Search the archive

Start for free →

LLMS ARE PICKING WINNERS. HERE'S HOW TO BECOME ONE

Posthog's traffic from LLMs grew 41x over the past two years. This converted better than almost every other channel Posthog runs. This post looks at how Posthog used Answer Engine Optimization, a mystical new field, to ensure models know about the business and its products.

SILICON VALLEY'S NEW BUYOUT PLAYBOOK IS HITTING WALL STREET

Venture capital firms are buying legacy companies to integrate AI, bypassing traditional enterprise software sales. This strategy, known as the AI rollup, targets industries with low software adoption, such as healthcare and accounting. VCs like General Catalyst and Thrive Capital seek growth by permanently transforming these businesses, contrasting with traditional private equity's focus on financial engineering.

BUILD FOR THE MARKET OF ZERO

NVIDIA once spent billions building a machine for a market of exactly zero customers. Its CEO told a Stanford class that this is the whole method, you reason out what the world will need and build it before anyone asks, rather than surveying customers about what they want today. His advice to founders is to reason forward to what has to be true, because customer research can only ever describe the present.

SIZE YOUR MARKET IN GROSS PROFIT, NOT REVENUE

Bling Capital says the old SaaS rules for judging a startup break in the AI era, because AI margins swing from negative to 90% and a big top-down market size hides what you can actually win. The fix is to size the business by gross profit instead of revenue, working out how many customers paying how much gets you to $100M of gross profit and what slice of the market that is. They also want the core problem written out as concrete before-and-after steps with real dollar ROI, because if you can't put it on paper, you don't really understand what your product is worth.

THE WEEKLY PLAN ALMOST NOBODY OFFERS

Fewer than 15% of subscription apps offer a weekly plan, yet weekly now drives 55% of all subscription-app revenue, a small slice of the field capturing most of the money. It works because a new user mostly wants to know whether it's worth risking $6 this week, and that small ask converts far better than asking them to commit $60 for the year. Adding a weekly tier takes about 20 minutes, and even the people who end up choosing monthly or annual convert better just because the weekly option is sitting there making the others look like a deal.

on-demand

THE FIRST MODEL THAT PUSHES BACK

Anthropic shipped Claude Fable 5 today, the first of the new Mythos-class models, and after five days of hard testing it's being called the best coding model in the world. The standout reaction is that it's the first model that felt like it was pushing the testers rather than the other way around, handling ambitious coding, writing, and agent work that earlier models couldn't touch. The bar for what one person can build just moved again, so the real question is what you'd attempt if the model could carry the whole build.

AGENT TOOLKIT FOR AWS

The Agent Toolkit for AWS gives AI coding agents the tools, knowledge, and guardrails they need to work with AWS services. It helps AI coding agents build, deploy, and manage applications on AWS. The toolkit works with Claude Code, Codex, Cursor, Kiro, and other coding agents developers already use. Over time, the best MCP servers, skills, and plugins from AWS Labs will be transitioned to the Agent Toolkit for AWS to ensure customers can access the broadest array of tooling and guidance for their agents.

DASH 2026: GUIDE TO DATADOG'S NEWEST ANNOUNCEMENTS

Datadog's DASH keynote this year showcased the company's Bits AI support for both developer and operations workflows. It also introduced dozens of new capabilities, including autonomous detection and remediation, AI-driven release validation and testing, and unified journey monitoring. Datadog is enabling teams to build better with AI. This post details all of the major keynote announcements.

HACKER NEWS SENT 10X THE TRAFFIC — TO THE ONE DEVICE IT COULDN'T INSTALL ON

Franz, a desktop app, got featured on both Hacker News and Product Hunt in the same week. Hacker News sent roughly ten times the traffic compared to Product Hunt. However, most of the traffic Hacker News sent was on mobile, while people were viewing the Product Hunt feature on laptops. Franz is a desktop app, and its site doesn't display a direct download link on mobile, so the traffic from Hacker News was less valuable.

