How many ads should you run on Meta? Weekly volume by spend level, and why the creative is now the targeting

Run as many new ads each week as your testing budget can give about three times your target cost per purchase each, which for most brands spending 10,000 to 100,000 a month works out at somewhere between 4 and 40 new ads a week. Keep 3 to 6 live in any one ad set, keep your proven ads running alongside, and expect Meta to put most of the money behind two or three of them. The right number is set by your budget and your cost per purchase, not by a rule of thumb about ads per ad set.

Almost every guide ranking for this search answers a narrower question: how many ads to put in one ad set. The common answer is 3 to 5, sometimes up to 15 for bigger spenders, as in this guide to ads per ad set. That is useful, but it is not what most founders are really asking. The real question is how many new ads you need to make and launch every week to keep the account growing, and how that changes as you spend more. That is what this piece answers, using what we have seen across the accounts we run.

Why the number of ads matters more than it used to

A few years ago the lever on Meta was the audience. You found the right interests and lookalikes, and the ad was a way of talking to them. That has flipped. Meta's own engineers describe their ad retrieval system, Andromeda, as a personalised engine that narrows tens of millions of candidate ads down to the few thousand most relevant to each person, built to scale with a growing volume of ad creative (Meta Engineering). In practice, with broad targeting, the ad itself does most of the work of finding the buyer. A video about a mum at bedtime finds parents of children who will not settle. A price-led static finds people comparing prices. Each different idea opens a different pocket of buyers.

That is why volume matters. One good ad can only find the people it speaks to. Ten genuinely different ads can find ten groups. We saw this in the first August Weekly: on a kids' supplement brand, general UGC video missed badly, while one ad built around a single bedtime moment returned nearly three times as much. Same product, same targeting, a different idea, a different buyer.

The catch is that "different" means different. Five versions of the same video with new opening words are one idea to Meta, not five. Across our accounts over the last 90 days we logged 109 lessons about angles, 92 about hooks, 83 about formats and 61 about personas. The angle and the persona, who the ad is for and what it says to them, are where the big swings come from. Hooks and edits fine tune a winner. They rarely create one.

How many new ads per week, by spend level

Here is the maths we use. Take the share of your budget you put on new ads, usually 15 to 25 percent. Divide that weekly test budget by three times your target cost per purchase. The answer is how many separate ideas you can afford to read properly each week. We explain where the three times figure comes from in our guide to creative testing budgets on Meta.

A worked example with a target cost per purchase of 40 and 20 percent of spend on testing:

Monthly spend	Weekly test budget	Ideas you can read each week	New ads to make (1 to 2 per idea)
10,000	about 460	about 4	4 to 8
30,000	about 1,380	about 11	11 to 20
50,000	about 2,300	about 19	19 to 35
100,000	about 4,600	about 38	38 to 60+

Swap in your own target. If your cost per purchase is 80, halve the right hand columns. If it is 20, double them. A lead generation account with a target cost per lead of 15 can read far more ideas on the same budget than a furniture brand with a target of 250.

Two things this table tells you that the "3 to 5 ads" advice does not:

• A small account launching 30 ads a week is wasting most of them. At 10,000 a month there is simply not enough money to show 30 ads to enough people to learn anything. You will get a read on a handful and the rest will sit unspent.
• A large account launching 5 ads a week is starving itself. At 100,000 a month, five new ideas will not replace winners as fast as they tire, and Meta will keep pushing the same few ads to the same people until results slide.

On the four accounts we run the most volume on, we launched between 6 and 9 new ads in each week we shipped a batch over the last 90 days. That sits inside the table for the budgets involved, and it is deliberately not more.

How many ads Meta will actually spend on

This is the part every guide skips. You can launch twenty ads, but Meta decides which ones get the money, and it decides fast.

Across the four accounts where we launched 25 or more ads in the last 90 days, the top 10 percent of ads took between 62 and 72 percent of all the spend. The top three ads on each account took between 56 and 62 percent on their own. On a UK discount beauty and home retailer, a new ad picked up about a third of the account's spend within its first four days live. That is Meta doing its job: it finds an ad that works and backs it.

