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 2,194 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 7 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
MacInnis: Product marketing rarely demands deep first-principles thinking
“If you look into the ocean of product marketers, it's kind of like shopping at Ross. It is not a job that demands a lot of really deep first principles thought from people in the job.”
MacInnis: All alpha and outperformance comes from anti-patterns
“All alpha, all outperformance relative to the index comes from anti-patterns. Patterns are the things people do over and over again... None of it is predictive of success. It's the anti-pattern.”
Nash: Wealthfront bet delegators would trust software over human financial advisors
“The contrarian play behind Wealthfront was that people assumed that delegators wanted to work with people, and that do-it-yourselfers would do stuff online, do it themselves. And Wealthfront, no, no, actually, there's a new generation of people. They are very …”
Entering existing software categories requires product completeness rather than product-market fit
“In this case, I don't think you're looking for product market fit because in the end of the day, product market fit is for a category, right? Do you need forecasting software? Well, I do. I already have one. So it's more about call it like maybe product comple…”
Virtual AI representatives will not replace human salespeople within ten years
“So I don't think they're going to go away. I mean, people are like, oh, can a virtual rep call you and sell you something? It's like, I'm not going to pick up the phone for what the next 10 years, but certainly make their work much more efficient.”
Habib: Writer's largest models match GPT-4 quality at lower hosting cost
“Now we've got 18 models. The biggest ones are GPT-IV level quality, but at a fraction of the cost of hosting. Super, super fast for the enterprise use cases that That our customers run them on.”
Habib: AI capabilities will outpace most organizations' capacity to absorb them
“And I do think the capabilities that are possible are gonna outstrip most organizations' capacities to absorb them, for most of their employees to absorb them without a lot of effort.”
Habib: Writer qualifies out enterprise prospects with weak champions, even CIOs
“And so our team is really good at politely declining to engage, right? Not just because the use case doesn't fit or it's not ICP, but we qualify the champion out, even if it's a CIO.”
Habib: Writer has the potential to become a 400 percent NRR company
“I think this should be a 400% NRR company.”
Habib: Current transformer architectures fail at enterprise agentic workflows
“You know, our kind of tests on inference on quality for agent use at work. It's just transformers. Just, it was current capabilities, really not able to do what we wanted them to do. And so new architectures that we're working on are super, super promising.”
Habib: Vector databases fail at RAG retrieval beyond 50,000 pages
“Like, it's fine when I am, I've indexed 50 pages, but when it goes to 50,000, you know, I can't get this needle in a haystack out of my kind of VectorDB approach.”
Masad: Mature tech platforms require ambitious breadth rather than narrow wedges
“In the early innings of a platform shift, you can be narrow and you can find a lot of product market fit, but towards when the thing mature, say, you know, web or cloud, what have you actually have to be a lot more ambitious.”
Masad: AI add-ons will become native within a couple years
“GUI was an add-on at some point, and I think this is where we are with AI. But I don't know, I think it's a transitionary period. I think at most couple years where most companies are going to take the approach we're taking.”
Masad: AI will bifurcate developers into 'creators' and 'engineers'
“I think that there's going to be a bifurcation of software engineer and software creator. Engineers are not going away. We need engineers in the same way that Instagram did not obviate the need for professional, you know, journalists and photographers and all …”
Building successful software is getting harder despite AI and modern tools
“No, I think it's getting harder, actually. Well, number one is because there's so many, when, back in the Barlow days, this is pre-web, this is software that you print on a disk and ship to humans, so you've got, you know, you're shipping it once a year, so th…”
Managers should encourage employees to leave Slack once they outgrow it
“And here's a controversial one. There's times we spent a year. I'm like, you don't need to be here anymore. You have learned everything that you're going to learn here at Slack. It's time to go somewhere else. And I wouldn't say it like that, but there's truly…”
Amin: Clay drove growth by forcing all customer support into public Slack
“Eventually the way that we got a ton of customers was actually by cutting off all of the different ways that we supported customers.
So we basically pushed all of our customers into the same channel.
So anyone that would sign up to Clay, we would say, join our…”
Running PLG and enterprise requires operating two completely separate companies
“Winning the enterprise is really, and winning a more traditional PLG motion, it's almost like two separate companies. I think it's a very different way you think about building product, different way you interact with customers, and certainly a very different …”
PLG companies mistakenly believe they can seamlessly blend into the enterprise
“I think the big mistakes companies make over and over again when I see is when they try to blend, they think they can just seamlessly evolve. And as they scale and Into an enterprise company. I think it's a lot more complicated than that.”
Jay: The 'PLG trap' happens when companies run out of organic accounts
“As you build the security stuff, it does unlock major expansion in, within your customer base, right? The companies where you have good organic usage, many of those companies will start paying you 25, 50, a 102 hundred K. And so you're taking in this data and …”
Jay: Pitching team productivity fails to resonate with enterprise executives
“As an executive, all I care about is driving enterprise value. And then really the only things that drive enterprise value are driving more revenue, lowering costs, risk mitigating, right, on that level. And so your pitch at that point is like, well, look look…”
Dropbox and Slack's enterprise value came from securitizing rogue employee usage
“And so Dropbox and Slack, if you ask me what's the company value, well, the company value was securitizing, in a way, the bottoms-up problem that we created for them, which is all this rogue usage.”
Software companies cannot become category leaders without winning the enterprise market
“All categories have to go through the enterprise. All winners have to win through the enterprise. There's very few, I really can't think of many of any companies that have become the category leader without winning the enterprise.”
