know, temperature or trucks or warehouses or. you know, factory equipment. That was the piece that we had to, to find out by getting product in customer's hands. And I think that had we done traditional market research, I don't think we would've ended up at the, the same place, or certainly not as quickly, but actually showing customers what was possible, showing them a totally different user experience, help them provide the feedback of, Hey, I get it, and this is where it can be most useful.

Brett: before you had the cowgirl creamery insight? What, what were you seeing in customers? Was it a bunch of, oh, this is kind of cool. that's interesting, but it's not solving a massive problem. Like what, how would you describe the before and after and what you were seeing and hearing fro customers?

Kiren: actually, it was not as linear as we tried refrigeration and then we tried refrigeration in trucks. I mentioned we were, you know, doing the refrigeration piece. part of our v vision was also that. We're not trying to find a single data point for a single industry. We're actually trying to build a platform.

and, you know, thinking about, you know, what did a company like Salesforce do for, for sales teams and for, for customer data in the early two thousands, how do you have kind of that equivalent for, the world of operations? So while we were kind of running this, this loop with, refrigeration, we were also talking to water utilities.

We were saying, Hey, I get, you know, leaks and it's hard to figure out where these leaks are. Or, you know, when I ain't ringing, when I need to run pumps at certain times of day, the energy is incredibly expensive. Is there ways that I can, I can smooth that out. we talked to, you know, we actually thought that that vehicles could be interesting and we talked to, transportation fleets about, you know, how do they track their trucks and what does that look like? So we were kind of running these experiments in parallel and then kind of seeing where the biggest engagement was, was occurring, where were customers saying like, Hey, this is something that's really impactful.

Like, I need you to build this. And that kind of created, created signal. it also helped us make sure that, you know, as we started building product, we were building something that could be broadly appealing and we weren't building something for a very, very niche audience. so then when we actually, you know, finally launched our, our product, we actually again had a, a, a few different use cases that, that we were selling into, refrigerated trucks being one of them.

But then we kind of kept running this, this feedback loop of kind of having a broad aperture getting signal. Doubling down where there was customer engagement. and, you know, I think that really helped make sure that we were building towards that vision of a broad platform for a broad set of industries and not hyper focusing on something that maybe ended up being, limiting over the long run.

and actually it's interesting, in our space, most of the technology providers, again, weren't Silicon Valley companies. They weren't even big tech companies like the, the Microsofts of the world. They were a lot of these, point solutions that were building technology for a very specific problem in a very specific industry.

But they were never able to get enough scale to really invest in r and d and kind of build the kinds of things that you or, and I would consider a, a modern technology platform. And we said, Hey, actually having that breadth is gonna be something that is, is really important. So something. From early on, we made sure we kind of balanced, you know, getting feedback from large sets of customers in large markets while also solving the problems in front of us where, where customers?

were saying there there's a need.

Brett: So what happened after you came across this insight from Cowgirl Creamery? Did you then decide to narrow in that use case Ke keeping in mind that the goal is, is always to build a broad platform, but that you wanted to nail that use case, drive repeatability around customers, and then land and expand.

What ended up happening, you kind of got a prototype working, they were seeing a lot of value, and then what happened?

Kiren: It was, it was a balance. On one hand we were saying we're getting signal here that, refrigerated, transportation is a problem, so let's go make this work well. But in parallel, while we're doing that, we're having our product on. Trucks. What are the other problems that people are having with trucks? They don't know where they're, either, you know, they have no way to track them or, you know, they have a, a GPS tracker that reports location every 15 minutes. We know from getting an Uber that you can track a vehicle in real time, we know how useful that is. So you're not waiting out in the rain for your car.

We should build that into the product, and that's something that probably customers other than cowgirl creamer are gonna be interested in as well. and then in the process of doing that, saying, Hey, what else happens with these vehicles? Hey, we're looking at 'em. And a bunch of them have their check engine light on and the driver's driving with it anyway.

That doesn't seem good. Let's figure out how, a way to connect into the diagnostics port and understand if there are fault codes. And, and so it was that, that process of continually pulling the string and seeing. What more can we do? And then who would this be useful for? and so as we were solving problems for cowgirl, we were also talking to lots of other fleets and operations businesses, trying things, and then kind of doubling down on, on what was working.

Brett: What you're sort of hitting on this, but what were the handful of things in that year after the cow cowgirl creamery insight that ended up being a part of the early sort of saleable product?

