Discover how Meter's CEO built a multi-decade networking infrastructure business by owning the entire stack, innovating on business models, and thinking long...
Building for Decades: The Full Stack Networking Company Strategy
Key Insights
- Long-term thinking matters: Building a business for 20, 30, or 40 years requires owning the entire technology stack, from hardware to software to operations
- Business model innovation is underrated: Most founders focus on product and technology, but reimagining how you deliver and charge can be transformative
- Vertical integration enables competitive advantage: Every major company that has succeeded long-term is vertically integrated, yet new startups often avoid this approach
- Time is more precious than money: When starting a company, you can waste money, but wasting time on wrong paths has irreversible consequences
- Customer obsession prevents market blindness: Being intensely close to customer problems helps you catch issues before they scale into company-wide crises
Understanding the Long-Term Vision
When you're building infrastructure for the next 20, 30, or even 40 years, the decisions you make today have exponential consequences. This is the core philosophy driving Meter, a networking infrastructure company founded by Anil Baranasi and his brother Sunil. Unlike many startups that optimize for quick exits or 5-10 year outcomes, Meter was built with a fundamentally different time horizon.
The challenge with thinking in decades rather than quarters is maintaining focus on both the long-term vision and the daily work required to get there. Anil describes this using what he calls a "barbell approach" to time horizons. On one end, you maintain conviction about where the company needs to be in 25 years. On the other end, you obsess over what happens in the next few hours and days. The middle ground—the 5-year plan—gets far less attention than traditional management frameworks suggest it should.
This creates an interesting rhythm in how work gets prioritized. The team at Meter doesn't use OKRs (Objectives and Key Results), the management framework popularized by Google and adopted across Silicon Valley. Instead, they focus on understanding the immediate next step required to advance toward the long-term vision. It's iterative discovery rather than prescriptive planning.
The Full Stack Imperative
One of the most counterintuitive decisions Meter made was committing to vertical integration from day one. In an industry where specialization is the norm—some companies build just switches, others build just wireless access points—Meter insisted on building hardware, software, and operations together.
This decision wasn't made in isolation. It came from studying why the networking industry has been dominated by the same few companies for 40 years. Anil observed a pattern: every startup that tried to compete would begin with a single point solution—maybe firewalls or switches. They'd gain early traction, but then they'd plateau. Rather than building the adjacent layers of the stack, they'd get acquired by one of the incumbents or remain a niche player forever.
The reasons for this are partly structural and partly psychological. A startup with limited capital and runway needs revenue quickly. The fastest way to revenue is to solve one specific problem exceptionally well. But this creates what Anil calls a "local maxima" trap. Once you're making money from your point solution, the incentive structure shifts. Every smart person on your team naturally gravitates toward deepening expertise in that narrow domain. Pivoting to build a broader stack feels like abandonment of your core business.
The solution, counterintuitive as it may seem, is to start broader. Not broader than the market will accept initially—Meter still started in small offices before expanding to warehouses and data centers. But broader in ambition. The insight is that building the full stack might only require 20% more effort than building a point solution, yet it enables you to capture dramatically more value.
Business Model as Product
Perhaps the most underappreciated insight from Meter's journey is treating business model as part of the product itself. In traditional thinking, you build a product and then figure out how to sell it. At Meter, the business model informed every hardware, software, and operational decision from the beginning.
The industry standard for networking was simple: build hardware as cheaply as possible, sell it as expensively as possible. Margin comes from the gap between manufacturing cost and selling price. This incentivizes cutting corners on hardware quality and pushing customers toward expensive support contracts.
Meter inverted this logic by studying how other industries had evolved. The security and alarm industry had pioneered "recurring monthly revenue" (RMR) models decades before SaaS became fashionable. The insight was elegant: instead of selling hardware, sell the outcome. The customer pays monthly based on what they get—in Meter's case, a per-square-foot pricing model that includes the hardware, software, installation, and ongoing maintenance.
This business model change had immediate implications for product decisions. Because Meter keeps responsibility for the network's performance, they could justify spending 20-30% more on components than traditional vendors. Those components weren't a line item cost to be minimized; they were an investment in delivering the outcome customers paid for. The higher-quality components meant better reliability, which meant fewer customer issues and lower support costs.