The flip side is that most ads barely get shown. On a kids' supplement brand, three new statics took almost nothing in two weeks because they sat in ad sets where one video was taking 89 to 99 percent of the spend. They were never tested. They were starved.

So when you count how many ads you are running, count the ones getting meaningful spend, not the ones you uploaded. Three practical rules:

1. Keep 3 to 6 ads in each ad set. More than that and the extras rarely get a look in.
2. Give a different format its own ad set. If your videos dominate, statics placed next to them will not get a fair read.
3. Treat an unspent ad as untested, not as a loser. Relaunch the ideas you believe in next to a different group of ads.

There is also a hard ceiling on how many ads a Page can run at once, tied to how much it spends. For Pages spending under 100,000 dollars in their highest month the limit has been 250 live ads, though Meta has been relaxing that cap for some accounts. Almost nobody should be near it. If you are, you are running too many ads that do nothing.

How many winners to expect

Founders often ask how many ads to run because they want to know how many it takes to find a winner. Our honest answer: more than you would like.

Of the 178 ads we launched across our accounts in the last 90 days, 31 earned a verdict of winner or scale. That is about one in six. More than half, 92 ads, were still marked too early to call, because they had not yet spent enough to judge. Only 14 had clearly lost.

That ratio is why volume matters. If roughly one ad in six turns into something you can scale, and you need a new winner every few weeks to replace the ones that tire, you need a steady flow of new ideas behind it. A brand that launches four ads a month will find a winner every month or two. A brand that launches eight a week will find one most weeks.

It is also why the definition of a winner matters. A winner is rare and it is relative to your own target. An ad that beats a weak account average can still be losing you money.

How long ads last, and when to replace them

The other half of "how many ads" is how long each one keeps working. The middle ad on our accounts stayed live for between 10 and 32 days, depending on the account. Strong ones last much longer. On the kids' supplement brand, one video has been live for 46 days and is still the account's top spender with the best cost per purchase among ads with enough data to read. We brief new work against it rather than switching it off to make room.

Fatigue is real, but it is not a reason to panic. On a UK luxury laundry brand, two ads were flagged as tiring, on rising frequency or falling click rates, yet both were still bringing in purchases under target. The right move there is to replace them with fresh ideas, not to pause them before the replacements have proved themselves. We wrote about what tends to sit behind a sudden drop in our piece on why Meta ROAS drops.

A simple rhythm that works at most spend levels:

What	How often
Launch a new batch sized to the table above	Every week
Read what has 10 or more results behind it	Every week, same day
Retire ads that have spent three times target with nothing	As soon as they hit it
Replace tiring winners	Before they drop under target, not after
Keep long runners live	As long as they beat target

What this looks like when it works

Volume on its own does not grow an account. Volume sized to the budget, with different ideas, read honestly every week, does. Two results from our case studies page show what happens when the flow of new creative keeps up with the spend. Liquidation Store: "Scaled Meta spend nearly 4x in a month while ROAS rose from 4.2x to 5.7x, peaking at about 5,000 orders a month from Meta." Le Luxe: "Orders from Meta up 3.4x in two months, with ROAS up 24% at the same time."

Neither of those came from one hero ad. They came from launching new ideas every week, letting Meta back the ones that worked, and replacing them before they tired.

If you want to see what a week of new ads looks like on your own brand, we will make you ads built from your account and your site.

FAQ

How many ads should I run per ad set on Meta? Three to six for most accounts. Meta will usually fund two or three of them properly, so more than six mostly adds ads that never get spend.

How many new ads should I launch each week? Divide your weekly testing budget by three times your target cost per purchase. That is how many separate ideas you can read. Make one or two ads per idea.

Is it bad to run too many ads on Meta? Yes, if your budget cannot fund them. Ads that get no spend teach you nothing, and splitting a small budget across many ads slows every one of them down.

How often should I refresh my Meta ads? Launch new ideas every week and replace winners as they start to tire, but keep any ad live for as long as it beats your target. Some last a week, some last two months.