The Ledger
Every statement that passed quotation and attribution checks. Mix any filter with any other: certainty 1/5, debate potential 5/5, or both at once.
why aren't all 41 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 1 checkable ones are still open, waiting for their date. predictions held up or didn't; assertions are supported or contradicted. on every card: ▮▮▮▮▮ certainty · ▮▮▮▮▮ debate potential. speakers are clickable
Varanasi: No single large company today is not vertically integrated
“What's ironic to really think about is there is not a single large company in the world today that is not vertically integrated. There's just none.”
Varanasi: Small teams should ask peers, not just managers, who to fire
“A lot of times when you go to a team that's six or seven or 10 people, more than the manager, I think it's actually better to ask everybody else that works with in that team who should be fired.”
Varanasi: If you think about the person, the person is the problem
“If you're spending your time thinking about the person instead of the problem, the person is the problem.”
Varanasi: OKRs only work for money-printing monopolies like Google
“If you have a money printing machine, you should have that kind of stuff, but for everybody else, it's different.”
Varanasi: Networking industry will move to square-footage pricing within 5-10 years
“I think now the rest of the industry will move there in the next five, 10 years is some scribbles and notebooks of me and Sunil where we came up with this.”
Varanasi: Networking startups fail to build full stacks and become acquired point solutions
“What we understood what happened with all the companies in the last 2030 years that tried to compete with Cisco was that they then ended up selling that. Then the best companies actually start having success. Then they never go out and build the rest of the th…”
Varanasi: Building a Very Large Business Requires Vertical Integration
“But if you do start a great, you know, point solution or something where you own horizontally one part of the stack, I think the chances of making it to a certain level are higher, but the chances of building a very large business, I think are very slim chance…”
Varanasi: Incumbent-heavy markets become much easier once the initial barrier is crossed
“In a market where incumbents like that don't exist, the initial ramp is easy to get in, but so many people get in, then it becomes really difficult. But with markets like this, with large incumbents, the initial ramp is really hard, but once you pass that chas…”
Varanasi: Meter focuses on a 25-year horizon rather than five-year goals
“I don't care about
Where meter ends up in five.
I care about where it ends up in 25.
But at the same time, the next few hours also really matter to me.”
Varanasi: Meter plans to develop proprietary ASICs and custom chips
“Over time, we can go even further down the stack where we should go make our own ASICs, our own chips.”
Varanasi: Most tech workers do not use the products they build
“I think another thing that I found very surprising, but then I'm pretty sure of is most people in technology don't use the products they help build.”
Varanasi: Leaders should be CC'd on every single customer issue
“I think if you help people understand that anytime a customer is having an actual issue, you should be at least CC'd on it for every single customer. You can actually drive a lot down.”
Varanasi: Meraki was the last great networking company
“The last great networking company was a company called Meraki.”
Varanasi: Managing hardware allows Meter to use 20% to 40% pricier components
“So what that enables us to do is because the hardware is for us to manage and us to make sure it's great, we maybe might choose components that are 20%, 30%, 40% more expensive than a traditional vendor might, but that is only marginal for us compared to the e…”
Varanasi: Core infrastructure software does not need customer design partnerships
“The job there is just get it working first, then make it fast, then make it secure. You don't need design partnership there. Can you actually process packets? Is it actually working? Is it fast and is it secure? Because no customer actually understands that pa…”
Varanasi: Meter's in-house chip plans will take eight years to yield
“Yeah, we have plans underway. We think we can do some great work there, but by the time we realize that it will take another eight years, I think from now before this will even be fruitful at all.”
Varanasi: Business Model Innovation Is Severely Underrated Compared to Product and Delivery
“If you look at businesses in the modern sense in the last 50 years, The three times businesses are usually rewarded is when they innovate on product and technology, or they innovate on how they deliver that. You know, somebody figured out SaaS, before that som…”
Varanasi: Expanding initial startup scope by 20% unlocks vastly more value
“If they had decided at the beginning that the scope was a little bit larger, generally my rough intuition is it's only 20% more. But it actually ends up enabling you to capture a lot more.”