The model also solved the chicken-and-egg problem of selling something new. Customers weren't evaluating a new piece of hardware; they were buying a service. This aligned incentives perfectly: Meter made money when the network worked. If the network failed, Meter didn't get paid.
The Hardware Manufacturing Journey
The path to building the first saleable product took approximately four and a half years—a timeline that would have been catastrophic if Meter had raised venture capital with typical expectations. The brothers had self-funded the company, which meant they could afford to take time to get things right.
The software work started first, since it seemed like the logical place to begin. But they quickly hit a wall: they couldn't achieve the performance they needed without building custom hardware. This led to a critical decision: where in the world could they iterate on hardware design fastest?
The answer was Shenzhen, China. For someone in San Francisco, the idea of designing a circuit board in the morning and having a prototype in your hands by nighttime seemed impossible. But that's exactly what the manufacturing ecosystem in Shenzhen enabled. Despite initial reluctance from established vendors—they all had minimum order quantities that Meter couldn't meet—some manufacturers took a chance on the young team.
What was supposed to be a one or two-month trip turned into a year and a half. Anil and Sunil lived in Shenzhen, learning everything about hardware manufacturing, supply chains, component sourcing, and production processes. Yes, the vegetarian diet was brutal. Yes, they learned that they could design a circuit board in the morning and have a working prototype by evening. But more importantly, they absorbed knowledge about how physical products actually get made—knowledge that remains embedded in Meter's operations today.
One of the most painful lessons came about a year into the software development work. The brothers realized they'd taken the wrong architectural approach to building the operating system and packet processing layer. They'd tried to optimize for an idealized version where every technology worked perfectly together. In reality, they needed something pragmatic that worked with the messy constraints of actual hardware and existing systems.
The realization came at an operating systems meetup in Santa Clara where an older engineer mentioned that a major tech company had been trying the same approach for seven years and had decided to open-source the resulting library. Once Anil and Sunil saw how this other approach worked, they understood instantly that they needed to scrap a year's worth of work. It took them another week to admit this to each other, but they did. That's a year out of the four-and-a-half-year timeline lost to taking the wrong path.
Customer Development Without Design Partners
As Meter built the hardware and software, Anil spent afternoons doing door-to-door sales near their small San Francisco office. The goal wasn't to generate revenue; it was to understand what networking companies actually needed. Most people weren't particularly excited about what Meter was building. This wasn't surprising—in an industry where new competitors rarely emerge, buyers aren't accustomed to evaluating novel solutions.
The insight Anil gained from this experience was that networking is similar to the automotive industry. People don't search for new car companies; they buy cars because they've seen them on the road or a friend recommended one. Similarly, networking hardware decisions are made through reputation and relationships, not by actively seeking innovations.
This realization changed their go-to-market approach. Rather than trying to convince the market of a better way, they focused on building something so obviously superior that customers would want it. For the software components—dashboards, management interfaces, and user experience—they did iterate with early customers. But for the core packet processing, wireless, and security functionality, they didn't need customer input. They needed to answer three questions: Can we actually do it? Is it fast? Is it secure? Customer feedback on these technical foundations wouldn't help because customers don't understand those layers the way engineers do.
Managing for Impact Over Process
As Meter grew from two founders to a team of dozens, Anil became increasingly aware of a dangerous tendency: the proliferation of what he calls "meta work." This is the work about work—status meetings, planning sessions, documentation, process design, tracking systems. In many companies, meta work crowds out actual work.
Anil is explicit about wanting to avoid this trap. The time people spend at Meter is their time—and Meter is buying both their brains and their hours. The goal should be to maximize the percentage of time spent on actual work and minimize time spent on coordination and process.
This philosophy extends to his approach to management. Rather than seeing managers as people who get promoted from individual contributor roles, Meter tries to keep everyone as "player coaches"—people who still do the work while also helping coordinate across the organization. This is possible at their current size, though Anil acknowledges they'll eventually hit a breaking point where this model becomes unsustainable.
He's also deliberately resisting the typical Silicon Valley management approach of separating management from influence. At Meter, you don't have to become a manager to shape the direction of the company. The goal is to allow great engineers, hardware designers, and operations specialists to have influence over decisions without the administrative burden of people management.
