PostHog

Fin's startup pack is now live, giving you $100,000 in credits for loads of AI tools. ($50,000 of those credits are from PostHog 😉) https://lnkd.in/dzV5n2qy

Sesame Street has 56 seasons. Watching it end-to-end would take about 83 days. Users of your product could produce a similar runtime of session recordings in just a few hours. Session Summaries gives you the sizzle reel (and it doesn’t cost $15.49 a month). It’s free while in beta. Just ask PostHog AI.

After weeks of researching how PostHog and other top startups validated their product ideas, I just published a simple guide to doing just that. Here’s the process: 1. Start with a list of ideas. They should be in the form of problems AKA jobs-to-be-done by your product for a specific customer. It’s ok if that customer is *you* and the problem is *one you’ve had*. 2. Validate your problem is real by talking to potential users. Not working hard enough is a failure mode here. So is caring too much too soon. Ask people if they've actually had the problem and how they tried to solve it. 3. Validate users want your solution. This is where having two co-founders really helps. In PostHog’s case, Tim built and James sold. Not explaining your solution clearly and lacking credibility are two problems you can face here. 4. Validate your solution works by retaining users. This signals you are repeatedly solving their problems, a rare and valuable sign. Leverage feedback like user interviews, surveys, session replays, and revenue to make your product better. I go into a lot more details in the guide like how to get meetings, common failure modes, and what to do after validation. You can read all this here → https://lnkd.in/dyvQpJvr

One of the biggest risks scaling startups face is scarcity mindset. It often creeps in without anyone noticing as you become afraid of: - Losing your culture, market position, and momentum - Making the wrong hire - Taking a bet that doesn’t pay off - Looking stupid to your board, investors, or the market So you get conservative in ways that feel responsible but are actually corrosive. - You hire the “safe” candidate with the impressive resume instead of the weird brilliant one who might be special. - You pursue the “proven” strategy that worked for other companies instead of the unproven thing that might work for you. - You optimize for not-screwing-up instead of for winning, and the irony is that this is exactly how you screw up. A process for shipping a product goes from an idea and one conversation to a whole process of pitches, presentations, and approvals. This evolves into a process of pitches, presentations, and approvals. By the time you get to actually building something, the original insight that made the idea good had been committee’d into something safe and unremarkable. The scarcity mindset is insidious because it masquerades as prudence, the kind of thing serious companies with serious responsibilities should be doing. But it’s poison. --- This was one of my favorite insights from one of my (anonymous) teammates in our latest newsletter on how startups lose their edge. Read the full thing here → https://lnkd.in/gUweGyeM

PostHog AI can now summarize session recordings in batches. You don’t have to watch hundreds of them anymore - unless that's genuinely your thing. https://lnkd.in/grZXZtXh Pro tip: you'll need to be using used Session replay (summaries only work when there’s something to summarize).

Introducing Workflows, now in open beta: – Build automations with a drag-and-drop interface – no YAML, no API juggling. – Send targeted emails, Slack posts, or webhook messages to users based on live product data. – Branch logic visually based on user properties, cohorts or random variations, enabling targeted actions or A/B-style automation paths. Workflows runs on the product data you already track in PostHog. No need to connect extra tools or import events. Workflows is free while in beta.

How are top engineers making the most of AI tools like Cursor and Claude Code? They are always experimenting with new tools, models, agents, and frameworks. They are never content with their existing workflows. But doing this at the highest level requires one thing: MONEY 🤑. These tools aren’t cheap. Luckily for engineers at PostHog, our team knows this. Our finance team recently raised the budget for software tools from $100/m to $300/m for engineers. This gives them plenty of budget to try out top tier models like Claude Max. The logic is: If doing this helps our engineers ship 1% faster, the compounding benefit of the value they create in the long run will vastly out grow the cost in the short run. This is one of the many things I learned while researching how engineers us AI tools at PostHog. To read the rest, check out my post on AI coding mistakes to avoid → https://lnkd.in/gmMwb_7t p.s. if you’d like to join a team like this, we’re hiring: https://lnkd.in/g4b6v6cw

Writing a press release about how you can now connect Supabase databases to PostHog? = does not spark joy Releasing a limited edition t-shirt so you can dress like a supahero? = does spark joy https://lnkd.in/ecyCaZA2

One of the highest signal pieces of attribution is this box. A simple, option “Where did you hear about us?” on signup. Although only a fraction of signups fill it out, it helps us figure out: - Which creator sponsorships are working? - Is there creators we should be sponsoring but aren’t? - How popular is AI search from ChatGPT, Claude, Perplexity, etc.? - Do events drive signups? - Is there some channel we should be caring about more? - How much do referrals matter? - Where are we spending money but not getting mentioned? Attribution is often a mirage. UTM parameters might tell you users come from one source, but the reality of where they heard about you is likely different. A user could read a newsletter, search “posthog,” and click a Google Ad. This would tell us “Google Ads are awesome” when maybe that isn't the case. The “Where did you hear about us?” is coming straight from the user, which is often the best we can do, making this an insight I check daily. --- This was a combination of a couple of the lessons we’ve learned about about marketing for developers at PostHog. Read the rest here → https://lnkd.in/giPjBFCp

🎄'tis the season (for loud, knitted sweaters that give just the right amount of “I work at a startup but also understand winter.”) 🎄 While stocks last, at posthog.com/merch