What is in, what is not
Marketing OS is the product for the brand that has decided to stop outsourcing and does not want to find out in month three that nobody owns the calendar.
| In the fixed fee | Not in it (and what we do instead) |
|---|---|
| Roles: who decides, who produces, who publishes, who reports, written for the team you have and the one you are hiring | Hiring people: the roles tell you who to hire; we can sit in on the interviews, as a favour, not a line item |
| A weekly rhythm (the Monday plan, the Thursday review) and a monthly one (the report, the calendar, the context update) | Running the rhythm for you: that is the Marketing Subscription, if you want a strategist inside |
| The tool stack: what you already pay for, what to drop, two or three tools that close gaps, an owner for each | Tool licences: on your accounts, in your name, at the vendor's price |
| The brand context: voice, facts, do's and don'ts, product and offer list, season, assets, in one package for every tool and every freelancer | A brand book or a new identity: if you have one, the context is built from it; if you do not, a branding studio comes first |
| Templates for what you publish every week and playbooks for what you do every season: a launch, a promotion, a press moment | The content itself, month after month: the Social Content Engine does that, or your team does, from the templates |
| Two training sessions with the team, the operating manual, and one check-in thirty days after handover | Ongoing support tickets: the manual is written so the answer is usually on page two |
The weeks, one by one
Week one, what you have. Interviews with the people who will run marketing, an inventory of the tools you pay for and the ones you use, a reading of the last three months of what you published and what it did. If a Discovery came first, its run model is the starting point. End of week one: a page that says what works, what is missing, and who is doing two jobs.
Week two, the operating system. Roles, written as a list of decisions and a list of tasks, with a name against each. The weekly rhythm: the Monday plan of what ships this week, the Thursday review of what shipped and what it did. The monthly rhythm: the report in units, the next month's calendar, the context update. The stack: what to keep, what to drop, what to add, who owns which account, and what goes into an AI tool on what terms (customer data only when the work needs it, under a data processing agreement and the data map; never unreleased numbers or anything under NDA). End of week two: the operating manual, first draft.
Week three, the brand context. The working file every tool and every freelancer starts from. Voice: ten examples of what you sound like, ten of what you never sound like, and why. Facts: the product list, the numbers you are allowed to say, the claims you are not. Do's and don'ts: from "always name the city" to "never say innovative". The current offers and the season. Where the assets live and which ones may be used. Then the templates: the post, the newsletter, the launch page brief, the monthly report, each with the context already inside. End of week three: the context and the templates, tested on a real week of content with your team.
Week four, the handover. Two training sessions, two hours each: one on the rhythm and the tools, one on producing with the context, live, on next week's content. The manual, final. The check-in date in the calendar, thirty days out. For a small team with a clear brand, weeks one and two compress into one, and the whole thing is two weeks.
What a brand context is, and why it stops content being generic
An AI tool given no context produces the average of everything it has read, which is what generic means. A freelancer given no context produces what worked for the last client. The brand context is the fix for both: one package, updated monthly, that says who you are in a form a tool or a person can use on the first day. Your voice in examples rather than adjectives. Your facts, so nothing gets invented. Your do's and don'ts. Your products, offers and season, current. Your assets, and which ones may be used.
Every template starts with the context inside it. Every freelancer gets it before the brief. Every AI tool gets it pasted in before the first prompt. And a person on your team decides what goes out, because context narrows the output and an editor chooses it. That combination is what makes in-house content sound like you rather than like everyone.
On top of the brand context we set up the agents: one AI agent on each part of the marketing, skeptic agents that check every draft against the facts and the voice, and a person who decides. The full version runs inside our own app: the AI system at Stadami, nineteen parts of a business on one brain.