The number is missed. The board wants an account of it. Sacking the agency feels like action. Then Monday arrives, and the target is still there.
The number is missed. The board wants an account of it. Somebody has to carry the blame before it travels further up the table. The agency is the cleanest name in the room: a contract, a notice period, a line that can be cut this quarter.
What comes next is treated as obvious. Source another partner. Hire in-house. Turn the channel down. The decision feels like action because something visible changes.
The part that does not change is the question the business still has to answer on Monday.
SO HOW ARE WE GOING TO HIT THE TARGET THIS TIME.
That question survives the firing. It survives the pitch. It survives the new starter in month one. Nine times out of ten it is the same question that sat under the old retainer, because the partner was never isolated as the problem.
The business still needs the capability. The next arrangement still has to live with the same offer, the same site, the same sales cycle, the same tracking, the same lag. There is no step-change waiting in a new logo. There is no wand.
Before you serve notice, look hard at the world after the retainer ends. Take that picture to the board. Do not kick the can. Do not plan to be gone in two years before the target is traced back.
Notice lands. Access gets listed. The Slack channel goes quiet. A handover is promised and arrives as a folder. The board pack for next month still has the same number on it.
Paid accounts do not hold their shape while they change hands. Tests reset. Naming conventions get rewritten. SEO compounding pauses while a new team learns what was already tried. Someone has to reconstruct the last eighteen months of decisions from change history and old decks.
The trough is not a rumour. It is 60 to 90 days of lower context and higher noise, against a target that was not rewritten when the supplier was.
If the plan for that trough is “the new people will be better”, you do not have a plan. You have a hope with a start date.
The incoming shop has sat in this meeting before. They will be wise. They will be savvy. They will tell you what the room wants to hear.
They will point at how badly the last partner missed. Sometimes that is true. It is also a sales move. A team with no history to defend can sound more commercial than the team that has been living with your constraints.
They will replay your projections as if those projections were a brief rather than a forecast. The number the board already believes in comes back to you in their slides, cleaned up and attributed to a better process.
Then the clause that makes the pitch safe: this will take time to fix.
NINETY DAYS IF THEY ARE BOLD. SIX MONTHS IF THEY ARE CAREFUL.
That window is not a turnaround plan. It is purchased silence. It buys them the period in which every account looks worse before it looks owned. By the time the grace period ends, they have the same lag, the same conversion rate, the same sales follow-up, and a target they helped legitimise in the room.
The tell is not optimism. It is a new partner treating your existing forecast as a fact while dating their own accountability.
A new agency can be the right call when the incumbent will not put seniors on the work, will not tell you the number is wrong, or will not change the plan when the assumptions break. It is the wrong call when the brief, the measurement and the commercial physics stay put and only the invoice changes.
The slide is tidy. Hire the capability. Bring it closer to the business. Cut the margin.
Year one can look cheaper. One salary instead of a retainer. No pitch theatre. “We own it now.”
Then the true cost settles.
The specialist needs cover for holiday, illness and the week the platform breaks. Cover needs a second person, or a contractor who quietly becomes a line. Somebody has to manage them. That somebody wants a title and a rise. HR arrives. Hiring freezes arrive. Politics arrive. The person who could run the account is now in meetings about the account.
Teams bloat for the same reason agency teams bloat: work expands to the size of the organisation around it.
In-house is not a discount button you press because the relationship got politically expensive. It is a decision to recruit, keep and manage craft inside your own walls. That can be the better model. It is the better model when you already know how to hire this skill, how to keep it, and how to review it with the same honesty you would demand from a partner.
It is not cheaper by default. It is cheaper only while the spreadsheet still has one name on it.
Ask it as a real question.
If the answer is yes, say that to the board with the revenue line attached. That is an honest commercial choice. It makes the gap visible instead of relocating it.
If the answer is no, because paid search or organic is still one of the few channels with the reach to scale the business, then you are no longer choosing “agency versus not agency”. You are choosing how this channel works here.
That is the forced logic:
WE WILL NOT TURN IT OFF → SO IT HAS TO WORK → SO THE PROBLEM IS THE PARTNER, OR THE TARGET, OR THE MEASUREMENT, OR THE CONDITIONS AROUND THE WORK
If those have not been separated, firing is a way of not saying which one.
The follow-up belongs in the same meeting. Are we better trying to figure this out with the people who already know the account, or do we want a new team to learn the same constraints from zero.
The current provider is not automatically right. They become the default once you admit the channel stays on and the number was never isolated to them.
“We’re not hitting target” is rarely only a performance sentence. It is a survival sentence. The board wants a number. Somebody has to own how it was set.
A target is not a fact. It is a stack of assumptions: conversion rate, close rate, average order value, lag, seasonality, tracking completeness, creative, stock, sales capacity, lead definition.
If the stack was never written down, the number is theatre. If only one party can be sacked for missing it, the number is also a transfer of political risk.
That is why the agency is such a convenient name. They sit downstream of the brief, the offer, the landing page, the CRM, the sales SLA and the budget stop-start. They are also the only party whose contract ends cleanly this quarter.
Blaming them can be rational politics. It is a poor diagnosis.
Joint accountability is not a values slide. It is a page that says what the brand owns and what the partner owns, what “good” means at 30, 90 and 180 days, and what happens when an assumption breaks.
Until that page exists, every miss will look like a supplier problem, because that is the only problem the process knows how to process.
Write the world after the retainer, on one page, before the email goes.
If that page cannot be written, the firing is not a decision. It is a deferral.
The line to take upstairs is not “defend the agency”. It is this: if we change supplier and change nothing else, we should expect the same number and a quieter first quarter. Here is what would actually have to change. Here is what we own. Here is what we are prepared to revisit in the target.
That conversation is harder than a retender. It is also the one that stops the loop: bloated brief, performative pitch, rigid number, missed number, new pitch.
None of this is a case for keeping a weak partner.
Fire them when they will not put the people from the pitch on the account. When they will not say the target is wrong. When they stop bringing a plan and start bringing a recap. When the relationship is only still alive because the notice period is awkward.
Do not fire them as the first response to a miss. Separate a bad month from a dead partnership.
Seasonality, an algorithm change, a tracking gap and a sales-cycle lag are not the same event as silence and junior swap-outs. If you cannot tell which you are in, you are not ready to change the model. You are ready to change the story.
You can change the name on the invoice. You can hire. You can cut the channel.
Only one of those is a commercial decision about demand. The other two are resource models. They do not invent a better number.
The capability still has to exist. The next partnership still faces the same challenge. Performance does not lurch because the slides are new.
What actually moves the number is a target tied to how the business makes money, measurement that both sides believe, and an honest split of what each side can control.
Explain that to the board now. The alternative is another twelve months of grace, another trough, another miss, and a marketing director who already knows how the next meeting will sound.
What agency scale actually buys you is operational distance, and six questions to ask before you sign.
Read →
Why growth needs unit economics, not a rigid ROAS target that switches off your next customers.
Read →
The belief REBEL was built on, and what it means for your account.
Read →This is what giving a damn looks like in practice. Tell us what's not working.