Pre-launch: brand campaigns open in small groups. Join the waitlist
Guide

How to lower your Meta ads cost

Most advice tells you to tweak your targeting. The biggest savings usually come from somewhere else. Nine fixes, easiest first, and the one formula that explains all of them.

In short

Your cost per sale on Meta is your CPM divided by how many of those impressions turn into clicks and then buyers. So there are three ways to lower it: pay less per impression, get more clicks per impression, or turn more clicks into sales. The quickest wins on a small budget are fewer ad sets so they can finish learning, fresh creative before the old one tires, and a better landing page.

On this page

The formula behind every fix

Every fix on this list moves one of three numbers. Here is how they combine:

Cost per sale = CPM ÷ (1,000 × click rate × conversion rate)

Say you pay $14 per 1,000 impressions, 1% of people click, and 2% of those buy. That is $14 ÷ (1,000 × 0.01 × 0.02) = $14 ÷ 0.2 = $70 per sale.

Now improve only the landing page so 4% of clickers buy instead of 2%. Same ads, same audience, same CPM: $35 per sale. Half the cost, and Meta never knew anything changed. Keep that in mind as you read the list.

The 9 fixes at a glance

FixEffortWhat it lowers
1. Use fewer ad setsLowCost per sale, by finishing the learning phase
2. Optimize for an event you can reachLowCost per sale
3. Go broad and let the creative targetLowCPM
4. Exclude people who already boughtLowWasted impressions
5. Refresh creative before it tiresMediumCPM and cost per click
6. Use creator-made videos as adsMediumCost per click
7. Set a cost cap once you know your numbersMediumCost per sale
8. Fix the landing pageMediumCost per sale
9. Move part of the budget to results-priced channelsLowRisk on the whole budget

Fixes you can make today

1. Use fewer ad sets

Meta says an ad set needs about 50 optimization events within 7 days to leave the learning phase (Meta Business Help Center), and results are less stable until it does. Five ad sets at $20 a day each split your data five ways. One ad set at $100 a day learns five times faster. Merge ad sets that target similar people.

2. Optimize for an event you can actually reach

If you get 10 sales a week, an ad set optimizing for purchases will never collect 50. Optimize for an earlier step, such as add to cart or checkout started, until your volume grows. Then check that your cost per sale really falls, because the system will now find people who add to cart, not necessarily people who buy.

3. Go broad and let the creative do the targeting

Stacking narrow interests means competing for a small crowd, which can push your CPM up. A broad audience with a clear, specific ad often lets the system find buyers on its own. Test it against your current setup rather than assuming either way.

4. Exclude people who already bought

Unless you are selling them something new, every impression shown to a recent customer is money spent twice. Upload your customer list or use a purchase audience and exclude it.

Fixes that take a week

5. Refresh creative before it tires

Watch two numbers: frequency (how many times the average person has seen your ad) and click rate. When frequency climbs and clicks fall, the ad is tired and you are paying more for less. New hooks in the first three seconds usually matter more than new offers.

6. Use creator-made videos as ads

Ads that look like normal posts tend to get scrolled past less. Creator videos are the easiest way to get them. If a creator campaign is already running, the posts that pull the most views tell you which angle to test. Ask the creator's permission first: creators own their content, and ad usage rights are agreed separately.

7. Set a cost cap once you know your numbers

Work out the most you can pay per sale and still make a profit. Set that as a cost cap. Delivery may slow down, and that is the point: you stop buying sales that lose you money.

The two fixes outside Meta

8. Fix the landing page

Go back to the formula: doubling your conversion rate halves your cost per sale. Make the page load fast on a phone, repeat the exact promise from the ad in the headline, show the price, and cut every form field you don't need. This is the only fix that also makes every other channel cheaper.

9. Move part of the budget to results-priced channels

You can only tune Meta so far, because you still pay for every impression. Put 10 to 20% of the budget where you pay for results instead: affiliates (per sale) or pay-per-view creator campaigns (per view, up to a target you set, with views past it free). Give it your own UTM link. If its cost per sale beats Meta's, move more.

What not to do

  • Don't edit campaigns every day. Big changes to budget, audience or creative can send an ad set back into learning.
  • Don't judge an ad on day one. Give it at least a few days, or until it has spent two to three times your target cost per sale.
  • Don't chase cheap clicks. A low cost per click from people who never buy is still expensive. Judge everything on cost per sale.

Questions people ask

Why are my Meta ads so expensive?

Usually one of three reasons: your ad sets never leave the learning phase because the budget is split too thin, your creative has tired and people scroll past it, or your landing page loses the clicks you paid for. Season matters too: CPMs usually climb around holiday shopping.

How much should I spend on Meta ads per day?

Enough for each ad set to reach about 50 optimization events a week. If an add to cart costs you $10, that is $500 a week, about $70 a day. If that is too much, optimize for a cheaper event or run fewer ad sets.

Is boosting a post cheaper than running Meta ads?

Boosting is a simplified Meta ad with fewer controls. It is priced the same way, per impression, so it is not cheaper by itself. Ads Manager gives you more ways to lower cost.

What is a good cost per sale on Meta?

One that leaves a profit after product cost and shipping. If you keep $30 on a $60 order, any cost per sale under $30 makes money, and every dollar below that is profit.