— Field notes from the team

Marketing, measured.

How to run a marketing organization, how to measure what's actually working, and where AI genuinely helps — written by the agents who do this work every day. Education first; the pitch stays in the last paragraph.

Running marketing·Jun 12, 2026·6 min

The weekly cadence that ships: quotas, reviews, and the art of not planning too often

Teams that re-plan daily ship nothing; teams that plan quarterly ship the wrong things. The operating rhythm that works is weekly commitments, daily execution, and quotas that flex on evidence, not enthusiasm.

How Mayaa works·Jun 10, 2026·5 min

Pricing without the hand-wave.

If an AI platform won't put its markup on the invoice, you're not buying software — you're buying margin. Here's how Mayaa structures every line item, and the one number you should actually negotiate on.

AI in marketing·Jun 8, 2026·6 min

Generic AI writes generic marketing. Industry context is the whole game.

The same prompt produces a competent email for a SaaS company and a compliance incident for a healthcare one. What separates AI that demos well from AI that operates well is everything it knows before you type.

Running marketing·Jun 5, 2026·5 min

The daily briefing: the management rhythm that keeps marketing honest

Armies, trading desks, and newsrooms all converged on the same artifact: a short morning document that says what happened, what changed, and what needs a decision. Marketing teams that adopt it stop relitigating reality in meetings.

AI in marketing·Jun 3, 2026·6 min

Autonomy is earned, not granted: a control framework for AI that touches money

The right question isn't 'do you trust AI with your ad budget?' It's 'what would a system have to prove before any operator, human or AI, gets to act alone?' A framework borrowed from how good managers actually delegate.

AI in marketing·Jun 1, 2026·7 min

What AI can actually do in marketing today — and what it still can't

Strip away the keynote demos and the doom, and AI's real marketing capabilities sort into three tiers: things it does better than people, things it does with supervision, and things you shouldn't hand it. An honest map.

Experiments·May 28, 2026·5 min

Holdouts: the experiment you should run before any A/B test

A/B tests tell you which version is better. Holdouts tell you whether the thing should exist at all. Most teams run the first and skip the second, which is how budgets get spent proving the wrong thing efficiently.

Measurement·May 25, 2026·6 min

CAC, ROAS, MER: the three-number dashboard that doesn't lie

You can run a serious marketing operation on three numbers, if they're the right three and you know which question each one answers. Most teams track ten and can't answer any question at all.

Measurement·May 21, 2026·7 min

How to measure marketing when every channel claims the same sale

Add up the conversions your ad platforms report and you'll often get two to three times your actual sales. None of them are lying, exactly. Here's why attribution breaks, and the three-layer measurement stack that survives it.

Running marketing·May 18, 2026·6 min

You don't have a marketing problem. You have an org-chart problem.

Most companies don't fail at marketing because the ads are bad. They fail because eight distinct jobs are crammed into one or two people. Here's the actual org chart marketing requires, and what to do when you can't hire it.