Kiren: one of the things that we did that was really helpful was we started pitching and, trying to sell the product to a pretty diverse set of industries early on.

And, you know, we were pretty intentional about kind of looking at the, the pie chart of, of industries we were engaged with. And, you know, again, I think we kind of did this intuitively at Meraki at the time, everyone was, was trying to deploy wifi networks. Some of our competitors were saying, Hey, we are gonna be the wifi provider for healthcare.

And we're saying like, now we wanna build products for everyone. and we ended up with a very diverse set of customers. We were, I'd say intentional about doing that at, at Samsara as well. And so, you know, we would, I remember we had, you know, transportation Tuesdays where every Tuesday we were focused on talking to, to customers in, in, in transportation verticals.

you know, we would otherwise be talking to food and beverage or construction firms, these, these different industries that had fleets and we thought could benefit from what we were doing. And then, you know, we were really kind of running the feedback loop of, okay, you're excited about this. What do you need in order to, to deploy it and pay for it?

By just running that feedback loop across a wide set of customers. We went from a million in a RR to to 17 and and then 70 the year after that. Along the way. What was fascinating was we were getting more and more feature requests that were related to, to tracking, trucks and primarily kind of around location and fuel and, and compliance.

and, you know, ended up taking us many years to build out all the features to be able to really sell to the whole market. But in parallel, we, started getting customers asking us, Hey, is there a dash cam that works with, with your, with, with your system?

they were asking us, Hey, like, who do you recommend? And we said, well, hey, if, if we built one would, would you wanna try it out? And a lot of them were saying yes. And this is an example I think of, how we were very conscious about trying to balance that, that depth and breadth. and it seemed like a, a really useful thing, right? To be able to have video

And again, there was that technology lens of, there's a couple of companies doing this. they're pretty clunky. They were built on the kind of two G and 3G networks. They were where you got. Very low res video. they're hard to use. But we saw that actually the, the technology that was being really being driven by mobile phone adoption was making video really high quality, really inexpensive, and you could actually start streaming it over the air.

So we prototyped a, a dash camera that was kind of integrated in the system. And then we started, giving it first to some of our existing customers and then letting some of the customers who had were asking us about it, gonna try it out. And they're saying, wow, this is exciting. And so, you know, we could have easily said, no, we're not gonna do that.

we've got this big list of of telematics fleet tracking products to, to go build. But there was a combination of, you start hearing the same question from more and more customers and then seeing that actually there was a potential technology inflection where you could build something that was. Truly different from, from the rest of what was on the market.

and so it was actually in that, that second year that we built our, our, our dash camera, and it's since evolved into this product that basically helps customers, really understand their, their risk and, and dramatically reduce it. and it's actually now our, our number one product?

by, by revenue and growth.

But it came by running that, that feedback loop and kind of being willing to find a way to say yes and figure out a way to, to get something done When we were hearing signal from customers and saw a way that we could, we thought we could do something differently than it had been done before.

Brett: When you're actually spending time with customers to un unearth something like, Hey, there might be an opportunity in dash cams, is it just you have a very close collaborative relationship with your customers and you're hanging out at their office, and what else is on your mind? What are you thinking about?

Like, how are you actually getting this information from the customers?

Kiren: Generally focusing early on, on building relationships with customers and you quickly figure out who are the early adopters, who are excited about what you're doing, want to be part of it, and it actually, you know, makes them more engaged and more loyal when they get to, provide feedback.

And, you know, as long as you actually act on it and, and, and, and do something with it, it, it creates this, this virtuous cycle. Part of it is just being willing to, you know, get on a plane and, most of our customers are not in, you know, San Francisco or New York or Boston, so we're going all over the place and it's, you know, indirect flight with a layover and then a three hour drive.

We had a, a customer advisory, group from, from very early on. We'd invite them to, to San Francisco, to our office and spend two days with them sharing our roadmap, getting our feedback, getting their feedback, sharing ideas, kind of doing allergy tests of, Hey, show of hands, how many of you would find this useful?

Like, here are three ideas. How do you rank them? Right? Getting that feedback with people who are already using the product, who actually have. kind of a vested interest in the company's success. And then there, there's trust. We did that very early on and, we, we still do it today. That has been very, very powerful.

