When to fire your agency.
Most companies fire their agency about a year later than they should have, and then do it over the wrong thing.
The wrong things are easy to spot because they are loud. A bad month. A missed deadline in a quarter that had four of them. A new marketing director who wants their own people. Somebody was short on a call. None of those tell you whether the work is any good. They tell you a relationship had a bad week, which every relationship does.
Four signals actually mean something.
You cannot say what they are working on this week without opening an email to check. Not the plan for the quarter — this week. If the honest answer is that you do not know, the work has drifted out of the business and into a workflow, and it has probably been there a while.
The reporting has gotten longer while the decisions have gotten fewer. Report length tends to grow in inverse proportion to confidence. Forty slides is not thoroughness. It is a search for something in the data that will read as progress.
The people who won the account are not on the calls anymore. This one is so normal that it has stopped registering as a problem. It is still a problem. You bought judgment and you are receiving coordination.
Every recommendation is a thing they happen to sell. This is the strongest signal on the list. An agency that has never once told you to spend less, wait a quarter, or fix something outside their scope is not advising you. Watch what they recommend when the honest answer would cost them money. That single observation is worth more than a year of reporting.
One thing that is not a reason to leave: a channel that stopped working. Sometimes the auction got more expensive, sometimes the category moved, sometimes the offer stopped being competitive and no amount of media buying was going to hide it. Firing over that trades a team that knows your business for a team that will spend four months learning it, and arrives at the same finding.
When you do go, say the real reason once, plainly, to the person who owns the account. Almost nobody does. Agencies mostly get told the budget got restructured, which teaches them nothing and lets the same failure happen to the next client. Ten honest minutes on the way out is the most useful thing a departing client can leave behind.
We are an agency, and this is a strange thing for an agency to publish. It is also self-interested. A client who knows exactly when to leave is a client who stays for the right reason, and that is the only kind of long relationship worth having.
Do not wait for a bad quarter.
Watch what they recommend when it costs them.