Varanasi: Silicon Valley culture pushes founders to optimize for local maxima
“Particularly in Silicon Valley, that this sort of thinking has accelerated, which is make a point solution as fast as possible to get to some revenue as fast as possible. And everybody that ends up optimizing for the local maxima, because if you end up giving …”
Varanasi: Founders mistakenly believe their specific problem is the world's hardest
“Almost invariably, every company tells you what they're working on is the hardest problem in the world. And I just don't think that's categorically true at all. But because, you know, you end up becoming either a slave of your own success or end up having some…”
Varanasi: Creating processes and tracking tools is not real work
“You put a process in place, you write a doc down, you talk about how to do the work. You talk about how to set up your tracking system on like, you know, Atlassian or linear or whatever. Doing all that feels like real work. I just don't think it's real work at…”
Meter aims to decouple management from decision-making authority
“And one of the things we're trying to do, and I don't know how successful we will be over the next five to seven years is, the other thing we disagree with in this vein, somehow people have bundled management and authority together in one. We're trying to deco…”
Anil Varanasi talks to his 24 direct reports every few days
“All the two dozen I talk to at least every two, three days.”
Varanasi: Chip development should move as slowly as possible
“When we go build chips, for example, I actually think the right thing there is to go as slow as possible because taping out a new chip, then building it is incredibly hard on ramifications on the mistakes you make.”
Varanasi: 90-day project timelines are artificial constructs tied to quarterly cycles
“I think with smart people, if you ask this question, you say, how is it that every problem that could be a sales problem, that could be an engineering problem or an operations problem, that could be a design problem. They can be exactly solved in 90 days, but …”
Varanasi: Cisco built its hardware stack via acquisitions, unlike Meter
“So if you look at our industry, Cisco has a rack like ours where they have, you know, 10 different pieces of hardware. But majority of that was done through acquisition. So they bought their switching business called Crescendo in the nineties, et cetera, Merak…”
Varanasi: SaaS recurring revenue models originated from the security and alarm industry
“Turns out it actually came from security and alarm industry. They were doing one-time sales and one-time revenue. Then there's this analyst consultant type person named Ron Davis in the security and alarm world. He started writing about this concept called RMR…”
Varanasi: Enterprise networking has multiple $200B+ giants yet no new entrants
“So what happened in the networking industry for the last 40 years is it's one of the only industries that has multiple 200 plus billion dollar businesses, yet no new companies come in.”
Varanasi: Enterprise networking buyers lack the muscle to evaluate new vendors
“Because new companies don't come into this industry, the buyers actually did not have a muscle to evaluate new customers, new vendors all the time.”
Varanasi: Hardware V1s are constrained by supply chains, not ideal designs
“You set out on a blank sheet of paper and you say you want to build the best hardware, but cost, MOQ, supply chain, which vendors will work with you, which chips you have access to, those all start narrowing your choices down to, you know, maybe just ok hardwa…”
Varanasi: Meter lost 100% of customer office activity starting March 12, 2020
“A hundred percent of our customers went to zero starting March 12th.”
Varanasi: Any business goes through roughly 1,000 days of severe hardship
“Studying businesses now the last two decades, I think for any business. I think there's about a thousand days of really bad time. Any business. We just happen to have all thousand at the beginning consecutively.”
Varanasi: Founders must lead sales when entering brand new markets
“If it's a brand new market, the founder has to do it, or it has to be part of it to convince the first few people to take the chance. Cause I think they're buying the person rather than the product.”
Varanasi: Academic RCTs prove management directly drives superior factory outcomes
“We have really great RCTs and A-B tests essentially of same problem, same factories, half management, half without management, and outcomes being entirely different.”
Varanasi: Fixed deadlines work by forcing teams to cut nonessential work
“And the most important thing is not that I think people like end up working hard or anything. I think people, if you have a great company, they're always working hard. What ends up actually happening is people end up not working on things that don't matter. Th…”
Varanasi: Full-stack hardware-software founders stay happier than pure software founders
“I have talked to other companies that have this like hardware software operations component. And I do think you're right. There's something there that these companies, founders tend to be a bit happier than others. Cause what you end up not liking about your o…”
Varanasi: Apple Silicon switched to two shared software-defined radios
“When they did that, usually hardware had three separate radios for like, 2.4 gigahertz, five gigahertz and Bluetooth. When they moved, they were using what this thing called software defined radio, which is using two radios for all three of those things.”
Varanasi: Alarm.com and ADT are valued more on RMR than GAAP revenue
“And today, if you go look at alarm.com or ADT or anyone, More than the gap revenue, this is the number that people care about. It's not even a gap metric, but what is the RMR?”
Varanasi: Shenzhen delivers manufactured circuit boards same-day
“You could design a circuit board in the morning and some kid would run over to your office by nighttime.”