Brett: Maybe we could go back to something you said a second ago. I think a lot of the conventional advice is you wanna start by my, by monopolizing a very narrow ICP, and then once you've dominated that, you move on to the other.

Once we do that, then we're gonna go and try to launch our financial services vertical, and then we're gonna launch, et cetera, et cetera. And it seems like one of the things that, that you all did was you wanted to be very broad from day one, and you didn't want to sort of take that strategy. 

Kiren: Going broad and, you know, finding those 80 20 solutions that actually work across verticals. I think it worked really well for us, both at Samsara and, and Meraki. I think part of it was actually, you know, working backwards from saying, Hey, how do we make this a meaningful business at scale? Right? We wanted to do something that could be a, a large self-sufficient company that could build lots of things and get them in front of lots of people.

working backwards from that, like you can kind of translate that into, into revenue goals, right? And could say, well, if we want our revenue to look like this and, you know. 10 years and five years and two years in one quarter knowing that the long term, there's a high error bar, but like, this is a vision of what, what success could look like.

You start working backwards and then you think about the, the size of the market and you say, well, if I sold this fleet tracking product to every cheese company, you know, how quickly am I gonna gonna run out of market? and then in parallel thinking through, well, if you hyper optimize there, there's a risk that actually you build something that is, is so narrow that it ends up not really providing leverage in other, other industries.

Whereas if you take the other approach and say, I wanna actually, engage with a broad market, you start end up finding these, these cross pollination opportunities where you're getting feedback from one industry, something that other industries aren't even asking for. But then you build it in the platform and all of a sudden everyone starts using it.

I'll give you a couple of examples. we were working with a, school bus fleet early on and said, we need to track our buses. We have no visibility. What's going on? We don't know if they're properly maintained, if they're late, if they're on time. but it's not just about us. We actually wanna be able to share location with, the kid's parents.

And yeah, we, we saw what you can do. It's real time. Like, like Uber, we wanna give that experience to parents so they know, you know, when to meet their, their kid at the bus stop. And we said, okay, good idea. So we built a simple, location sharing, system where you could share the real time location of a, of a school bus with someone who isn't a user of our platform.

They get a live tracking link, kind of like you would get with, with an Uber or Lyft. We started showing that to other industries and all of a sudden they're saying. Oh, I can tell you all the different ways that I would use that. I'm a construction company and I pull a dump truck up to a site, and now that truck is sitting around for half an hour waiting for a crew to come move that pile of debris and get into the truck.

People are sitting around, everything's getting gummed up. If they could have a tracking link, then they could have everything ready right when that truck arrives. you think about all the different examples like that, we would not have prioritized that feature if we weren't talking to lots of different customers.

but by finding areas where certain industries were giving us feedback and kind of applying the judgment of this could be broadly applicable, ended up being just an amazing source of kind of, product ideas that, that benefited everyone. Our mobile apps was another one. we started building mobile apps for, for drivers and. early on, it was customers who had very specific compliance requirements. Hey, they need to complete these forms. These forms are getting, mandated to move from paper to electronic format. you know, can you, can you build a, a, an app for drivers to do that? We started thinking about, well, if you put a smartphone in the driver's hands, what are all the other things you could do with it?

Right? All of the other paperwork they're doing, you could, you could digitize, you could start attaching photos to things. They can be, messaging with their dispatcher or they can, you know, they can get training and coaching. How do we just focused on.

feedback from one industry? We would've ended up with a compliance app.

Now we've got this, you know, really broad set of, of, of tools for frontline workers. And so I think that, you know, by starting with, okay, we have the ambition of we want to build something that's gonna have really broad impact. Engaging with the broad market, we're able to kind of pick and choose what are the, what are the things that we are gonna build that are gonna have the most impact.

and ultimately, if you don't meet the needs of an industry, they're not gonna buy your product, Right.

So it's not like we could go so broad but not actually connect the dots all the way to making the useful product. and, you know, revenue is a, a way to keep yourself honest there, but I think being able to say, okay, we're, we're trying to maximize revenue growth, but we want to be doing it in a way that is pulling all the best ideas from all these different types of customers.

Ended up being really powerful.

Brett: What else? What else about. The setup allowed you to grow so quickly in a market that I don't think was a traditionally good buyer of software, right?

Kiren: Ease of use and time to value. And that's still something that we really focus on on today.

we really wanted the product to be something that, you know, in 15 or 20 minutes you could get it set up and, and see the value. If you're trying to deploy it across a large organization, of course that's gonna take time and change management and we've gotta physically stick, you know, boxes into, vehicles, construction equipment, et cetera.