YOU COMMISSION THE WORK NOW

An early tester got access to Fable and watched it work on its own for up to nine hours on a single brief, spinning up helper agents to research, code, and check each other. The shift he describes is that he now sets the goal and judges the result, while the model handles every step in between. The power is real and so is the unease, because hundreds of choices happen in a black box you never see, so the skill that matters becomes writing a sharp brief and knowing a good result when you get one.

pay attention. Learn more about sponsorship opportunities.

IN 2026, LESS THAN ONE THIRD OF GOOGLE SEARCHES STILL SEND A CLICK

There's been a 12.5% growth in clickless queries over the last two years due to AI features, instant answers, UI elements that keep searchers in the results, and shifting user preferences.

GET A LIFE, IT'S A CAREER STRATEGY

Specialization wins in stable, fast-feedback worlds like chess and golf, but building a company is the opposite - a messy place where the rules shift and no single skill covers it.

WHY GOOD COMPANIES GO BAD

A new book called Incorruptible, from the author who wrote The Lean Startup, digs into why good companies slowly go bad, drifting from serving customers to serving only the stock price.

YOUR MODEL PROVIDER IS NOW YOUR COMPETITOR

If your startup sells an AI product built on OpenAI or Anthropic, your supplier is now your rival. Both just launched billion-dollar consulting arms that send their own engineers into the same big companies you're selling to, building AI systems on the very models you license from them, the same way AWS watches what sells on its platform and then ships its own version.

THE STARTUP THAT LAUNCHED ITSELF

A startup inside Airtable ran its own launch almost entirely on its own agents, from cold outreach to the landing page to qualifying leads. The bet is that the model is now a commodity and the durable product is the harness around it, the skills, memory, and checks that make each run better than the last. The founders getting real leverage aren't aiming agents at busywork, they're rebuilding core workflows, so choose your scaffolding as carefully as your model.

AI JUST TOOK 40 OF THE APP STORE'S TOP 100

AI has crossed from hype into real consumer spending. More than 40 of the top 100 App Store apps now have AI features built in, and they grew sales four times faster than the rest of the top 100, inside a $1.4 trillion market. What's pulling the money is that an AI upgrade pays off the instant you buy it, a higher limit or a smarter coach, so the spending is landing on everyday apps like photos, health, and shopping, not just the chatbots.

YOUR AI STRATEGY HAS A TRUST PROBLEM, NOT A TOOLING PROBLEM

Many companies now have the technology they need to go faster. The blocker is the systems companies design to prevent things from happening. Rapid innovation can't happen in environments where employees aren't trusted. Companies need employees with agency - the ability to make decisions on behalf of the organization - or their AI innovation efforts will be wasted.

THE INEVITABLE FAILURE OF ONE-SHOT PROJECT FUNDING

Software requires ongoing investment, but this is a feature, not a bug or mistake. Software is a depreciating intellectual asset, and companies need to recognize and budget for this. Every line of code in software needs to be supported until the last customer deletes it. One-time project-based budgets ignore this reality, which inevitably leads to financial and technical crises.

YOUR NEXT SEO HIRE IS A PRODUCT MANAGER

Search is moving from ten blue links to AI answers, and the usual SEO hire, a channel manager who files tickets and reports rankings, is built for the old game. What works now is product thinking, understanding what a person actually means when they ask, then shaping the product so an AI model trusts it enough to cite. AI made the technical grunt work cheap, so hire for judgment, the person who can walk into a product review and defend a roadmap, not a keyword report.

WORDPRESS.COM CHANGELOG: WORDPRESS 7.0 AND WAYS TO REPURPOSE YOUR WRITTEN CONTENT

WordPress 7.0 shipped last month. WordPress.com sites have been automatically updated. Changes include post previews, navigation overlays, responsive block visibility, more design pieces, a font manager, and much more. The release also ships with AI features designed to assist in the editor, as well as real-time collaboration on selected plans.

