How long does a Meta ad last before it stops working for a newsletter?

Everybody says you may have a great ad for a week. I've used the same ad for three months in places, and it's still pretty darn good. Not amazing, but pretty darn good.

"It slowly degrades over time, but for way longer than you think."

How long has one ad actually lasted for me?

In August 2025 I was spending $500 to $600 a day on Meta ads and had one creative. One. Same one all over the country for six months, with only the newsletter name and location changed. By late 2025 I'd been running that same ad for nine months in every location.

• It took 14 of my newsletters to 10,000 subscribers.
• A partner's newsletter hit 15,000 on it and it was still working.
• It works across markets because local ads aren't national. Nobody in Michigan sees the Louisiana ad.

That last point is why local media is so good for this. The same concept can run 20 times for 20 different brands.

Why do costs rise after launch?

Because Meta finds the easiest people first. In the first two to four weeks you get people who love events, are on Facebook or Instagram a lot, and like signing up for things. Numbers are usually really good. The 15 to 20 cent numbers people share are real, but you might only get them for a week or two if you're lucky.

Once those are gone, it gets a little harder and costs go up a bit. Every local market has a capped audience, so the primary campaign's cost drifts up as the reachable audience gets used up. At one point my newest newsletter was getting subscribers for 20 cents while the market I'd advertised in the longest, on the same ad, had climbed to 60 or 70 cents. A guest on my podcast warned that in a small area you can burn out the audience quickly, so you have to keep fresh creative coming.

The wall for my one-ad approach is somewhere around 5,000 to 10,000 subscribers in a market, sometimes 10,000 to 20,000. By then everybody in town has seen it multiple times. Past that you need new creative, and that's the hard part.

When should I refresh the creative?

When the numbers say so. The pattern on my event-angle concept: cost per subscriber starts amazing, creeps up, and stays in a good place for a long while, even on brands I've run for a year. When it gets closer to a dollar plus, I push new variants.

• Refresh the creative before you change anything else.
• Don't widen the radius first.
• Don't blame seasonality before you've changed what you control. One creative left static in every location forever is not best practice, not even close.

Context matters too. Jas from Winnipeg Digest told me on my podcast he was at about 33,000 subscribers and still getting them for under a dollar. At that size, around 75 cents is pretty good.

Should I trust Meta when it says my ad is fatigued?

No. Meta once told me my creative was fatigued while I was getting subscribers for 19 cents, which is basically as low as you can get. To me, fatigued means I'm spending too much money.

I do look at the trend. Going from 4 cents to 12 cents in a week is a big jump. Going from 11 cents yesterday to 12 today is noise, so keep doing what you're doing. If I'm unsure, I leave the current ad on and launch another next to it. If the new one is way worse, turn it off. If it's way better, turn the old one off.

How often should I make new ads?

For local media, shoot for new creative once a month, maybe every other week. The real crazy marketers do a hundred or even five hundred ads a week, but that's national e-commerce spending $50,000 a week. Better video and better text take real manpower, so I wouldn't push it.

Expect most new ads to lose. The creative is by far the most important thing, and most tests won't beat your standard winner. When I tried AI creatives and agency-style concepts, they flopped horribly, up to 10 times worse than what I already had. My fallback was to keep the old ad and eat the higher cost. Even a tired winner beat them.

That's why I want a creative specialist once I'm running the same setup in 40 to 50 locations. At that scale I can pay someone about $3,000 a month, because they build it once and duplicate it in every city.

How else can I slow fatigue?

• Keep the concept, change enough to catch the people you missed. New background videos of fun local things, new text angles. Two things stay in every version: the city names and the word free.
• Rotate small elements. On one news brand I planned a female-voice version of the ad's song to rotate creative.
• Run two copy variants at once over the same video.
• Exclude existing subscribers once your list is big enough, so you stop paying to reach them.
• Bring a second, different concept to reach the next tier of people after the eager signers are gone.
• Or just stop. When the ad for my own town's newsletter started dying, I turned it off. I had enough subscribers for the moment.

If the cost suddenly blows up instead of creeping, that's a different problem. See what to do when Meta ad costs spike.