But if you've got a, a small branch or division, can you do a, a, a, a free trial and you set it up in an afternoon and see why it's useful. that was part of our sales motion from day one and actually dictated how the product had to work.

Brett: Hmm. So that, did you, how did you come up with that?

Kiren: you know, that was actually a big part of Meraki as well. Most of the networking industry at the time it was, it was pretty complex. And if you wanted to deploy an enterprise grade network, you needed network architects and design. And there was a lot of configuration and customization and it was a pretty big expensive project.

we said we're gonna make a network that's simple, that's plug and play, that's still usable by a business that has security and management, but you can actually get it up and running in an afternoon. And that was really what created the, the core product fit, market fit at Meraki. 'cause it turned out that it was not just the, you know, big slow moving it groups who were deploying networks, it was everyone and people didn't have time for a multi-week or month project.

So we saw that work and saw that it was a really key part of, of rapid adoption and all of us experienced it as consumers as well. Right. You know, and I don't have tolerance to try something in our personal lives if it's gonna take weeks to set up. So that was kind of, again, one of those things that we sort of felt fundamentally we should build things that you can get up and running quickly and, and see the value quickly.

we also focused on how do you actually make it economically valuable to customers very quickly. operations industries traditionally have not spent a lot of money on, on software and technology, but they have massive budgets, right? Our customers, they spend tens, sometimes hundreds of millions of dollars a year on just fuel, right?

we're really trying to, say, Hey, if you take a little bit of that, put it into technology, you're gonna see big savings. But they have thin margins, big budgets, thin margins. You're not gonna spend money on a product if it is not economically valuable, quickly. that was actually a really great forcing function. Like I remember in in our lives building products for IT organizations. There were some customers who would buy something 'cause they thought it was neat and the IT department had discretionary budgets.

They're technologists, they thought that something was cool. And it's actually really dangerous because as a startup you think, oh, I'm building something that is getting traction. But not very many people make decisions like that. But you can kind of fool yourself into think they can get a product market fit and then all of a sudden you can't, you can't scale and customers aren't renewing in operations.

Customers don't part with their money if there isn't a very clear ROI. So it actually helped us. Very early on, hone in on those use cases that customers really, really cared about, that provided real value. So, hey, I tried this and in the first week, we had a, a truck, someone left the door open, we will have, would've lost $15,000 worth of cheese 'cause of your system.

I didn't, or, you know, we were able to cut our accident rate in, in half and we're able to take a chunk of that?

savings and, put it into a driver rewards program. And we've lowered our, our turnover rate, those types of economic impacts. Actually making them clear and fast was really important I think in, in, in getting product market fit.

And then in ultimately growing from there.

Brett: When you thought about pricing your product, was it simply an ROI model based?

Kiren: Yeah, we would always kind of start with what is the customer's next best alternative. And in some cases there was nothing on the market. In a lot of cases there was a competitive product. I think most of our products, we weren't necessarily the first, we just had a different approach. And so even if we had a, a very different approach and it was more modern, easy to use, et cetera, the customer does have a, an alternative which is, Hey, I could buy this, legacy app.

Maybe it doesn't do everything that, that yours does. And so we think about how do we operate in, in that market context and, you know, what is our ability to price versus the the next best alternative? And, you know, early on it was actually like, how do we make this an easy decision for customers, right?

Which means if there's existing more opportunity, if there's existing vendors in a space, maybe there's the really expensive one. There's a really cheap one. Aim to be in the middle, right? And if we can actually deliver a better product than the most expensive product, but at a better price, that's a lot less convincing.

You have to do. As we've grown over time and started building more things that maybe didn't have a direct analogy, you know, we've gotta do a bit more discovery and it gets a bit more nuanced. But we thought about are the unit economics sustainable and is it a no-brainer for the customer? And if the answer is yes, go sell it and grow.

I'd say, you know, even now we're not focused on extracting every dollar possible. We're focused on actually, how do we have customers say like, that was the best technology investment I ever made. And that means leaving some value on the table. But the market is so big that we end up doing better by, you know, optimizing for, for growth and making an easy decision for customers while also making sure that we've got healthy unit economics around along the way.

