Seasonal vs Evergreen Facebook Ads (2027)

The two-layer plan: evergreen base, seasonal spikes on top

The cleanest way to think about a year of Facebook ads is as two layers stacked on top of each other. The bottom layer is evergreen: a small, always-on set of ads that sells your core product on its own merits, with no expiry date and no reference to any particular week of the calendar. It runs in January and it runs in July, and its job is to carry the baseline demand that exists whether or not anything special is happening. The top layer is seasonal: short bursts you switch on for a specific occasion, a Black Friday sale, a product launch, a gifting window, a category moment, then switch off again once the occasion passes.

Most advertisers only think in the top layer. They plan campaign to campaign, sale to sale, and between the peaks the account goes quiet or drifts. That is the expensive way to run ads, and the rest of this guide is about why. The base layer is not the boring part you tolerate until the next sale. It is the part that keeps your cost per result stable, holds your hard-won delivery data, and stops the account starting from zero every time you want to sell something.

The planning question is not "seasonal or evergreen." It is "how much of each, and how do I tell them apart when I read the results." Get the split right and the two layers reinforce each other: the evergreen base keeps demand warm and delivery efficient, and the seasonal spikes capture the extra intent that shows up around an occasion, on top of a base that never went cold.

What counts as evergreen, and what is genuinely seasonal

The test for an evergreen ad is simple. Could it run, unchanged, in any month of the year without looking dated or wrong? A product demo, a customer testimonial that does not mention a holiday, a founder explaining why the product exists, a clear how-it-works walkthrough, a review-led social-proof ad: none of these expire. They speak to the reason someone buys the product at all, which does not change with the season. That is the layer you build once, optimize hard, and leave running.

A seasonal ad is the opposite. It is tied to a moment and stops making sense outside it. A Black Friday discount, a gift-guide carousel for December, a new-year-new-you angle in January, a back-to-school bundle, a festive theme built around a holiday: each is sharp inside its window and dead the day after. Seasonal creative earns attention precisely because it matches what the buyer already has on their mind that week, and that relevance is also its short shelf life.

The two layers behave differently on almost every axis, which is why it helps to plan them separately rather than lumping everything into one calendar.

One nuance is worth stating plainly. Evergreen does not mean "the same ad file forever." A message can be timeless while the specific creative that carries it wears out. The evergreen layer is a permanent slot in your account that always has a fresh, working ad in it, not a single video you upload once and never touch. When the current evergreen ad fatigues, you swap in a new execution of the same durable idea. The slot is permanent, the file inside it is not.

Why the always-on layer protects your learning phase

This is the part most split decisions miss, and it is where Meta's own mechanics make the case for you. After you publish or significantly edit an ad set, it enters the learning phase while the delivery system works out who responds. The ad set exits that phase once it gathers roughly 50 optimization events, a purchase, a lead, or whatever you optimize for, within about seven days, according to Meta's Business Help Center. Until it does, delivery is unstable and your cost per result is inflated and noisy.

Here is the problem with a seasonal-only account. A campaign that runs for ten days around a sale, on a fresh ad set, often never gathers 50 events before you switch it off. It spends its entire short life in exploration, at the least efficient point of its existence, and then it is gone. You paid the learning-phase tax and never collected the reward of a stabilized, optimized ad set. Do that occasion after occasion and you are perpetually restarting from zero, always paying the exploration premium and never banking the return.

The evergreen layer fixes this by being the one part of your account that gets to finish learning and stay finished. An always-on ad set that has been live for months has long since cleared its 50 events, and as long as you do not make significant edits to it (a change to the audience, the optimization event, or the creative all count, and any of them restart the phase), it keeps delivering at its stable, efficient cost. That stability is an asset. It means your account always has a proven, low-variance baseline running, so a soft month is not a soft account.

It also changes how you should treat seasonal campaigns. When you bolt a seasonal burst on top of a healthy evergreen base, you are adding to a stable system, not rebuilding one. You are not touching the ad sets that already work, so you do not reset them. And because the base has kept your pixel and conversion data flowing all year, the delivery system has fresh signal to lean on when the seasonal campaign starts its own, shorter learning. A base that never went cold is the difference between a seasonal launch that finds its feet in days and one that flails through the whole occasion still trying to learn.