On one-on-one meetings—the standard practice in most tech companies—Anil is skeptical but not dogmatic. He doesn't do scheduled recurring one-on-ones; instead, he picks up the phone and calls the 24 people who report to him when he's thinking about something. He talks to each of them at least every 2-3 days, but in real-time conversations rather than structured meetings.
This raises the question: how do you give feedback, coach people, and help them grow? Anil's answer is that most of that should happen immediately and peer-to-peer. If a colleague sees something that needs correcting, they should address it directly. The manager's role is to be a mirror—pointing out patterns they see across the organization that individuals might not see themselves, and ensuring that high-quality work is happening at the expected pace.
Speed and Slowness
Early in Meter's life, Anil was obsessed with speed—moving as fast as possible toward product-market fit. Over time, his thinking has evolved. He's now convinced that you need different speeds for different types of work.
For some activities, speed is paramount. If a prospect inquires about Meter's products, the response should be lightning-fast regardless of company size. Customers who are paying the bills deserve immediate attention to their issues. Sales interactions need velocity.
But for other decisions, slower is better. When you're choosing what hardware to build and how to architect it, decisions have ramifications that last for years. Rushing through these choices creates technical debt and architectural mistakes that haunt you. Similarly, when Meter eventually designs its own chips—which Anil expects will take another eight years—the right approach is to move deliberately and carefully, minimizing the chances of expensive mistakes.
The way Meter instills urgency without demanding constant speed is through deadlines for meaningful events. They run a user conference in November where they release new product lines. This deadline forces prioritization—people naturally cut work that doesn't matter when they don't have time for it. It's the same principle Apple uses with its annual September keynote. The deadline itself becomes the mechanism that ensures focus.
Anil has also noticed that when he asks teams "Why does every problem in the world take exactly 90 days to solve?" people often realize their estimates are arbitrary and come back with more realistic assessments. It's a question that forces reckoning with whether the timeline actually matches the work required.
Navigating Incumbent Markets
The networking industry is one of the few sectors where multiple $200+ billion companies exist, yet virtually no new companies have entered the market successfully in decades. This seems to violate basic market efficiency principles—shouldn't new competitors emerge to challenge entrenched players?
Anil's hypothesis is that the barrier isn't the quality of the incumbent products; it's the nature of the competitive moat. When a startup tries to compete in networking, it typically starts with a narrow point solution. The incumbent vendors can afford to lose money on that segment temporarily, subsidizing their own solution to kill the competitor. For a well-funded startup, that's a speed bump. For a capital-constrained team, it's fatal.
But here's the counterintuitive insight: once you pass through that initial gauntlet and achieve real product-market fit, suddenly there's almost nobody else there competing with you. The incumbents have moved on to their next threat, assuming you'll either fail or get acquired. The difficulty is inverted—early is incredibly hard, late is easier.
This explains why Meter delayed channel sales for years. The networking industry sells 90% through the channel—resellers and distributors who have relationships with enterprises. But channels are also how incumbents kill new entrants. By selling direct, Meter could build proof points and case studies. By the time they entered the channel, they had hundreds of customers and evidence that their approach was dramatically better. At that point, the channel partners wanted to work with them.
Learning to See the Whole System
Perhaps the most important lesson from Meter's journey is how to recognize problems before they cascade. Early on, every customer hated Meter's products for about six to eight months. The team's initial reaction was to rewrite everything—they assumed they were the problem.
It turned out that the issue wasn't Meter; it was a specific implementation detail in how Apple had implemented software-defined radio in their M1 chips. Once that issue was fixed, customers immediately went from hating the product to loving it. But during those eight months, Meter could have made different decisions—different hardware choices, different software architecture—that would have permanently constrained their ability to iterate.
The lesson was to become unreasonably close to customer issues. Anil and Sunil review every customer problem, regardless of size. A small customer paying $20,000 per year gets the same level of attention as a major enterprise. This doesn't mean they personally fix every issue, but they make sure they know about it, understand why it happened, and learn what it reveals about systemic problems in their product or operations.
The Role of Great Markets
Looking across all of Anil's decisions—the choice to build full stack, the decision to stay focused on networking, the timing of when to enter different market segments—he credits one crucial factor: choosing a great market.