Brett: Is part of it that there's, that there is a surprising amount of anchoring in the customer's eyes. Like I, I assume if you looked at a, a, a, a few of your early products. They were vastly better in every dimension, which ultimately meant that there would be more cost savings than the best available alternative.

and so one would imagine that you should charge more o other than wanting to make it, I guess a no brainer buy, do peop do end buyers. Just think about it, you know? Yeah, your dash cam is vastly better. It's gonna save us all this money, but this is what it costs for a dash cam and it's hard cognitively to get off of the anchoring of existing products in the category.

Kiren: There's a little bit of anchoring. There's also, I think, skepticism of customers who have been burned in the past. and you know, I think as we have developed more. Reference customers and built a brand and, and, and reputation, you know, that allows us to have more, pricing flexibility. But, you know, ultimately we're still very much just focused on, you know, what is gonna drive the most growth.

and finding that, that right balance.

Brett: What about, is there anything you figured out in, in pricing a platform where I assume many of your customers buy many of your products and it's all integrated and that's one of the special parts of the business. Is there anything about how you price that or it's like, you know, the average dash cam is this, so that's generally what we're gonna charge you for that part of the, for that product.

Kiren: not always something that you can quantify and say, I'm gonna get a higher ROI by having these things connected. But they say, Hey, if I've got a single set of users, all of the data, ties together. all of the APIs and integrations work together. actually the products can, can talk to each other and, make each other better.

That provides a ton of differentiation. so we don't charge for the platform itself. Exactly. It's really, Hey, the reason you would like to buy these five products from Samsara as opposed to one product from Samsara and four from, you know, four other companies, is, hey, each of these products kind of stands on its own as, as, as best in class.

But then you put them all together and it's, it's way better for the customer and the end user.

Brett: In talking to different folks on your team, over the years, I heard some people talk about the way that you goal around revenue, at least product managers that seemed very customer focused, but also many of them I think had, or at least in the past, had a revenue number that they were trying to hit for their line of business.

Kiren: We've tried kind of different goal structures over over time, but I think one of the things that

we kind of figured out early on is that especially for newer products, revenue is the best metric for product market fit.

Brett: customer value.

Kiren: Exactly. it's a proxy for, for customer value.

and impact.

And if we've got high revenue growth with solid unit economics, that means that things are working. And I think that that has really helped product managers, especially product managers who are working on new products and new apps, say, Hey, I know that you know, good looks like this, great looks like that.

And this is what off track looks like. and actually it becomes, I think, really fun to be able to have a, a number that you can hit and, and beat and get over, and then everyone kind of, everyone appreciates that, but it also translates really well. I think, you know, there's a lot of areas where that metric worked really well in the early days, but, you know, when we look at our more mature products, there's a lot of different factors that go into a customer, deciding whether or not to buy one.

And actually you could have revenue look healthy, but the customer experience is degrading and that's gonna catch up with you. And in a few, quarters or a years, you're gonna see that in, in, in revenue. On the flip side, you can be building things that are truly delighting users and you can't directly attribute them to revenue growth because that might be tied to did marketing generate enough leads and we have enough sales capacity, et cetera.

But you can know with high conviction if you're seeing usage increase, if you're seeing engagement, that's gonna translate to revenue over the long run. So I think that it's a really valuable tool, but you gotta use it in the right, right, products in the right scale.

Brett: What are the things that you thought would generalize from Meraki and did not?

Kiren: Yeah. Meraki, we sold through channel partners and, it turned out that that was the way that people like to buy networking. You wanna buy your networking equipment from the same person who's selling you your. In your laptops. And those reseller, partners were a great way to go to market. We tried that at, at Samsara and, a lot of actually specific partners that would been very successful with at Meraki didn't work.

and we found that actually these customers wanted to primarily buy directly from us and have a relationship directly with us. So that's something we, we tried. It wasn't working very quickly, moved on. So I think the things that really translated were around, you know, culture and values and, and what I was talking about in terms of being customer centric, being long term oriented, having this, iterative approach to building and then certain things around the user experience.

Brett: As someone who, who did a lot of marketing as a non classically trained marketer at Meraki, how did some of the marketing ideas apply? And, and I'm, I'm interested in the fact that I assumed there was this element of category creation and you had the chance to sort of build something new at that time.

Did you think much about how are we gonna educate, how we're gonna create a movement about the, around this sort of, those types of things?