Deciding the split: a budget you can actually run

There is no universal ratio, but there is a useful starting point from outside Facebook. In their long-running analysis of advertising effectiveness for the IPA, Les Binet and Peter Field found that the most profitable campaigns split spend roughly 60/40 between long-term brand building and short-term sales activation. The parallel is not exact, but it rhymes. The always-on work that compounds gets the larger, steady share, and the burst-driven activation gets the smaller, sharper one. Sixty percent to the base, forty percent to the occasions, is a sane place to start before your own data moves it.

Work it as a reserve, not a fixed monthly line. Take a store spending $10,000 a month, or $120,000 a year. Put roughly 60%, about $6,000 a month, into the evergreen layer as steady daily spend that never switches off. That leaves about $48,000 a year for the seasonal layer, and the trick is that you do not spread it evenly. You pool it and concentrate it into the handful of windows that actually pay for themselves.

If you have five real occasions in your year, each seasonal window can carry close to $9,000 of concentrated spend, far more firepower than the same money dribbled across 52 weeks would ever have. That concentration is the whole point of a reserve. Seasonal demand is spiky, so seasonal spend should be spiky too. During a window, your total daily budget might briefly run two or three times its normal level as the reserve pours in on top of the untouched base.

Three things pull the ratio around. A business with genuinely strong seasonality, a gifting brand that does much of its revenue in the fourth quarter, will run leaner than 60/40 on the base and hold more in reserve. A business with steady, year-round demand, a consumables or subscription product, tilts further toward the base, because that is where nearly all the revenue lives. And a very small budget should tilt hard toward the base for a structural reason: you need enough steady spend for at least the evergreen ad set to clear its 50 events a week, so splitting a tiny budget across many short seasonal bursts guarantees none of them ever learn. When money is tight, one always-on ad set that works beats five seasonal ones that never stabilize.

The price of the calendar: seasonal reach costs more

Concentrating spend into occasion windows comes with a cost you have to plan around. Those are exactly the weeks when everyone else is bidding too, so reach gets more expensive. The auction is a live market, and seasonal demand inflates the price of impressions across the board.

Gupta Media's cost tracker, which runs on tens of billions of impressions, puts the blended Meta CPM for full-year 2025 at about $8.19. But the average hides the seasonal shape. The quietest stretch, January 2024, sat near $6.05 per thousand impressions, while the Black Friday peak in late November 2024 hit about $16.85 and Cyber Monday reached $17.70, the year's high. In round terms, a thousand impressions during the peak week cost close to double the annual average and nearly triple the January floor.

Source: Gupta Media cost tracker.

Two planning consequences fall out of this. First, a dollar of seasonal spend buys roughly half the reach a dollar of off-peak evergreen spend does, so your seasonal reserve has to be sized in impressions, not just dollars. Budget for the inflated CPM or you will reach far fewer people than the raw number suggests. Second, the arithmetic quietly rewards the evergreen layer. The base does much of its work in the cheap months, accumulating reach and conversions while impressions sit at their $6 to $8 floor, rather than crowding into the $17 peak with everyone else. Meta also reported that its average price per ad rose about 9% across full-year 2025, so the whole curve drifts upward year over year, which makes cheap off-peak reach more valuable, not less.

Going dark is the expensive option

The strongest argument for a permanent base is what happens to brands that let theirs lapse. The temptation is real. Between occasions, when nothing is on sale, it feels efficient to switch everything off and save the money for the next push. The effectiveness research says that is a false economy.

A study published in the Journal of Advertising Research, replicating and extending earlier work from the Ehrenberg-Bass Institute across 365 brands in 22 consumer-goods categories, tracked what happened to brands that stopped advertising entirely. Market share fell on average about 10% after one year dark, 20% after two, and 28% after three, measured against the last year the brand advertised. An earlier study in the same journal, looking at 41 beer and cider brands over two decades, found sales dropped about 16% after a year without advertising, 25% after two, and 36% after three. The decline was steeper for smaller brands and for brands already losing ground, which is precisely the profile of most businesses running Facebook ads to grow.

