You have replaced the marketing agency at least once. You have changed the channels more than once, probably twice, because the last set stopped working and the new set was where everyone said the attention had moved.
And you have hired, and parted with, at least one head of marketing, maybe two. Each one arrived with a plan that sounded better than the last.
After all of it, you are roughly where you started. Activity, yes. Things get made. Posts go out, the pipeline number wobbles, a campaign lands and then it is over and nothing it built survives into the next one.
You are spending real money to stay in the same place.
So let me ask the question your last three decisions were quietly trying to answer. What if none of the things you replaced was ever the problem?
The Math You Already Ran Without Noticing
You ran an experiment over the last three years. You did not call it that, but that is what it was.
You changed almost every variable. Different agency, different channels, different person running it, different budget, different tools.
The one thing that did not change across all of it was the result. Nothing compounded. Year two started from roughly the same line as year one, and year three from the same line as year two.
When you swap every variable in a setup and the outcome refuses to move, the outcome was never tracking the variables you were swapping.
It was tracking the one thing you left in place the whole time. The constant.
And the constant was not the agency, because you changed the agency. It was not the channel mix, because you changed that twice.
It was not the person, because you changed the person, and the new person, who was genuinely good, produced the same non-result the old one did. The constant was something underneath all of those, something none of your three decisions ever touched.
The constant is that your marketing function has no structure for any of this work to accumulate inside.
Talented People With No Structure Still Produce Random Acts
This is the part that is hard to accept, because you did the responsible thing every time. You hired well. You picked credible agencies. You followed the budget where the results seemed to be.
A talented marketer dropped into a function with no spine does not produce compounding marketing. They produce well-executed random acts. Sharper than the last person’s, sometimes. Still disconnected, because disconnected work is the default output of any function with nothing to hold it together over time.
Structure is the thing that holds it together. I mean something specific by the word, not a reorg chart:
- A single source of truth for strategy that the whole function actually operates from, so a decision made in March still governs the work in September.
- An operating rhythm that repeats, so planning, production, review, and reporting happen on a cadence instead of when someone remembers.
- A filter that protects the plan from the idea everyone had in the shower this morning.
- A small set of numbers tied to revenue that the function is actually steered by.
- A clear map of who owns what, so the work has somewhere to live when a person leaves.
This is the same root underneath problems I have written about from other angles. A CEO who reaches for a new dashboard when what they need is a decision is reaching for a visible lever instead of the missing structure beneath it. Replacing the agency is the same reflex, one level up.
The lever you can see is rarely the thing that needs changing.
None of your three changes added any of that, because none of them was about that. A new agency brings its own process and takes it with them when they go. A new channel is a place to put work, not a way to make work add up. A new head of marketing inherits no structure, so they build their own from scratch, and when they leave it leaves with them, and the next one starts from zero again.
Every change reset the clock. That is why three years produced one year, lived three times.
Why Consistency and Continuity Are the Only Things That Compound
Marketing compounds the way trust compounds, slowly and only if it is the same thing showing up again and again.
A position the market hears once is noise. The same position, held across two years, becomes the thing you are known for. A piece of content is a piece of content. A point of view returned to, sharpened, and built on across forty pieces is an asset that sells while you sleep.
That return, that sameness over time, is consistency and continuity. And consistency and continuity are not things you get by hiring a more disciplined person or by everyone trying harder.
They are produced by structure, or they are not produced at all.
A function with a spine stays consistent through a personnel change because the consistency lives in the structure, not in anyone’s memory. A function without one loses its thread every time the org chart twitches.
The cleanest version of this I ever saw was in editorial. An issue went out every month not because anyone was heroic but because the work ran on a fixed structure: a calendar, a standard, a sequence of handoffs that did not depend on who was inspired that week.
The heroic version, the one that runs on a few talented people pushing hard, works right up until one of them is out, and then it produces nothing. Structure is what lets ordinary effort accumulate into something that outlasts the people who did it.
The clearest proof I have ever had of this came from running marketing for ten companies at the same time. That is only possible if each function runs on structure rather than improvisation, because no one can improvise ten things at once.
The ones that worked were not the ones with the most talented person attached. They were the ones where the structure carried the weight, so a good person could be senior and occasional instead of present and frantic.
The Seam Where It Shows Up First
There is one place the missing structure costs you in cash before it costs you anywhere else, and it is the line between marketing and sales.
With no shared definition of a qualified lead, no agreed handoff, and no common revenue goal, the two functions renegotiate their relationship every time the people change. Marketing throws volume over a wall. Sales calls the volume garbage.
Both are partly right, and neither argument ever resolves, because there is no structure holding the agreement in place from one quarter to the next. I have argued before that the marketing-sales gap is a trust problem that process alone will not fix. Structure is how that trust gets built and kept: a shared definition and a shared goal that outlast the two people who first agreed to them.
Whatever working understanding the last head of marketing built with your sales lead walked out the door with them. The new one starts the negotiation over.
The deals that fall through the gap in the meantime are the most expensive thing your missing structure produces, and they never show up on a marketing report as a marketing failure.
What This Actually Costs, and the Test to Run This Week
The cost is not the wasted budget, though that is real. The cost is the compounding you never got.
Three years of work that should have stacked into market position, into a function that runs without you watching it, into pipeline that builds on last year’s pipeline, instead reset to zero three times.
You did not buy bad marketing. You bought one year of marketing, three times, at the price of three.
So before you replace anyone or anything a fourth time, run one test.
Take your last three marketing efforts, the campaign, the hire, the channel bet, whichever three come to mind. Name the one thing each of them built that is still operating in your marketing today. Not the activity, which is over. The thing it left behind that the next effort started from.
If you can name something for each, you have more structure than you think, and the problem really might be the people or the channels.
If the honest answer is that nothing survived, that each effort started clean and ended clean and left nothing for the next one to stand on, then you finally know what the constant was. And you know that the fourth agency is going to produce the same year you have already paid for three times.
The people were never the variable worth changing. The thing holding them was missing the whole time.