Kiren: I think the, the aspect that really carried over was the idea of showing the product and then letting customers try the product themselves and optimizing everything around that. and. Both at Meraki and, and, and Samsara. And, and really to this day, we start with a demo, right? And that can be over, zoom, it can be in person, but we start with, we're gonna show you a live working system.

It's not a it know, a sandbox, it's not Figma. We're gonna show you live, a live system. when we do our customer conferences, we're only demoing live products. And there were some, you know, people came in from big companies and they said, well, what are you doing? Like, it, it, it could, it could go down. It might not work.

You should be using, mockups. We're like, no, we, we show the real thing. and then making it easy for customers to get it, get their hands on it and try themselves. That is the piece that, that really carried over and I think is a core element of, of how we go to market.

Brett: What is the reason for that?

Kiren: I think there's definitely some, counter positioning against, a product where you can't demo it because it's gonna be a a nine month deployment. I think some of it is actually just, it reflects the, the pride in, in, in what you've built. And, you know, I see a lot of, startups today that lead with demos and it's awesome.

And, you know, we have our, venture program at, at Samsara where we fund a number of startups who're building in the operation space. I get to see a lot of the, the founders pitch and the vast majority of them are leading with demo. And you can see the pride they have in the product that they've built.

And some companies I think lose that as they scale. But there's a handful that, that hang onto it. I think it really. Maintains that connection with the people who are building the product and how it's being introduced to customers.

Brett: Is there anything else you all have done that you sort of pioneer on the go to market side?

Kiren: You know, I think, in, in the early days, one thing that we did that was really helpful was we focused on building a scalable sales motion. And, you know, we never did the pure founder selling, like we always had?

a, a sales team, and it started off very small. We said, you know, the way to to, to make this successful is to build something that we can sell through a sales team that we can scale.

Brett: So you didn't do the normal thing. We get to 5 million with founder led sales, then we hire our first sale AE or sales leader. That's not how it works.

Kiren: That's not how it worked. In fact, we hired our first, a DR or SDR before we had a product to sell, and their job was to book demos. so myself and, and one of our product managers could basically do live demos of what we were building and get people to say, Hey, I will try this out once it's ready. and, and you know, that allowed us to say, Hey, go call a hundred water utilities and see if, this pitch is resonating.

Right? Go call a hundred food and beverage distributors and see if this pitch is resonating right. And you can see at the very top of the funnel where there's resonance. So when we got to the point where we thought, Hey, we have a product that we can try to sell, we started with a, small inside sales team and then a couple of field sales reps and said, okay, we are going to build a small, but end-to-end sales machine here.

Something that if it's working, we can scale it. And so that means we start with generating leads, we then turn those leads into demos that our, our sales reps are doing and turn that into a free trial and then turn that into, a contract and, and then a, a deployment. that process requires you to actually, get a bunch of fundamentals, right, that you might be able to work around if you're founder selling.

So, for example, the product needs to be really easy to demo 'cause you need. The sales team, not just the people who are setting the roadmap and working with the engineers every day to be able to just, you know, click, log into a dashboard and, and, and show the the newest features. you need to figure out what are the pain points that are causing people to pick up the phone in the first place and take a meeting and then take that meeting and translate it to, you know, a to a, a trial and eventually, a purchase order.

And you're getting that data, you know, really kind of at, at scale. you're, you're making sure that you have a value proposition and, a pitch that multiple people can deliver, right? Not just the person who has the, you know, the, the, the founder zeal, right? And then when it's working end to end, what's amazing about it is you can say, great.

Now let's go double this house the sale team. Now let's just go double it again. and so I think doing that early on. Helped us get a lot of those fundamentals right. similarly, you know, on the, the post-sales side, we actually said early on we're gonna start with that customer success because that is a very easy way to paper over things that make a product hard to use.

we eventually added customer success because it turned out that there were a lot of things that we really needed to help customers with in terms of change management that were not, you know, technology centric, but by saying, well, we're gonna have a, a support rep who can take tickets, but then we need to actually, not work around them, but get them resolved in the product itself.

Brett: something you just hit on, what did you figure out about change management?

Kiren: we figured out, you know, especially as we started going up market. That it was a real concern for customers. And if you think about these, these customers, they might say, okay, this product looks great. I've gotta go figure out, first of all, how do I go, you know, install devices in, you know, maybe 5,000 vehicles or assets that are all over, the place and they're always in use.