The takeaway for the two-layer plan is direct. The evergreen layer is not just an efficiency play for the learning phase, it is insurance against the slow erosion that sets in the moment a brand goes quiet. Demand does not politely wait for your next campaign. Between occasions, the base keeps you in the consideration set, keeps the pixel warm, and keeps the account from having to rebuild recognition and delivery from scratch every time you want to sell. The seasonal spikes capture attention; the base is what there is to spike on top of.

Measuring which layer carries baseline versus occasion demand

The planning is only half the job. The other half is reading the results so you can tell which layer is actually earning its budget, and the two layers demand two different questions.

Start by tagging them cleanly. Keep evergreen and seasonal in separate campaigns with names you can filter, so you can pull each layer's numbers without untangling them later. For the evergreen base, the question is steady-state efficiency: what is its cost per result and return on ad spend over a rolling multi-week window, at rest, with no occasion running? That number is your true baseline, the level the account produces on its own. Read it as a trend across weeks, not a single day, because the auction reprices constantly and one day proves nothing.

For the seasonal layer, the question is harder and more important. Did the burst actually add sales, or did it just take credit for sales the base would have made anyway? This is the incrementality problem, and it is where seasonal reporting flatters itself most. During a sale, last-click attribution happily assigns conversions to whichever ad was seen last, so a seasonal campaign can post a gorgeous return that is partly cannibalized from your evergreen base and partly demand that existed regardless of the ad.

Three practical reads get you closer to the truth. First, watch your blended ROAS, total revenue against total spend across both layers, not just the seasonal campaign's in-platform ROAS. If the seasonal campaign reports a 6x return but your blended return barely moved during the window, much of that 6x was borrowed from the base. Second, compare the occasion window against your established evergreen baseline: the incremental contribution of the seasonal layer is roughly the lift above what the base was already producing, not the seasonal campaign's gross number. Third, where the stakes justify it, run a geographic or audience holdout, a matched group you deliberately do not show the seasonal ads to, and compare. The gap between the exposed and held-out groups is the closest thing to a clean read of what the seasonal layer truly added.

Over a year, this reading habit tells you whether your split is right. If the evergreen base holds a healthy, stable cost per result and the seasonal bursts consistently show real incremental lift, the two-layer plan is working. If your seasonal campaigns look spectacular in-platform but never move the blended number, you are paying a premium in the most expensive weeks of the year to harvest demand your base already owned, and the ratio needs to tilt back toward the always-on layer.

A twelve-month view of the two layers

Zoom out to the whole year and the plan gets easy to run. The evergreen base is the flat line that never breaks: continuous daily spend, one or two always-on ad sets that have cleared the learning phase, with the specific creative inside them refreshed whenever it fatigues so the slot always holds a working ad. On top of that flat line you draw a handful of deliberate bumps, one per occasion that genuinely matters to your category, each sized from the seasonal reserve.

Pick those occasions on purpose rather than chasing every date on the retail calendar. Four to six real windows a year is plenty for most businesses: the ones where your buyers are already primed to act and the extra CPM is worth paying. Between them, resist the urge to invent a spike. A quiet month for the seasonal layer is not a quiet month for the account, because the base is still running, still converting at its cheaper off-peak CPM, and still holding the ground the going-dark research shows is so costly to give up.

The rhythm, then, is steady underneath and spiky on top. Keep the base always-on and optimized, hold a reserve for the occasions that pay, and read each layer with the question that fits it: efficiency for the evergreen line, incremental lift for the seasonal bumps. Keeping the whole job of running a Facebook ad, research, creative production, launch, and measurement, in one place, whether that is a platform like AdPlay.ai or a workflow you assemble yourself, makes the two layers easier to keep straight, because you can see the base and the bursts against each other instead of in separate tools. The split is not a one-time decision anyway. It is a ratio you tune every year as your own numbers tell you which layer is doing the work.