Marketing Automation
Not another tool. The system underneath the tools — agents, a data repository, and the workflows that connect them.
Sevenam builds the system underneath your marketing — a repository of what the market is telling you, a set of agents with real job specs, and the workflows that connect them to the CRM and the tools you already own. Scoped and quoted after we map what you run.
Four fields, and no call booked yet. We read what you send and tell you whether you qualify for a free marketing technology, systems and performance audit — offered at our discretion and in limited numbers.
What marketing automation actually is.
Software that does the parts of marketing that do not need a person. Not the thinking — the collecting, the drafting, the chasing and the reporting, which is most of the week for most marketing teams. Three examples of the same idea.
None of it removes the person deciding. Every one of those steps ends with someone approving what happens next — which is the difference between automation you can leave running and automation you have to supervise.
The systems audit, and who we can do it for.
We run a small number of free marketing technology, systems and performance audits each month. It is a real piece of work — a read of your stack, your data and what is actually firing — and because it takes real time, it is offered only where we can see something worth finding. The call is fifteen minutes and decides whether you qualify.
Enough there to read
- A real marketing stack, whatever shape it is in
- Someone accountable for it internally
- Enough activity that the data means something
- A written read, and a call to go through it
- Yours to keep and act on, with us or without us
Not yet, and we will say so
- Pre-launch, with no CRM or ad data yet
- Looking for a document to take to another supplier
- Nobody who owns the stack day to day
- Both are fine positions to be in — neither is helped by this
- We say so on the call, and tell you what would change it
To be plain about it: the audit is offered at our discretion, in limited numbers, and nothing on this page is an offer to supply it — qualifying is decided on the call and we may decline for any reason, including simply being full that month. If we say yes, you get the scope and the dates in writing before we start.
Three parts, in this order.
Signal goes in on the left, the repository keeps it, agents work from it, and nothing that spends money or reaches a customer leaves without a person saying yes. Build them out of order and you get an agent guessing, which is what most AI marketing already is.
Drawn from our own design tokens rather than a wall of vendor logos, because the tools underneath a build differ for every client — the repository and the job specs are the system, and whatever software sits beneath them is an implementation detail we choose with you.
The engagement, in four phases.
Bespoke does not mean vague. This is the shape of every build, and what actually changes hands in each part of it.
The mapping workshop
- A full day in the room with your team
- Josh flies in for it
- Every tool, handover and manual step on the wall
- The first time most teams see their own operation drawn
The written audit
- What we found, ranked by what it costs you
- What it would take to fix, and in what order
- A call to go through it line by line
- Yours to keep and act on, with us or without us
The build
- The marketing data repository, on your infrastructure
- Agents with written job specs and run schedules
- CRM and workflow automation into your existing tools
- Approval gates on anything that reaches a customer
Training and handover
- Two live sessions with your team, recorded
- Written runbooks for changing a job spec or adding an agent
- Every account and credential in your name
- It keeps running if we stop working together
Training is on this list for a reason. A system only we can operate is a dependency rather than an asset, and your people have to be able to change it without calling us.
Why most marketing automation stalls.
Most companies already own the pieces. There is a CRM, an email platform, a scheduler, a dashboard, a folder of decks, and somebody using an AI assistant in a browser tab. What is missing is the thing that joins them, so the learning never accumulates: sales hears one version of the market, support hears another, the ads account sees what got clicked, and the founder remembers the one call that stuck. Everyone arrives at the growth meeting with a different version of reality.
The second failure is subtler. Work done in a chat window disappears when the tab closes. Ask for ten posts on Monday and by Friday the assistant has forgotten the objection that came up on every sales call, the phrasing that actually earned replies, and the offer that did not land. It starts from nothing every time, which is why AI output so often feels generic — it is generic, because nothing it needs is in front of it.
What we actually build.
Three things, in order. A marketing data repository — one place where what the market is telling you is written down and kept: the objections, the winning language, the offers that worked, the ones that did not, and the rules the business runs by. Then a set of agents with real job specs: what each reads, when it runs, what it produces, what needs a human before it goes anywhere. Then the workflows that connect the two to the systems you already pay for, including the CRM.
The point is not to remove people. It is to take the part of the work that is reading, collating, drafting and chasing — which is most of it — and give it to something that does it every day without getting bored, so the people you have spend their time on judgement, taste and the decisions that need a business context no system has.
What you own at the end of it.
Everything is built inside your own Business Manager from the first hour — the ad account, the pixel, the audiences, the creative files and the reporting. There is no agency-owned account that you rent access to, and nothing to negotiate over if the relationship ends.