And if they're not in use, I'm not making money off of them. This new system's really easy to use, but I have, you know, hundreds or thousands of frontline workers who I need to train on it. actually being able to make change management?

part of our pitch was super helpful, right? Especially in, in the enterprise, being able to say, we have the technology, but we will also help you be successful with it.

And, you know, you can get this done end to end in. X months, right? that means training thousands of people, you know, getting thousands of assets connected. Here's someone you can talk to who's gone through this, who looks like you. and actually making that kind of peace of mind, part of our value proposition, I think was something that emerged over time.

You know, especially as we really started to go up market. And I think, you know, that also really helped establish that that relationship where we're not just trying to solve the customer something and move on, we're actually trying to partner with them for, for many, many years to come, which then turns into product feedback, which allows us to build more products and then sell them to these customers as well as new products.

Brett: What are some of the tactics that have been interesting that have really worked, like, I think it's, it's, it's such an interesting point where, you know, you have drivers of these trucks that have been doing this a certain way for a long time. You have so many stakeholders that have been doing their jobs in a specific way for a long period of time, and once you bring data and analytics and real time feedback, someone then has to go do something with that to sort of have a positive impact on the business.

Kiren: Yeah. You know, some of it was actually taking what we heard back from customers and then saying, how do we get this out to everyone? So for example, when we started with our, our safety cameras, you had some drivers who had been driving a truck for 40 years and they say, this is Big brother, I don't want this in me, in my face.

And we actually had customers say, oh, you know, what we did was we started by first putting it in the, you know, the, the VP of Safety's own car so they can see that we're doing it ourselves. Then we start with putting it on the five safest driver's, cars, the absolute best ones. And invariably there is an accident or a near miss where that camera shows that they did the right thing and, and they're not at fault.

And then you take that video and then you share it with everyone, and then you go roll it out. And then as you're rolling it out, you also roll out a safe driver bonus program where we say, we're gonna save all this money with this technology. We're gonna take a portion of it and give it as, bonuses for, for drivers who have, high safety scores. That's something that, you know, we started hearing these anecdotes for customers and kind of turning them into, into a playbook. and now it's something that we have like, you know, really, well packaged up that's easy for us to, to onboard customers to. but that kind of keeps evolving as we. Get new ideas from customers as we build new products.

and we almost think of that as like, it's not a technology product, but it, it is like a product in that way of we're hearing about what works. We're finding ways to make it scalable and then baking it in.

Brett: What are the things that you've had to change your mind about as the company has gone sort of in each chapter of growth?

Kiren: how we use. Revenue to, to, to, motivate and align product managers is something that's changed a lot. Early on, revenue was a really, really good indicator of, of product market fit, and it was something that, you know, we could kind of rally everyone behind. We got to a point where actually we had really strong product market fit across multiple products. The customer experience was actually starting to, to degrade in some ways. We had some rough edges to the product, which kind of makes sense. when you think about how fast we grew, but it wasn't hurting revenue growth. And we actually had to say, okay, no, we need folks to just really focus on customer experience and, obsess over that.

And we got to a point where we said, actually. Our customer experience is, is great. We talk to customers, they're really, really happy. We're still focused on how do we kind of get from 99.9 to 99.99%, and they're saying, we'd love you to go build these new products and solve new problems for us. We need to reorient folks back towards saying, okay, what are these new things that we can, we can build?

And, you know, how are we either, you know, driving adoption or, or, or driving revenue growth? So, you know, the, the thing that we need to focus on, I think has, has changed over time, and I'm sure it'll continue to change. It's also different in, you know, different parts of the business, different products, different geographies.

so that something, that thing requires us to, to hit refresh. I think, you know, an area that, that we sort of had to go through a bit of a, a, a journey on is, how we think about. Organizing and communicating our, our roadmap. And when the company was very small, it could be very dynamic. And everyone's, you know, literally in the same building, having, having lunch together, kind of typical sort of startup, feel and the superpower of startups is you can be super dynamic and everyone stays in sync.

I think we went through a period where we had grown from, you know, a few hundred people to a few thousand people. We're seeing a lot of strain from that. So we put in a lot of, process to have much more predictable, well-defined, roadmaps. And it solved some problems, but it actually meant that we were slowing down, right?