Jay: PLG companies must start executive value sales concurrently with expansion sales
“Phase three starts the same time phase two starts, because it takes that much time to identify the problem, identify the path, and then start building the right team to execute that, that play. It takes a long, long time.”
PLG companies mistakenly segment sales teams by company size instead of motion
“This is something that I see a lot of PLG companies get wrong, which is they confuse segments and motions. If you take a step back, traditionally enterprise companies are set up and they are split by segments, SMB, mid-market enterprise, right? And of course y…”
Notion would not exist if Evernote had executed on team-level value
“One can argue if they executed better, there would be no notion as well, right? But they just got stuck on the individual user value, and the team user value, like, they didn't really, it doesn't really, it didn't really resonate, right? The product didn't rea…”
Very few PLG companies outside of DevOps have truly broken out
“There really aren't many successful PLG companies outside of like the DevOps ones like Datadog and Atlassian that have actually really broken out.”
Blank: Chaos-driven founders self-sabotage company scaling after product-market fit
“When the company finds product market fit and you need to build repeatability and scalability in the company, these same founders throw organizational hand grenades into their own organization and culture, trying to keep that chaos going on inside their compan…”
Steve Blank: Steve Jobs's second Apple tenure succeeded via operational executive buffers
“Jobs was still insane, but the second coming of Steve Jobs had an organizational level of world-class executors. Tim Cook, John Rubinstein, Avi, et cetera, were all people who could take that craziness and filter it down to an organization that provided a leve…”
Steve Blank: 20th-century practice of replacing founders killed long-term startup innovation
“The mistake that people used to make was only replace the founder in the 20th century to take a company public. We forgot that that innovation cycle might be every, you know, two to three years and you could get liquidity for the investors, but then the compan…”
Blank: Customer discovery was designed to test vision, not run focus groups
“I invented customer discovery for me, which was to inform my vision, not to create a goddamn focus group.”
Blank: World-class number twos almost never make world-class CEOs
“They were world-class number twos and almost never are world-class number ones.”
Blank: Ballmer collapsed Microsoft OS share by missing mobile and search
“He basically took Microsoft operating system share from 98% to probably two percent today, right? If you now count mobile devices, right? They're just irrelevant. Because he was so focused on extending the windows ecosystem. They missed mobile. They missed sea…”
Blank: The entire Lean Startup methodology will be enabled by AI
“The entire lean method. Is going to be end to end enabled by AI.”
Blank: Generative AI innovation will die if big tech stops competing
“So I sure hope OpenAI, Google and Facebook, and even Apple continue to slug it out at some point. Because if they don't, innovation will die in the browser or like the browser did, it will die in generative AI.”
Narkhede: Confluent ignored narrow ICPs to capture the broad Kafka base
“Kafka, you know, that was one of the reasons why we couldn't or we didn't sort of narrowly focus on an ICP like traditional companies do for good reason. It was because, you know, the Kafka adoption base was so broad and for Confluent to quickly convert those …”
Narkhede: Commercial open-source requires all project creators to found together
“For the success of an open source company, all the creators should be part of it. And when that doesn't happen, you have two companies and they're competing on things they shouldn't be competing on, which mostly comes down to pricing and it's just a race to th…”
Narkhede: Confluent chose not to separate on-prem and cloud sales teams
“We actually decided to not split it. And that was because we ultimately wanted to transition to a cloud first business. So we put in a, instead a lot of thought and a lot of investment in sales enablement you know, internal education of why this is important, …”
Narkhede: Confluent probably would have failed if it started as SaaS-only
“Would it have just been feasible if we started with a SaaS only offering? And the answer is probably not because, you know, one of the most important things an open source company needs to do is quickly establish yourself as the market leader, quickly establis…”
Google frames Bard as an experiment because LLMs aren't fully baked products
“It's part of why we talked about Bard as an experiment at this point. We don't think that this technology is yet ready to be a fully baked product.”
Information retrieval is not the majority use case for Google Bard
“Some people, and it turns out not the majority use case, but some people are going to use it to try to find information.”
Persistent memory for LLMs will take much longer to solve than expected
“And so I am beyond Fascinated and intrigued by the research that's taking place around memory and being able to reference things from previous conversations, but I think it's going to take way longer than we expect.”
LLM hallucination is a core generative feature, not a bug
“And so finding the way of harnessing that as a feature rather than a bug, I think is going to be challenging, especially when one of the perceived shortcomings of your product is actually going to be, is like the core of what makes it function. I don't know th…”
Tech companies dodged engineering resource allocation for a decade through overhiring
“One challenge of the last decade, right, is that resource allocation was largely dodged by many companies because they just Decided to spend more and hire more folks. And so oftentimes the real constraint was, like, hiring velocity on a given team and your abi…”
Believing engineering managers shouldn't get into technical details is an anti-pattern
“I think one of the lessons in management over the last decade has been that You know, managers are people leaders in that it is micromanagement that it is bad to get into the details. I think this is like an anti pattern because it creates disengaged and like …”
Larson: Engineering executives must educate peers instead of shielding teams
“The most valuable thing as an executive you can do is educate like the other execs on how engineering works, not kind of protecting engineering from visibility from anyone else.”
Starting from AI capabilities beats traditional customer-interview-driven product development
“What we realized as we kept building those products was starting from like research that just like Became possible and identifying some coming, having some hypothesis of the ways in which it could really, we could build like really strong and valuable tools ar…”
Future websites will be dynamically generated just-in-time for individual users
“The thing that we want to accomplish with the web is that we want to make it truly an AI driven web where when you go to a website, you get a copy of that page that is just for you generated just in time. And it knows it has all this context about what you wan…”