Name what’s slow
Thirty minutes, no deck, no discovery theatre. You describe what's slow. I tell you whether it's two hours of work or six weeks — including when the answer is "nothing, you're fine".
I diagnose why consumer marketing teams with good tools still ship late, and I fix the workflow before recommending anything new, in anything from a two-hour session (from €250) to a two-week pass. These are the six things I hear most often. Each usually traces back to one of three causes, and often two at once. Pick the one closest to yours and see where it tends to lead.
Thirty minutes, no deck, no discovery theatre. You describe what's slow. I tell you whether it's two hours of work or six weeks — including when the answer is "nothing, you're fine".
Not the process document — the actual briefs, the actual approval chain, the files people email each other at 22:40. Where the hours go is never where the org chart says.
What the machine owns, what stays human, and what it is never allowed to touch. I draw it on one page before anything gets built, because that page is the actual product.
Not a demo dataset, not a sandbox — this week's real campaign, run through the thing while your team watches and tries to break it. (They will.) If it can't survive a Tuesday, it doesn't ship.
Code, prompts, documentation, and the two people on your team who can now change it without me. If you need me again next quarter, I designed it wrong.
Two hours compresses them into one call. The two-week pass runs them in ten working days. A rebuild runs them over six weeks or more. The depth changes, the sequence never does, and you can see where you are the whole way.
One question, answered honestly: what is actually slowing you down? If a session solves it, you get told that instead of sold a project.
I map the workflow as it really runs, not as the process doc claims. Where hours go, where approvals stall, which 80% a machine should own.
What the system does and what stays human, both decided before I build anything. Automate the 80%, pour yourself into the 20%.
I build it instead of speccing it, and I test it against real work, not a demo dataset. A tool, a content system, a workflow your team runs on Monday.
Code, prompts, documentation and a trained team. An engagement that ends in dependency is a failed engagement.
Direction, not a scoreboard. These are the shapes that change when a content operation becomes a system — measured against your own baseline in week one, so you can see the move rather than take it on faith.
One pass replaces the round trip. The review that used to happen three times happens once, with the obvious corrections already made.
The mechanical eighty percent moves to the system. Your team's time lands on the twenty that actually carries the brand.
Code, prompts and documentation transfer with the build. Nothing about the system requires calling me again.
That's the one worth thirty minutes. No deck, no discovery call about the discovery call, no proposal you have to read twice.