GOOGLE MEET'S AI NOTE-TAKER NOW LETS YOU CUSTOMIZE YOUR MEETING NOTES AND TRACK DECISIONS

Google has rolled out an upgrade to its 'Take notes for me' feature in Google Meet that gives users more control over what gets captured, making it easier to track outcomes after meetings wrap up. Users can now toggle specific sections, like Decisions or Next Steps. The new Decisions tracker helps to categorize outcomes like 'Aligned' or 'Needs Further Discussion'. The Summary section has been tightened up to be more concise and scannable. Screenshots of the changes are available in the article.

WE ACCIDENTALLY BUILT A SECOND CODEBASE

Skills are small bundles of instructions handed to models so they can run the same workflows every time. Teams will often notice they keep doing the same things every time, so they write the workflow down as a skill and stop doing it by hand. Over time, this can result in a huge library of skills. These skills will eventually become obsolete as models improve. Developers need to monitor skill development so they don't build up and become a second code base that nobody uses.

OUR FIRST CUSTOMERS WERE THE EXCEPTION

The market always wins, even if a great team has a great product. You can't change a market, but you can always pivot to a more favorable sector. This requires a high level of execution and can mean essentially starting the company all over again.

99.1% TOTALITY

YC CEO Garry Tan recalls running low on Tesla charge in rural Texas and settling for 99.1% of the 2024 eclipse, only to learn that 99% is not 99% of the experience but zero percent of it. The same boundary applies to founders he now meets weekly with identical decks since AI collapsed the cost of building: the wedge, screenshot, and TAM math all match, but only some founders can carry the idea across into a real artifact where users feel the totality.

YOU'RE A FAR BETTER BET THAN YOU THINK

Paul Graham likes to show founders the odds their own valuation actually requires, and they're absurdly low.

AI-NATIVE SERVICES: THE DEFINITIVE GUIDE

Hanover Park has recreated core accounting systems for AI-native operation, enhancing efficiency by significantly reducing labor.

BECOMING YOURSELF IS A SOCIAL PROJECT

The people around you would help more than they do, they just can't see what you actually need.

TOP SAAS VENDORS ON RAMP (JUNE 2026)

Ramp reports DeepSeek as the top trending software for June, indicating renewed interest from American firms despite past security concerns. Companies are also increasingly using open source AI models and platforms like Fireworks AI and DeepInfra, moving away from OpenAI and Anthropic. Despite AI developments, design tools like Figma and Paper remain popular, showing resilience against new AI solutions.

APPS ARE BECOMING MEMES

When building software gets as cheap as posting a meme, the unit stops being a product and becomes a drop: made fast, judged on whether it lands, forgotten by Friday. Roadmaps stop mattering; taste, timing, and an audience start to. That's why some VCs now scout founders the way labels scout musicians: can you make a hit, and can you do it again?

HOW TO INFLUENCE BUYERS WITH LINKEDIN CONTENT

External voices are having a growing influence on buying committees. Decision makers trust podcasts, communities, thought leaders, and online trends more than sales reps. Companies that understand this stop being just sellers and start creating educational content in their industry. This post discusses how to use LinkedIn as a channel to help influence buyers.

HOW TO BUILD AN OUTCOME-BASED PRICING PLAN

The SaaS CFO Ben Murray lays out 12 design decisions for true outcome-based pricing, drawing the Sierra-style line that distinguishes outcome from consumption dressed up as outcome. Real examples include Intercom Fin at $0.99 per outcome, Help Scout at $0.75 per AI resolution with a free 3-month training window, HubSpot Breeze at $0.50 per resolved conversation and $1.00 per recommended lead, Zendesk's 72-hour LLM-verified resolution window, and EVE CoreGuard's $37.5K-to-$1.2M risk-avoidance ladder.

