Notes
Notes from the work.
Not a blog. Just things I keep saying to founders, written down. Short, plain, and from inside the accounts, not above them.
I dug out my 2019 BFCM plan. Eight things changed by 2026. Three didn't.
It was called the BFCM plan. Two days in the title, and that's where the thinking stopped. The season now starts in September and runs past Christmas into Q5, and Black Friday weekend is about a sixth of the money. What moved, what never did, and an interactive map of the seven moments that actually make up the quarter.
Read →Why European brands struggle to crack the US, and what it actually takes.
I've launched brands into new markets on both sides of the Atlantic, and the crossing into the US fails the same way every time. It's rarely the product. On the ad auction that breaks your math, the trust Americans need loud and up front, the plumbing nobody decks, and the one question most brands skip: whether the numbers even work before you spend.
Read →AI should be making your agency deliver more, not just earn more.
A couple of years ago it took me a team of five and a few agencies to run what I now run on my own. That same efficiency is sitting inside every agency in the world. The good ones hand it back to you. The comfortable ones keep it, same invoice as 2023. How to tell which you've got, and why the question only matters now.
Read →The 11 levels of Claude, for founders, operators and growth marketers.
A viral chart said there are seven levels of using Claude. It stops at the waterline. The whole ladder redrawn for founders, operators and growth marketers (output, cost, automations, the P&L), where I sit on it (graded honestly, by Claude itself), and a two-minute quiz that draws your map.
Read →Forecasting, explained for founders.
Most forecasts are a wish typed into a spreadsheet. How to build one that isn't: contribution margin first, new and returning revenue split, and three paths, an existing brand, a cold launch, and a new-market expansion (with the bedsheet lesson).
Read →Incrementality, explained for people who've been nodding along (like me).
I nodded at that word for a year before admitting I didn't know what it meant. It wasn't what I thought. The whole idea in three numbers, the rooster test, and how to run your first holdout without a data team.
Read →I dug out my 2019 DTC playbook. Four things changed. Three surprisingly didn't.
It wasn't wrong, it was right for its time, and the ground moved. What changed, what never did, and why I write this quarterly now.
Read →They told me the lifetime value metric was flawed.
A high-eight-figure business that couldn't name its own cost to acquire a customer. On the number that hides your costs, the one that doesn't, and the two P&Ls I run.
Read →No single tool tells you the whole story. So I stopped looking for one.
The analytics stack I actually run, which tool is for which job, why I built my own dashboard over the top, and the one read none of them can do.
Read →Agency, marketer, fractional. The order founders get burned in.
Three hires, one shape: each fixes the last gap and opens a new one. Why it's not bad luck, and what the fourth attempt has to be.
Read →What I actually look at in an audit.
Two to three weeks, your numbers, your accounts. Here's what I'm hunting for, in the order I hunt for it.
Read →Anyone can have the playbook now.
AI handed everyone the same frameworks. What it didn't hand out is reps. On what's left when the playbook is free.
Read →