Networking touches literally every person on Earth. Packets flow through networks constantly. It's a market that existed for 40 years when Meter was founded, so it's mature. It's growing, particularly as e-commerce has expanded and companies have automated warehouses and operations. It has enduring purchasing power because networks are critical to business operations regardless of economic conditions.
Most importantly, it's a market that will continue to matter for decades. There's no risk of the problem becoming irrelevant. This is why Anil believes market selection is often underappreciated in startup advice. Founders focus on product and technology innovation, but the market you choose determines the ceiling on your ambitions.
Many of the best markets—the ones with the most growth potential and that touch the most people—are dominated by entrenched incumbents. This seems discouraging. But Anil's observation is that the difficulty is front-loaded. Breaking into an incumbent-dominated market is brutally hard. But once you've broken through and proven you have something materially better, the incumbents often can't respond effectively. They're optimized for the existing market, not for defending against a better approach.
Sustaining the Infinite Game
One of the most striking aspects of Meter is that it doesn't feel like a typical startup grinding toward an exit. Anil and Sunil work with intensity, but they're playing an infinite game. They're genuinely excited about running the company for the long term, not just for the time until acquisition.
Part of this comes from the nature of the business itself. Hardware, software, and operations create constant opportunities to learn new domains. Another part comes from having a co-founder—Anil notes that in a relationship, "neither of us wanted out at the exact same time." When one person is down, the other finds a way to stay engaged.
But the deepest reason is the work itself. Anil genuinely loves understanding how systems work. He has a standing practice with his brother of using the entire product end-to-end every weekend—every feature, every workflow. He reviews every customer issue. He talks to his team constantly. This creates a level of engagement that many founders lose as their companies scale.
The lesson for others building companies is that if you're doing work that genuinely interests you, with people you trust, in a market that matters, the path to a satisfying long-term outcome becomes possible. The goal isn't to optimize for happiness or fulfillment explicitly; it's to choose the right work, the right people, and the right market, and those things follow.
Market Entry and the Macroeconomic Index
Meter initially planned to focus exclusively on office networks—the obvious starting point for a networking company. But COVID-19 changed everything. Office occupancy dropped to zero, and the supply chain for hardware became severely constrained. Larger companies got priority; Meter couldn't get components.
But Anil noticed something: e-commerce was growing from 12% of economic activity to 24%. This drove a massive expansion in warehouses, manufacturing facilities, and automated logistics centers. These spaces needed networks. And because they were being built and automated, they needed better networks than traditional offices.
Macroeconomics pushed Meter into a new vertical almost by accident. What appeared to be a crisis—the office market dying—forced expansion into a segment that ended up driving significant growth. This is why Anil thinks of Meter as an index on the economy. As economic activity shifts, the company's growth opportunities shift with it.
Building Conviction Through Diverse Influence
When Anil chose George Mason University as an undergraduate, he was exposed to economists like Tyler Cowen who believed the world wasn't a black box—that you could actually understand how systems worked if you invested time in learning. This had a profound impact on his worldview.
The insight was simple but powerful: the world's complexity isn't a barrier to understanding; it's an invitation. You don't have to accept things as they are. You can learn how they work and think about how they could be better. This mindset is evident throughout Meter—from the choice to build full stack rather than accepting the industry's point solution model, to the willingness to scrap an entire year's worth of work when a better approach appeared, to the conviction that you could sell networking on a recurring revenue model despite decades of one-time licensing.
Conclusion
Meter's story isn't about discovering a killer product feature or finding a massive untapped market. It's about making a series of deliberate decisions that most companies avoid: committing to vertical integration despite its complexity, treating business model as part of product design, choosing to stay in a single market and dominate it rather than diversify, being unreasonably close to customers and their problems, and building a company culture that values doing real work over talking about work.
The path to building a decade-spanning business requires a fundamentally different mindset than optimizing for a five-year exit. It requires believing that you can understand how complex systems work, that ownership of the entire stack creates value that point solutions can't capture, and that business model innovation is as important as technology innovation. Most importantly, it requires having the conviction to make choices that seem harder in the short term because they're more defensible in the long term. For founders building for decades rather than quarters, these principles offer a blueprint for thinking about what actually creates enduring value.
Original source: Building Meter for decades, not an exit | Anil Varanasi (Co-founder and CEO)
powered by osmu.app