HOW TO INFLUENCE COMPLEX BUYING GROUPS

One excited champion isn't enough to close a deal anymore. Even a mid-sized sale now has to get past product, finance, security, legal, and the execs, and each of them cares about something different. Product wants clarity, finance wants the numbers to work, security wants to know it's safe, and your champion wants to not look bad for picking you. The trick is to tell one clear story and adjust it for each person, instead of using seven different pitches. Instead of asking buyers which message they like, look at how your existing customers actually decided to buy.

HOW MY NON-ENGINEERING TEAM AT SENTRY LEARNED TO SHIP

Sentry had a non-engineering team migrate around 2,500 pages of its marketing site to Markdown and code in a git repository over four months. Before the project, most of the team hadn't written code before. This post discusses why Sentry decided to have a non-engineering team complete the task and the lessons the company learned during the migration.

GOOGLE SEARCH ADDING PROFILE PAGES FOR WEBSITES AND CREATORS

Google's new 'Search profiles' allow website owners and creators to shape their own presence on Search. Search profiles can be accessed via a creator or publisher's Knowledge Panel, Discover feed, or through a direct URL. They can contain page and social media links, videos, and posts. More capabilities will be introduced in the coming months. The feature is coming first to the US before further expanding to more publishers and creators worldwide.

ASANA LAUNCHES AGENTIC WORK MANAGEMENT PLATFORM, PITCHING ITSELF AS THE ‘EASY BUTTON' FOR AI-DRIVEN ENTERPRISES

Agents fail in the enterprise due to a lack of organizational memory and understanding of roles and responsibilities. Asana has been building the infrastructure that fixes these issues even before agents existed to use it. Its Agentic Work Management Platform is a system designed to align humans and agents around the same plan, context, and governance. It is composed of a chief-of-staff agent called Dash, a new generation of AI Teammates, and three vertical applications that target IT, software development, and professional services.

SOME INVESTORS SHAPE MARKETS, WHILE OTHERS ARE SHAPED BY THEM

Homebrew's Hunter Walk pushes back on the "competitors are raising bigger funds, so I have to" reflex he hears from legacy and new VCs, arguing the chorus signals commoditization of capital and a slow regression toward mediocrity collected via larger fee paychecks. The first-principles questions VCs should ask before scaling AUM: why does the firm exist, what playbook is it best suited for, what resources are already in hand, what needs to be acquired, and what needs to be shed.

THE AI TREADMILL

AI will reduce repetitive, tedious work and make us better at many things. The question is now what we choose to do with the space it creates. You could use the efficiency gains to add more things to work on, or you can spend the extra time to think carefully and meaningfully about the things that actually matter. Creativity and ingenuity come from making space to think, daydream, and wonder.

HOW WE'D MARKET TO SOFTWARE DEVELOPERS AT STARTUPS

Audience research won't make developers love marketing, but it prevents budget waste and points you towards places where they're actually paying attention.

WITH ALL THE AI MONEY SLOSHING AROUND, SKIN IN THE GAME ISN'T ENOUGH ANYMORE

Startups can no longer rely just on "skin in the game" due to shifts in incentives from easily monetized options like secondaries and SPVs.

WHY SHIPPING FASTER IS MAKING YOUR PRODUCT WORSE

3 YEARS OF IMPROVING AN AI AGENT

Don't build for today's model, build for next year's.

Read the full guide →

THE TOKEN RECKONING IS HERE AND IT'S NOT WHAT YOU THINK

AI token costs are rising dramatically, leading companies to reconsider or cancel AI initiatives. EntelligenceAI reports that only 18% of AI spending results in productive output, causing concerns about the sustainability of current AI practices. Uber and others highlight the difficulty in justifying excess spending without direct productivity gains, signaling possible slowed growth in the sector.

THE NEW PRIMITIVES OF AI-NATIVE DEVELOPMENT

There's a quiet land grab happening in AI dev tools right now. Everyone wants to be the platform, but the market decides who actually becomes one, and that's whoever you can't leave without it feeling like surgery. The way you get there is to own a "primitive": one small, painful, unavoidable job that everything else ends up depending on, the way Stripe owns payments or Twilio owns texts.