What Instagram and
Facebook ads cost.

You set a budget, not a price. A daily budget is an average: Ads Manager says Meta may spend up to 75% more than the daily budget on days with better opportunities (older developer documentation still says 25%), but over a week it won't spend more than 7 times the daily budget. A lifetime budget is a ceiling for the ad set: Meta says you won't be charged more than it unless you change the delivery settings.

Each ad set's settings show what you're charged for. For most campaigns it's impressions, counted per 1,000 views (CPM), whatever result you've asked Meta to find.

If you need a hard limit across everything, set an account spending limit in your billing settings as well as your budgets.

Yes, but not one number. Meta's documentation says minimum daily budgets depend on your account's currency and on what the ad set is optimised for: impressions have one floor, frequent actions like clicks another, and rarer actions like app installs another again. Ads Manager tells you the minimum when you enter a budget.

So treat any 'from a dollar a day' figure as someone else's account. The minimum Meta will accept and the budget that gets you useful results are very different numbers.

Mostly the wrong question. Both run through the same Ads Manager and the same auction, and by default Meta spreads an ad set across its placements (feeds, Stories, Reels and more) to find results at the lowest cost.

You can limit an ad set to Instagram only. That's worth doing when the creative only makes sense there. Usually it narrows the places your ad can show, which tends to push costs up rather than down.

Compare placements after a campaign has run, using the placement breakdown in Ads Manager, and judge them on cost per result, not cost per view.

Cost per result is three numbers multiplied together. The cost per 1,000 views (CPM) is set by the auction. The click rate (CTR) is set mostly by your ad. The share of clicks that act is set by your page and your offer.

That's useful, because it tells you where to work. If the click rate is low, a cheaper audience won't save you: the ad needs to change. If people click but don't act, the ad's fine and the page is the problem. Double either rate and your cost per result halves.

Put your own numbers in from Ads Manager. The starting numbers are made up to show how it works.

The highest bid doesn't automatically win. As Jon Loomer explains Meta's auction, a lower bid can beat a higher one when Meta predicts more people will respond to it, or rates the ad's quality higher. Clickbait, engagement bait and ads that hold back key information score worse.

Competition matters too. Costs usually rise in the weeks when everyone advertises at once, like the lead-up to Black Friday, Christmas and the end of the financial year, because more advertisers are bidding for the same people.

And small audiences cost more to reach repeatedly. The narrower you target, the fewer people Meta can choose from to find the ones likely to act.

Meta says an ad set needs about 50 results in a week to finish its learning phase, where delivery settles down. If your budget buys 10 results a week, splitting it across five ad sets means none of them gets there.

So the useful budget is set by your cost per result, not by a minimum. The calculator above shows results per week for your budget. The learning budget calculator on the social media ads page works it the other way, from the results you need.

When the number is out of reach, use fewer ad sets, or optimise for an earlier step that happens more often, like a lead form opened rather than a sale.

Meta bills from overseas, so the tax side works differently from a local supplier. The ATO's page on GST on imported services and digital products explains the rules. Show your accountant a Meta invoice once, so you know how to treat it.

Keep the ad spend separate from anyone's management fee in your books. It's the only way to see what the ads cost per result over time.