That single structural choice is what makes the rest of it possible. An arrangement where the provider owns the asset has to be renewed; one where you own it has to be earned.
What gets built, in order.
Open whichever part you care about. Nothing here is a product you buy off a shelf — it is the same five or six pieces, assembled around what you already run.
The first build and the one that makes the rest worth doing. One place where what the market is telling you is written down and kept: the objections ranked by how often they come up, the language that earns replies, the offers that worked and the ones that did not, your thresholds and brand rules, and the corrections — so a mistake is made once rather than every week. It lives on your own infrastructure. Almost no company has this, which is why almost every AI output they generate reads generic.
Not a chat window. Each agent has a specification the same way a person would: what it reads, when it runs, what it produces, what good looks like, and who approves the output. One might read every sales call transcript and update the objection list. Another might draft creative angles from what performed last month. They write back what they learned, so the repository gets better rather than staying still.
Usually the fastest return in the whole build, because the data is already sitting there and almost nothing is being done with it. Records that update themselves from real activity rather than from someone remembering. Lifecycle stages that move on evidence. Follow-up that fires on a trigger instead of a calendar reminder. Lists that are correct on the morning you send rather than correct last quarter.
Nothing that spends money, emails a customer or changes a record leaves the system without a person agreeing to it. This is not a setting that can be switched off, and it is the reason the system can be trusted with a live account and a real list. An agent that can send on its own is a liability dressed as a productivity gain.
The connective tissue: your CRM, your email platform, your ad accounts, your scheduler, your warehouse if you have one. We do not migrate you to a new stack as the opening move, because the tools are rarely the problem — the absence of anything joining them is. Where a tool genuinely has to change we will say so and explain what it buys.
One document that says what moved, what it probably means, and what to do about it. Written by the system from the repository and the accounts, then read by a person before it reaches you. It exists because a dashboard tells you what happened and almost never tells you what to do.
It does not replace judgement, set your strategy, or decide what business you are in. It is not a compliance function — we build the approval gates, but the person approving has to be yours. And it does not fix bad data: if the CRM is empty or nobody owns it, the honest answer is that you are not ready yet, and we would rather say that on a fifteen-minute call than after a month of work.
Yes, and it is worth being straight about the shape of it: the advertising side of Sevenam is older than the automation side. The repository, the job specs and the approval gate come out of running our own operation that way — the daily decisions our Meta clients receive are produced by exactly this kind of system. If you want the automation line's own case studies, ask on the call and we will tell you honestly what exists and what does not.
Bespoke, quoted as a fixed figure in writing after we map what you run. There is no list price the way there is for the Meta products, because a company on one clean CRM and a company running four disconnected tools are not the same job, and pretending otherwise means one of you is being overcharged.
Questions people actually ask.
What does marketing automation cost here?
It is bespoke, and deliberately so — the work depends entirely on what you already run, how clean the data is and how much of it can be trusted. We map that first and quote a fixed figure in writing before anything starts. Unlike our Meta products, there is no list price for this, because there is no standard shape to it.
Is this just Zapier with extra steps?
No, though it will often use tools like that underneath. The difference is the repository and the job specs: a connector moves data between two apps, and what we build accumulates what the business learns so that every agent gets sharper rather than repeating itself.
Will this replace our marketing team?
It replaces a large amount of the work a marketing team does by hand — the reading, collating, drafting, reporting and chasing. What it does not replace is judgement about what should exist in the first place, which is the part worth paying people for. Most clients redeploy rather than reduce.
Do you work with our CRM?
Yes. CRM automation is usually where the fastest return sits, because the data is already there and almost nothing is being done with it. We work with what you have rather than moving you onto something new, unless what you have genuinely cannot do the job.
How long before anything works?
The repository and the first agent are usually live inside a few weeks. Whether that is worth having depends on your data, which is exactly what the mapping session is for — and if the answer is that you are not ready, we say that.
Who owns what you build?
You do, on your own accounts and infrastructure, the same as everything else we do. If the arrangement ends, the system keeps running.
What is the catch with the free audit?
That we choose who gets one. It is real work, we run a small number each month, and we offer them where we can already see something worth finding — which we decide on a fifteen-minute call rather than after a week of digging. If we decline, you get told on the call and you get told why. There is no obligation on you either way, and the written read is yours whether or not you hire us.
See whether your systems are worth automating yet.
Fifteen minutes with Josh, no pitch deck. Some accounts are not ready, and that is a faster answer than a proposal.
Get started Four fields, twenty seconds →Four fields, and no call booked yet. We read what you send and tell you whether you qualify for a free marketing technology, systems and performance audit — offered at our discretion and in limited numbers.