Evergreen Facebook Ads: Keep One Ad Live, Queue Its Replacement

Evergreen Facebook ads are ads built to run continuously, not for a sale or holiday. They come in two kinds, and the right one depends on what you sell and what you can measure.

If you sell a product people need every month, you want the always-on ad. If you sell something that starts with a sign-up, you probably want the funnel. Many accounts run both.

An evergreen funnel targets people who took a specific action within a recent window, so new people keep entering and older ones leave. A custom audience, a list of people Meta builds from their actions, does the sorting.

Jon Loomer's own evergreen campaign is a working example, with a webinar registration as the trigger. It uses seven ads in sequence across rolling windows of 1-5, 6-10, 11-15, 16-20, 21-25, 26-30 and 36-40 days since the trigger, plus exclusions so each person sees only the ad meant for their stage. It is optimized for Reach (Meta's objective for showing ads to as many people as possible) with a daily frequency cap, and it runs on feed placements only.

His simpler version starts with a single 10-day window. The problem it solves is stated plainly in his basic setup: "Those who don't convert keep seeing your ad, and it eventually becomes white noise."

For a store, the same logic applies with a trigger such as a product view or a started checkout. That is our adaptation, not part of his campaign.

Among 20 active beauty ads from three brands, counted on 7 November 2026, 18 had run under 90 days and the oldest had run 137 days. These are ages so far, not lifespans: every ad was still running, retired ads are not in the sample, and several ads share one launch date.

Here is how we counted. We took the first 20 of the 60 ads on our US Beauty trending-ads page, which sorts by best performing. They come from three brands: CeraVe, Sol de Janeiro and The Ordinary. For each ad we counted the days from its listed start date to 7 November 2026, when we opened the page.

The median age was 36.5 days and the oldest ad was 137 days. The ages, sorted, were 14, 14, 15, 16, 16, 17, 26, 30, 33, 36, 37, 61, 61, 63, 65, 65, 65, 65, 92 and 137.

This is a small sample: one category, three brands, one day. Four CeraVe ads share a 3 September start, two Sol de Janeiro ads start on 22 October and two Ordinary ads start on 24 October, so the 20 ads are fewer than 20 independent launches. Meta's Ad Library, the public database of active ads, shows no spend or results for US commercial ads, so the count shows what these advertisers keep live, not which ads win.

A start date also belongs to one listing, not one idea, as our Facebook Ad Library guide explains. A concept relaunched as a new listing gets a new date, so these ages understate how long a concept has run. Our recommendation is to plan for rotation, and to let your own numbers decide how fast it needs to happen.

Evergreen ads usually slow down because the same people see them again and again, or because the opening has worn out. Leave a winner alone only while its cost per purchase holds. Once it slips, change the surface, not the offer.

The first cause is frequency, the average number of times each person has seen the ad. In a funnel, the rolling window limits it. In an always-on ad, nothing limits it except budget and audience size.

The second cause is the concept wearing out for the people who already scrolled past it. A third cause is unrelated to the ad: a price change, an out-of-stock product or a slower landing page can look like fatigue.

Our recommendation is to rule out the third cause first, then compare frequency against results. Do not edit a healthy ad just to feel productive. Launch the new version beside it as a separate ad, so the old one keeps carrying spend until the new one earns it.

Run this audit once a week on every ad that has spent in the last 90 days. It tells you which ads to keep, refresh, replace or retire, using your own numbers.

Our creative audit scorecard and test capacity planner run these checks for free. For when to end a test, see our creative test stop rules.

A fading evergreen ad should be replaced by a ranked queue of challengers, not by whatever idea is available that day. Each challenger changes one thing about the winning concept: the hook, the proof, the format or the offer framing.

This is the step we built Deepsolv for. Our Creative Intelligence platform for Meta advertising teams combines ad-performance data, competitor activity, customer signals and past ad learnings into a ranked weekly test plan and execution-ready creative briefs. The audit above tells you what is fading. The plan tells you what to test next, in order.

Keep each challenger close to the original so you can tell what changed. Our guide to creative iteration covers how to vary a winner, and a creative test queue keeps the challengers in order. Meta's Andromeda update, its ad-retrieval system, favors creative diversity, so treat one floor ad plus two or three challengers as a starting point and keep the challengers visibly different from each other. Raise the budget on a floor ad only once the queue behind it is full, which is the order our guide to scaling Facebook ads follows.

Start with the audit on your top five ads by spend, name your floor ad and line up two challengers. That takes one afternoon and tells you whether your "evergreen" ads are holding.

If you want the challengers ranked for you, book a personalized demo with our founding team. See Deepsolv pricing.