A marketing team can have a perfectly sensible strategy and still make the work much harder than it needs to be.
Useful knowledge lives in old presentations, CRM records, documents and people’s heads. Every new campaign starts by reconstructing the same audience, proposition and evidence. Someone exports something from one system so somebody else can upload it into another. Reporting tells you what happened, but not necessarily what to do about it.
None of these things sounds particularly dramatic on its own. Together, they can consume a surprising amount of time.
Marketing systems is the part of my work concerned with how marketing actually operates: how knowledge is retained, how decisions get made, how information moves, how campaigns become executable and where repetitive work can be removed without removing the judgement that matters.
Sometimes technology is a big part of the answer. Sometimes it really isn’t.
A lot of marketing friction is hiding in plain sight.
Marketing teams accumulate ways of working over time.
A new tool gets added. Somebody builds a useful spreadsheet. A campaign process develops around whoever happens to be running campaigns. Sales has one version of the customer. Marketing has another. Important decisions get made in meetings and slowly become less important as people forget why they were made.
The result isn’t necessarily a broken marketing function. It’s often just unnecessary friction.
You see it when people have to search five places to find something they know exists. When every brief needs the proposition explaining again. When campaign planning means starting with a blank document. When useful customer evidence is technically available but practically impossible to find.
Or when a team has plenty of data but still needs somebody to work out what the data means before anyone can act on it.
This is one of the more interesting opportunities.
A lot of campaign planning is not new thinking. It’s remembering things the business already knows.
Who are we trying to reach? What do they care about? What’s our proposition? Which claims can we actually substantiate? How do we talk about this? What have we tried before? What did we learn? What does sales need from the campaign?
If that knowledge has to be rediscovered every time, an idea can spend a lot of time becoming a brief before it gets anywhere near becoming marketing.
A better system can retain more of that context.
Then, when somebody has a campaign idea, they’re not starting from nothing. They can start with the business, audience, proposition, evidence, brand principles and previous learning already available.
From there, the system might help turn an idea into an initial campaign structure, audience choices, messaging, channel thinking, briefs, first drafts and the work required to get it moving.
That does not mean pressing a button and allowing a machine to run the marketing department.
It means spending less time repeatedly rebuilding what is already known.
The interesting bit isn’t automated content production. It’s reducing the distance between thinking and doing without losing the thinking on the way.
AI is useful when it has a useful job.
AI has made some of this considerably more practical.
It can be good at retrieving context, organising knowledge, applying agreed thinking consistently and producing useful first versions. It can help interpret patterns in information, move data between systems and remove repetitive steps from a process.
Used properly, it can also make marketing knowledge more accessible.
Instead of the reasoning behind your proposition sitting in a deck from last year, it can become something the team can use while planning a campaign. Instead of somebody having to remember every relevant proof point, the system can retrieve the ones that fit the job.
That’s useful.
It still doesn’t decide whether the proposition is any good.
AI doesn’t remove the need for strategy, good information, clear ownership or sensible processes. It doesn’t know that a metric is distracting you unless somebody has established what actually matters.
And making a bad decision happen automatically is not much of an advance.
The technology is useful when it supports better marketing. It isn’t the proposition in its own right.
Not every marketing system needs much technology at all.
Some problems genuinely need connected tools, automation or more sophisticated use of AI.
Others need a better planning template. A clearer CRM rule. A shared source of truth. A meeting where people make the decisions they have been avoiding.
The point is to make the work easier to do well, not to make the solution look technically impressive.
A CRM problem can turn out to be an ownership problem. A reporting problem can turn out to be a decision problem. An AI problem can turn out to be that nobody has agreed what good looks like.
Start with how the work actually happens.
I don’t begin by deciding which system you should build.
First I want to understand the existing one, whether anybody has deliberately designed it or not.
How does a campaign currently get from idea to market? Where does the information come from? What do people repeatedly have to find, recreate, copy or check? Where is human judgement genuinely important? What already works?
From there, we can improve the parts that matter.
That might mean creating a reusable marketing knowledge base. Making campaign planning more repeatable. Improving how CRM and marketing information connects. Building better rules around signals and actions. Reducing manual reporting. Creating an AI-assisted workflow around an existing process. Or simply making the process itself clearer.
I’m comfortable building the first useful version myself.
Where something needs deeper engineering, platform expertise or technical implementation, I can work with the people who already own that internally or with the right specialist.
The aim isn’t to create dependence on a clever system. It’s to leave marketing easier to run.
You don’t need to know whether you need a marketing system.
You may just know that too much time disappears between having an idea and getting something done.
Or that the team keeps reconstructing things it ought to know already.
Or that you have plenty of tools and data, but the work still feels more complicated than it should.
That’s enough to start with.
We can work out where the friction actually is before deciding what, if anything, needs building.
If that sounds familiar, get in touch.