I used to hate Google Analytics 4 until I sat down and created these 5 reports. I use all of them weekly and so should you…or just keep whining about how much GA4 sucks. Lets dive in...

fun dinner!

they made it complicated to create an industry around it with agencies that are incentivized to spend more (% of spend fees) Now they’re making it easier (MCPs) so everyone can be their own agency and spend more. The middle man is being cut out

good night for good sweet prince

openclaw is all fun and games until you have to refi your house to fuel your addiction Did these 7 things and now on track to save about $550 this month so I can keep clawin: 1) Model tiering (Sonnet for cron/sub-agents) — $400-$500 saved 2) ChatGPT subscription as secondary provider — $100-$200 saved (hoping for custom OC plan soon with higher limits) 3) Killing unnecessary heartbeats — $50-$150 saved 4) Cron jobs replacing constant polling — $50-$100 saved 5) Sub-agents for heavy tasks — $30-$80 saved 6) Trimming system prompt/context — $20-$60 saved 7) Local tools over paid APIs — $10-$30 saved

I’ve used Granola in meetings for a while. Wiring it into my AI agent is what finally unlocked it. I set up my OpenClaw agent to automatically read every meeting I'm in, extracts action items, posts summaries to Slack, and gives me a daily briefing at 8am with my priorities. Here's how it works: 1.⁠ ⁠Granola transcribes my meetings locally on my MacBook - no cloud, no API, just a local file (more secure, nothing exposed to internet) 2.⁠ ⁠My AI agent (Ace) polls every 30 min from a separate machine (my mac mini) over a private Tailscale tunnel 3.⁠ ⁠Pulls the raw transcript via SSH not Granola's AI summary, the actual transcript file (Granola has great speaker tagging, not great summaries) 4.⁠ ⁠Agent generates his own summary with strict rules to prevent hallucinations, every action item gets verified against the transcript before he posts 5.⁠ ⁠Extracts action items with owners, due dates, and priority then feeds them into a central tracker 6.⁠ ⁠Posts to Slack automatically - team meetings go through an approval gate so I review before anything goes public (just started this and working nicely) 7. Agent follows up with teammates as a thread reply to action items post, making sure no balls are dropped 8.⁠ ⁠Morning briefing at 8 AM - what's due today, what's overdue, what I'm waiting on from others Key learning was Granola has no API or export for transcriptions. But the data lives locally on disk. So my agent just reads the file directly over SSH. No copy-pasting. No manual note-taking. No worrying you forgot to do that one thing. Granola OpenClaw Tailscale = fully automated meeting ops.

exactly

People love pushing this conspiracy narrative, not favoritism if they did it for other matches at the same time

as business owners, confidence in sales from marketing has more to do with how easy it is to prove the last click came from each channel. the further the channel is from last click, the harder the attribution, and least confidence in the channel. but we know the last click channel isnt the only thing influencing a purchase decision so that view is too near sighted. develop a growth system to fill all parts of the funnel but your last click channels have to work well enough to fund everything else.

Was watching yesterday's ep of MFM on systems and Sam nailed something here most founders ignore. You're doing too much. Most companies can only take on one real growth project every six months. One ICP. One channel. Master it before you touch anything else. It's exactly how AppSumo scaled in the early days. We didn’t have a channel mix. We had one engine: email One customer type (the agency). One motion that kept compounding (get more emails, send more emails). Everything else came later. Ads. Affiliate. Influencers. Video. None of it mattered until the first channel worked like a machine. Most founders think "doing ads" means trying meta, google, linkedin, x all at once. But just ends up distracting the team, dilutes the budget, and too many channels to even know whats working. If you’re under $10M, you don’t need more channels. You need six months of deep focus on the one channel your ICP responds to most. Growth is simpler than people want it to be. It just requires more patience than most people give it. s/o @thesamparr aka the "Big System" for the video