Why Your Beauty Brand's Ads Stop Working (And the 3 Things That Control It)

Pennock Growth Notes · Paid Media

Why Your Beauty Brand's Ads Stop Working (And the 3 Things That Control It)

An account that scaled last quarter and stalled this one feels like the algorithm turned on you. It almost never did. Three levers control the outcome, and only one of them is the platform.

Here is the pattern we see every week. A beauty brand finds a winning ad, spend climbs, return on ad spend holds, everyone is happy. Then, usually without warning, the same campaign that printed money starts bleeding it. The instinct is to blame Meta, blame the auction, blame a mysterious update. So the team duplicates the campaign, resets the learning phase, and waits. Nothing improves, because the reset did not touch the real problem.

Ads stop working for three reasons, and the platform is only one of them. Creative fatigue, audience saturation, and signal decay. Learn to tell them apart and you stop guessing. Confuse them and you burn a month resetting things that were never broken.

First, it is almost never the algorithm

Modern buying on Meta and Google is stable. The systems do not quietly decide to punish a good advertiser. What changes is the input you feed them: the creative people see, the audience still available to reach efficiently, and the conversion signal that teaches the model who to find. When performance drops, one of those three inputs degraded. The algorithm is doing exactly what it always did, optimizing against a worse input. That reframing matters, because it moves you from waiting on the platform to fixing something you actually control.

Control 1: Creative fatigue is the usual killer

For beauty brands, creative is the single biggest lever, and it wears out faster than anyone plans for. Your audience saw the hero ad, then saw it again, then again. Novelty fades, the thumb keeps scrolling, and the click through rate slips before the return on ad spend does. By the time the revenue line dips, the creative has been tired for a week or two.

The tell is in the early metrics, not the late ones. Watch hook rate, the share of viewers still there at three seconds, and click through rate. When those fall while frequency rises, the creative is fatigued, full stop. The fix is not a bigger budget on the same asset, it is a fresh angle: a new hook, a different format, a new problem to open on. We wrote the full diagnosis in ad fatigue explained, and the way creative production itself has shifted in how Meta creative development changed with Andromeda. The brands that never seem to fatigue are not lucky. They have a queue of tested angles ready before the current winner slips.

Control 2: Audience saturation and frequency

The second lever is reach. Every audience has a pool of people who will convert efficiently. As you spend into it, you reach the easy buyers first, then the system has to work harder and pay more for each additional one. Frequency climbs, cost per acquisition climbs with it, and the campaign that looked broken is simply out of cheap demand at that spend level.

This one hides behind a volume trap. Pouring more budget into a saturating audience does not find new buyers, it just shows the same people the same ad more often and lifts your cost. We broke down what actually scales versus what only looks like scale in the ads volume trap. The fix is to widen the pool, not the budget: new audiences, new geographies, new platforms, or fresh creative that speaks to a different buyer inside the same audience. Saturation is a demand problem, and you solve demand problems with reach, not with a higher bid.

Control 3: Signal decay, or your measurement is lying to the algorithm

The third lever is the quietest and the most damaging. The algorithm only optimizes as well as the conversion signal you send it. When that signal degrades, from a pixel that fires inconsistently, a checkout change that broke an event, a server side setup drifting out of sync, or attribution windows quietly shifting, the model starts learning from noise. It keeps spending, but it chases the wrong people, and performance erodes in a way no creative refresh will fix.

Signal decay is why measurement sits at the center of everything we do right now. If you cannot trust the number, you cannot trust the optimization built on top of it. Start with the fundamentals in what return on ad spend actually measures, then audit the plumbing: is the pixel firing on every key event, does platform reported revenue roughly track your commerce backend, and are you reading a blended number rather than three platforms each claiming the same sale. When the signal is clean, the algorithm gets smart again on its own.

Quick triage: if hook rate and click through rate fell first, it is creative. If frequency and cost per acquisition rose while click through held, it is saturation. If the platform still shows a healthy return but your real revenue does not, it is signal. The symptom tells you the lever.

How to diagnose which one is hurting you

Do not fix all three at once, because then you never learn what worked. Diagnose in order. First look at the early creative metrics, hook rate and click through rate, over the two weeks before the drop. If they slid, you found it. If they held, check frequency and cost per acquisition for saturation. If both of those look fine and the account still underperforms against real revenue, you are looking at a signal problem, and it is time to audit tracking rather than touch the campaign at all.

This order matters because the three failures masquerade as each other. A saturated audience can look like tired creative. A broken pixel can look like a bad algorithm. Diagnosing in sequence stops you from refreshing creative to fix a tracking bug, which is the most common and most expensive mistake we see.

The fix sequence

Once you know the lever, the move is obvious:

  1. Creative fatigue: launch a fresh angle from your tested queue, do not just raise budget on the tired asset.
  2. Saturation: expand the audience or the channel before you expand the budget, and let new creative reach a different buyer.
  3. Signal decay: fix the tracking first, then judge everything against a blended revenue over spend number you actually trust.

Ads that stop working are not a mystery and rarely a crisis. They are one of three known problems with three known fixes. The brands that scale steadily are not the ones with a secret audience or a magic creative. They are the ones who can name which lever slipped this week, and reach for the right fix instead of resetting the campaign and hoping.

Frequently asked questions

How do I know if my ads stopped working because of creative or the algorithm?

Look at the early metrics before the revenue drop. If hook rate and click through rate fell while frequency rose, it is creative fatigue, not the algorithm. The platform optimizes against whatever creative you feed it, so a tired asset produces weak results even on a healthy account.

Will increasing my budget fix ads that stopped performing?

Usually not. If the cause is audience saturation, more budget just shows the same people the same ad more often and raises your cost per acquisition. Expand the audience or the channel before the budget, and refresh creative so it reaches a different buyer.

What is signal decay in paid media?

Signal decay is when the conversion data you send the platform degrades, from a misfiring pixel, a broken checkout event, or drifting attribution, so the algorithm starts optimizing toward the wrong people. Performance erodes even though creative and audience are fine. The fix is to audit tracking and judge results against a blended revenue number you trust.

Free Growth Audit

Not sure which of the three is draining your account? We will diagnose creative, saturation, and signal for you, and show you the fix. On the house.

Get your free audit
Nikki Lindgren