AIGENCYFLOW

How much should an e-commerce brand spend on Meta ads?

Set your Meta ads budget from your margin, not from a round number. First decide the most you can pay in ads for one order (your target cost per order). Then fund enough orders a week for Meta's system to learn, which Meta puts at around 50 conversions per ad set per week. With an €18 target that is about €900 a week for one campaign. Most brands we talk to pick a budget by feel ("let's try €1,500 a month") and then judge the result by feel too. Below is the method we use for every e-commerce account, so the number makes sense before the first euro is spent.

In short

  • Target cost per order = about one third of your contribution margin per order. Manage against this one number.
  • Learning budget = 50 × target cost per order, per week, per campaign.
  • Three stages: learn (weeks 1 to 3), stabilise (weeks 4 to 8), scale (week 9 onwards).
  • Scale in steps of 20%, at most twice a week, and only while cost per order is under target.
  • If cost per order drifts up, change the ads before you change the budget.

Start from the order, not from the budget

Take one product, or your average order, and write down three numbers:

  • Average order value: what a customer pays, including the shipping they pay.
  • Contribution margin: what is left after product cost, payment fees, shipping and packaging. For most DTC brands this sits between 50% and 70% of the order value.
  • Target cost per order: the maximum you are willing to pay in ads to win that order. A safe starting point is one third of the contribution margin. If you sell a €90 set with €55 margin, a target cost per order of around €18 leaves room for returns and for the days when ads simply perform worse.

The target cost per order is the one number you manage against. ROAS, CPM and click-through rate help explain why something happens, but only cost per order tells you whether it is working.

You can afford more if customers reorder. A brand whose customers buy three times a year can set the target on the first order closer to half the margin, because the second and third orders cost almost nothing in ads. Be honest about the reorder rate and use what your store data shows rather than what you hope.

Stage one: learning (weeks 1 to 3)

Meta's delivery system needs roughly 50 conversions per ad set per week to leave what it calls the learning phase and optimise steadily. Multiply that by your target cost per order and you have the minimum budget to learn anything. With an €18 target, that is about €900 per week, or €130 a day, for one campaign.

You can start lower, in two ways.

  • Accept slower learning. With half the budget the numbers swing more and it takes twice as long to read them. In our experience, brands that start below half the learning budget tend to quit after a month. They never reached a volume where the results are readable, which is different from ads that do not work.
  • Optimise for an earlier step for a short time. A store with very few orders can let Meta optimise for "add to cart" or "initiate checkout" first, because those happen more often, and switch to purchases once volume allows. Treat this as a temporary bridge. The system will find people who add to cart, and that is a different group from people who pay.

Keep the structure simple at this stage. One campaign with one or two ad sets, broad targeting and several different ads teaches the system more than six small ad sets that each get five orders a week.

Stage two: stabilise (weeks 4 to 8)

Once cost per order has been under target for two consecutive weeks, keep the budget flat and put your attention into creatives. A new batch of ads now moves the result far more than any budget change. Fatigue, far more than the auction, is the reason most accounts drift upward in cost per order, which is why we produce a weekly batch for every client. How large that batch should be is a calculation too, explained in how many new creatives a week you need.

Stage three: scale (week 9 onwards)

Raise the daily budget 20% at a time, at most twice a week, and only while cost per order stays under target. A larger jump can push the ad set back into learning, and that usually costs you a week. If cost per order rises above target for five days in a row, hold the budget and swap creatives before you cut spend.

Scaling also changes who you reach. Your first buyers are the easiest ones, and each step up reaches people a little less ready to buy, so a small rise in cost per order as you scale is normal. You need to know whether the extra orders are still profitable at the new cost, which is why the target comes from your margin.

What a healthy account looks like

  • Cost per order under target for the trailing 14 days.
  • At least 30% of spend on ads launched in the last 30 days.
  • A checkout with express payment and no account requirement, so the orders you pay for actually complete. The nine settings we check are in Shopify checkout optimisation.
  • Pixel and Conversions API both active, so Meta sees every purchase.

If any of these is missing, more budget will not fix it. Close the leak first and scale afterwards.

A worked example

Illustrative figures, not a client result.

A bedding brand sells sets at €189 with a €110 contribution margin. Its target cost per order is €35, so the learning budget is 50 × €35 = €1,750 per week. After four weeks the account runs at €28 per order. From week nine the budget grows 20% at a time, and by month four the brand spends around €12,000 a month at €31 per order. That is roughly 390 orders and around €73,000 of monthly revenue from Meta alone. The brand got there without a bigger budget on day one. It needed a target, patience in the first month and fresh creatives every week.

Common mistakes

  • Judging after seven days. The first week is the noisiest. Read the trend over two to three weeks.
  • Splitting a small budget over many ad sets. Each one starves. Use fewer ad sets and more ads.
  • Changing budget, audience and ads on the same day. You will not know which change did what, and each large edit can restart learning.
  • Counting only the platform's numbers. Compare Meta's reported purchases with the orders in your store every week.
  • Scaling on a discount week. Sale periods flatter cost per order. Set budgets from normal weeks.

Setting this budget and managing it week by week is the core of our paid advertising work.

Frequently asked questions

What is a good starting budget for Meta ads for an online store?

There is no universal figure. Multiply your target cost per order by 50 to get a weekly learning budget for one campaign. A store with a €20 target needs about €1,000 a week to learn at full speed, and less still works, only more slowly.

How do I calculate my target cost per order?

Take your contribution margin per order, after product cost, shipping, payment fees and packaging, and start with one third of it. Raise the share if customers reliably reorder and lower it if returns are high.

How long does the Meta learning phase take?

Meta's own guidance is that an ad set leaves the learning phase after around 50 optimisation events in a week. Plan on about three weeks before results are stable enough to judge.

How fast can I increase my Meta ads budget?

In steps of about 20%, at most twice a week, while cost per order stays under target. Larger jumps can restart the learning phase and make results unstable for several days.

Should I look at ROAS or cost per order?

Use cost per order against a target that comes from your margin. ROAS is useful for comparing campaigns, but a high ROAS on a low-margin product can still lose money.

What should I do when cost per order goes up?

Check tracking first, then the store, then the age of your ads. In most accounts the cause is creative fatigue, so launch new ads before you cut the budget.

Sources

Let’s talk about your project.

Book 30 minutes with the founder. We prepare the call by looking at your ads and your website beforehand.

Yousrireplies in under 30 min

  1. We look at your numbers
  2. We tell you what we’d change
  3. Quote within 48h
  • 30-minute video call
  • No commitment

Prefer to write? Message on WhatsApp

Do you run a gym?Free audit with the founder

Live calendar30-min video call, email confirmation