08 Aug Cost Per Result Suddenly Spiked? Don’t Kill the Campaign
Here’s a pattern I see constantly: a campaign runs well for two weeks, cost per result sitting comfortably around $10. Then, out of nowhere, it jumps to $15, $18. The instinct is to assume the campaign has stopped working and kill it. That instinct is usually wrong, and acting on it usually costs more than it saves.
Why killing it fast is the wrong move
If a campaign worked well for two weeks and then got expensive for a couple of days, that’s not the same story as a campaign that never worked. Killing it immediately does two things, both bad: it throws away whatever learning the ad set had already built up, and it discards a proven performer based on a few days of noise rather than an actual diagnosis. The right question isn’t “is this still working” — it’s “what changed, and is it something inside my control or outside it.”
Check CPM first
Before touching the campaign itself, check CPM. A meaningful share of cost-per-result spikes — commonly somewhere in the 50-80% range of cases — trace back to CPM shifting with broader market conditions, not to anything actually wrong with your creative, audience, or offer.
Meta runs on open supply and demand. During festive periods or other high-demand windows, more advertisers compete for the same placements, and CPM rises across the board as a result. If CPM is the culprit, that has nothing to do with your campaign specifically — it’s a market condition you have to pull through, not a problem to fix. Build your own CPM baseline over time so you can recognize when a jump is genuinely unusual for your account versus expected seasonal movement.
If it’s not CPM, look at where the clicks are actually going
If CPM looks normal and the spike is coming from somewhere else, the next place to check is whether your clicks are converting into the right next step.
Check the gap between clicks (all) and link clicks. If overall clicks are healthy but link clicks — the ones that actually matter — are comparatively small, that’s a sign the creative is pulling in the wrong kind of attention. This does happen: one case involved a client who adapted a TikTok-style creative for Facebook. It generated excellent overall engagement metrics, but very little of that translated into the actual action wanted. Good engagement numbers don’t automatically mean good results — sometimes a creative is simply attracting the wrong response.
Check landing page view against content view. If people are reaching your landing page but not browsing further, something’s disconnected between what the ad promised and what they found. In one case, this literally traced back to an ad promoting one product while the link sent people to a completely different product page — an easy mismatch to create, and one that quietly tanks performance without looking like an obvious setup error.
The mindset that actually works
Every fluctuation is a diagnostic opportunity, not a verdict. Build a hypothesis about what likely changed, check it against the data, and only then decide what to adjust — the same one-variable-at-a-time discipline that applies to any optimization: change one thing, give it time to show a real signal, then move to the next.
Campaigns can run far longer than most people assume
Worth remembering while you’re deciding whether to kill something: a well-structured campaign with the right exclusion strategy doesn’t have a natural expiration date. One example ran for roughly 25 to 26 months continuously. The pattern that keeps a campaign alive that long is straightforward — it stays evergreen because engaged people get excluded and retargeted, and fresh people keep flowing in at the top. Killing a campaign after a few rough days is often ending something that had years of life left in it, over a problem that wasn’t actually about the campaign at all.
A quick self-check before you kill anything
- Has this campaign been genuinely broken, or worked well until recently? Those call for very different responses.
- Have I checked CPM against my own historical baseline before assuming the campaign itself is the problem?
- Is there a seasonal or market-wide reason (a festive period, a high-competition window) that could explain a CPM jump?
- Is there a gap between clicks (all) and link clicks that suggests the creative is attracting the wrong response?
- Does my landing page actually match what the ad promised, or could there be a mismatch between what people expected and what they found?
This breakdown is written by Jason Gan, a Meta Certified Professional and Badged Meta Business Partner who has personally audited over 1,000 advertiser accounts since 2010. You can see real account breakdowns on the Jason Gan YouTube channel.
Frequently Asked Questions
Should I kill a campaign that’s suddenly getting more expensive after weeks of good performance? Not immediately. A campaign that’s worked well for two weeks and then gets expensive for a couple of days is far more likely to be reacting to a temporary market shift than to have genuinely stopped working. Killing it right away throws away a proven performer based on a few days of noise, and forfeits whatever learning the ad set had already built up.
What’s the most common reason cost per result suddenly increases? CPM shifting with market conditions — a large share of cost spikes trace back to this, not to anything wrong with your campaign. Festive periods and other high-demand windows bring more advertisers into the market competing for the same placements, which pushes CPM up across the board, independent of your creative or targeting.
If it’s not CPM, what else could cause a sudden performance drop? Look at whether clicks are converting to the right next step. A large gap between clicks (all) and link clicks suggests the creative is attracting the wrong kind of attention. A drop from landing page view to content view suggests the traffic reaching your page isn’t matching what the ad promised — sometimes literally landing on the wrong product page relative to what the ad advertised.
How long can a well-managed Meta ads campaign realistically run? Well beyond what most advertisers assume. With the right exclusion strategy keeping it evergreen, a single campaign can run for well over a year — one real example ran for roughly 25 to 26 months. Campaigns don’t have a built-in expiration; what usually ends them early is advertisers reacting to short-term noise instead of diagnosing it.
Book a Meta Ads Audit — $80 / 30 Minutes →
No login required. If your cost per result has spiked and you’re not sure whether to kill it or wait it out, we can find the real cause together in one session.
If you want ongoing help reading these signals correctly instead of reacting to every fluctuation, that’s exactly what 1-on-1 coaching is for.