Firing someone over ad performance is a strange decision to make, because the thing you're judging, Meta Ads results, is genuinely volatile even when everyone involved is doing everything right. iOS privacy changes, seasonal demand, a competitor doubling their budget, a creative that simply wore out: all of these tank performance with zero input from your media buyer's competence.
That volatility is exactly what makes this decision so uncomfortable. You don't want to fire someone for a bad month that wasn't their fault. You also don't want to keep paying someone who's been coasting for six months and blaming the algorithm every time you ask. The honest answer is that you need a way to tell those two situations apart, and "how the numbers feel" isn't reliable enough to do it.
The Wrong Question and the Right One
The instinct is to ask: "is performance bad right now?" That's the wrong question, because bad performance alone doesn't tell you whether it's a market problem or a person problem.
The right question is: "When performance got bad, what did they do about it, and did it make sense?"
A media buyer who's doing their job well, faced with a genuine downturn, should be able to tell you specifically what changed, what they checked, and what they tried. A media buyer who isn't should default to vague reassurance, "it's just the platform right now," "everyone's seeing this," "let's give it more time", without pointing to anything concrete.
Five Concrete Things to Check Before You Decide
1. Do they know their own numbers without looking them up on the call? Not perfectly, but a media buyer actively managing your account should have a rough, current sense of CPA, ROAS trend, and top campaigns without scrambling. Constantly needing to "check and get back to you" on basic account state is a real signal, especially months into the relationship.
2. Has anything actually changed when performance dropped? Ads Manager keeps a change history. If ROAS has been declining for a month and nothing in the account, budgets, audiences, creative, bidding, has changed in that window, that's not a market problem being weathered, that's an account not being actively managed.
3. Can they explain a decline in terms of a specific mechanism, not a vibe? "Performance is down because frequency climbed and the creative fatigued" is a real, checkable claim. "Performance is down, the algorithm's been weird lately" is not. The difference between those two answers tells you a lot about whether someone is actually diagnosing the account or guessing out loud.
4. Are they testing anything, or just running the same three ad sets? Some stability in a working account is fine. Total stasis for months, especially alongside declining results, usually means nobody is actively iterating.
5. Does their story match an independent read of the account? This is the check most people skip, because it means either learning Ads Manager well enough to verify claims yourself, or getting a second opinion. It's also the most decisive one, a media buyer's explanation either holds up against what the account actually shows, or it doesn't.
Where an Independent Score Helps
That fifth check is exactly what WizeScale exists to make easy. Connect your Meta Ads account, read-only, nothing changes, nothing gets touched, and WizeScale's deterministic rules engine reads the same data your media buyer sees and produces a plain-language health Score with specific, named issues if there are any: creative fatigue, audience overlap, a tracking mismatch, a budget change that didn't pay off.
You're not getting a competing opinion. You're getting a rules-based read of the actual account, which means when you sit down with your media buyer, you can compare their explanation against what the data independently shows. If the two line up, you have real confidence in the relationship. If they don't, you have something specific to raise, instead of a vague sense that things aren't working.
What a Bad Month Looks Like vs. What a Bad Fit Looks Like
A bad month, managed well: performance dips, your media buyer notices quickly, points to a specific likely cause, tests a response, and reports back on whether it worked, even if the fix didn't fully solve it.
A bad fit: performance dips and stays down, explanations stay vague across multiple conversations, changes in the account are rare or absent, and every check-in ends with "let's give it more time" without a specific plan for what changes in that time.
The first pattern is a relationship worth keeping through a rough patch. The second is worth acting on, regardless of how the next month's numbers happen to land, because the problem isn't this month's ROAS, it's the absence of active management underneath it.
Deciding With Evidence, Not a Gut Feeling
Don't decide based on how bad the numbers feel this week. Decide based on whether the person managing your account can explain, specifically and checkably, what's happening and what they're doing about it, and whether that explanation holds up against an independent look at the account itself.
If you want that independent look before your next conversation, check your Meta Ads account's health Score with WizeScale. Read-only, ready in minutes.