We weren't able to iterate as quickly. We weren't able to ship as quickly and said, actually a lot of what we were doing in those early days of how do you get something to a demo, a demoable form quickly? How do you get in front of customers? How do you run that feedback? That's what we need to be doing now.

We need to figure out how to make that work across, you know, a multi-thousand person company, but we need to be doing that and not, you know, acting like IBM. Right. and so that's an area that I think, has been really exciting, you know, over the last, you know, you know, period of time kind of figuring out how do we pull in those elements from, from really operating as a startup with customer scale and, you know, scale of, of engineers and product managers and designers.

And actually what we've been able to ship has just exploded. Right. Which has been really, really fun.

Brett: What's just, what's happening with, with LLMs and, and more broadly with ai. Does it feel like the early days of Samsara when you had all these, converging technologies. at least on the hardware side, dropping costs. You could do all sorts of incredible things with Bluetooth, et cetera. Does it feel like that or it feels more profound or different or?

Kiren: A hundred percent. I, I think, you know, when I think about Samara's product, even today, you know, with, with, more and more AI built into the product, the data that we're collecting from our customers, operations, it is like an iceberg. We're helping customers action on just the very tip. And, you know, they've had really amazing results with that, but there's so much data in the system that is still untapped and LLMs and all of the advancements in AI are just radically changing.

How we could make that, that data useful for customers in a way that doesn't require them putting more staff and personnel behind, understanding the data and acting on it. And, you know, it's, it's really, radically changed our ability to make data useful. and, you know, rewind five years, we, ago we started building AI and, and products for, for the first time really started with a safety product.

we had these dash cameras and, they would capture accidents and near misses. These were detected by someone, slamming on the brakes. So a high GForce event. He would upload video and you'd look at these and, and so many of them were drivers on their phone, distracted, right? So we started building AI to basically detect distraction and, and mobile usage and, you know, give their driver a.

A reminder to put their phone down, make it such that you could, you know, run a, a program where you rewarded drivers for not using their phones while driving. And that had a massive difference in, in accident rates. We kind of stacked on different, features on top of that, like drowsiness or, tailgating or rolling stop signs, all these different kind of safety behaviors. Each of those were pretty slow and expensive to, to, to build because we had to basically build new AI models from each of them. Were kind of training them on top of all this dash camera data. now we can use LLMs as a starting point and we can start to do things that just would've been impossible before, right?

So for example, now if you slam on the brakes and it's because, someone cut you off, or?

a pedestrian ran into the road. An ambulance, crossed in front of you even though you had a green light. The AI can actually watch that video and say, Hey, rather than lower your safety score for, for slamming on the brakes, I'm gonna increase your safety score for, for driving defensively.

That's something that if you have a big enough team and you have someone reviewing every video clip, you could do manually, but most customers don't. Right? So whether it's, you know, that you can apply ideas like that to maintenance, to fuel consumption, to training, to doing safety inspections at a job site, like all of these different areas and all of the data's there, but now we can actually just make it meaningful in ways that that was really not possible before.

Brett: who has influenced you the most as a company builder that's not a family member, and what's the thing that they imparted on you that's like a big part of your worldview of, of company building more broadly?

Kiren: Well, the, the practical answer is that, you know, Sanjit is, is someone who I've now known for, gosh, over 25 years, he's been my boss for the last 15 plus. and, you know, just by, by virtue of, of working together for so long, he's had a incredible impact. you know, we started doing computer science classes together as, as undergrads.

and then, you know, going from, from that to, to building Meraki and, and, and, and now samsara and. I think, you know, a lot of it has been the process of, of building together, you know, a lot of his, superpowers. I won't say they've rub rubbed off of me, but there's something that I can, I can try to try to emulate.

but, you know, definitely a, a huge impact there.

Brett: What are, what are the superpowers?

Kiren: Oh, there, there's a, there, there's a long list. but I think, you know, probably the, the, the biggest are his ability to be very strategic and be thinking about markets and technology trends and multi-year cycles and, and then kind of get into the details of, you know, this model could run on this chip set and could enable this feature which would be useful to this customer.

And that ability to kind of span, span those different altitudes. And then this is kind of nuclear powered, creativity and, and sense for what's possible and kind of never, never being complacent and always looking to say how, how can we do more? so that's the one that I've, you know, got to, to just, be building with and, and, learn by doing.

Brett: Great. Thanks for spending all the time with us. This was awesome.

Kiren: Thank you. It was a lot of fun.