How much should a small ecommerce store spend on Meta ads per day?
A daily budget worked out from what a sale can cost you and how many ads you test, not a round number.
By Mark Tran · Reviewed
The short answer
There is no right daily budget for Meta ads, and no minimum below which they stop working. The right number for your store comes from two things: what a sale can cost you, and how many ads you want to test at once. Give each ad enough money to spend about twice your target cost per sale inside a week, and add it up.
For a store where a sale can cost A$40 and you want to test four ads, that is about A$46 a day. For a store where a sale costs A$140, the same test needs about A$160 a day. That is why one founder is told A$20 a day is fine and another is told it is pointless. Both can be right.
Below is the method I use, step by step, with what I found running and auditing Australian ecommerce ad accounts between July and October 2026, with the brands' names removed.
1. Work out what a sale can cost you
Start from your margin, not your ads. Take your average order, remove product cost, packaging, shipping and payment fees, and what is left is the most a sale can cost before you lose money on it. Then decide how much of that you want to keep.
Two things catch Australian stores out. Shopify's sales figures usually include GST, so a sale looks bigger than what you keep. And a discount raises the bar: at 15% off, you need more sales for the same ad spend just to break even.
The break-even ROAS calculator does this maths, with GST and a management fee included.
2. Decide how many ads you will test
More ads means more chances to find a winner, but each one needs its share of the money. Spread a small budget across many ads and none of them gets enough spend to tell you anything.
What I found:
- A jewellery brand ran 30 to 40 creatives in one campaign, each in three copy versions, at A$35 a day. Most got A$2 to A$3 over the whole run. That campaign averaged about A$197 a sale.
- Its next campaign cut back to 7 creatives at the same budget, and cost per sale improved to about A$140. But Meta put roughly 99% of the spend on one video, so the other six never got a fair read. Both campaigns had under ten purchases, so treat the comparison as a direction, not a result.
- A clothing brand's first launch had the same problem: one ad took 72% of the prospecting budget, A$297, for no purchases. A later test gave each of four ads its own ad set, starting at A$8 a day, and spend came out roughly equal across the four.
For a small store I usually test three to five ads at a time, and swap the weakest one or two out after each read.
3. Give each ad enough to reach a verdict
My rule for stopping an ad once purchases are coming in: when it has spent twice the target cost of a sale without a purchase, it goes. That rule sets the floor for your budget, because every ad needs room to reach it inside one read window, usually a week.
Daily budget floor = number of ads × 2 × target cost per sale ÷ 7
| Target cost per sale | Ads tested | Daily budget floor |
|---|---|---|
| A$25 | 3 | about A$21 |
| A$40 | 4 | about A$46 |
| A$70 | 4 | about A$80 |
| A$140 | 4 | about A$160 |
Where a sale costs about A$35, this works out at A$10 a day per ad, which is the rule of thumb I quote on calls. It rises quickly as the cost of a sale goes up.
Meta has its own, much higher bar. Its guidance is that an ad set usually finishes its learning phase after about 50 results in the week after its last significant edit. At A$40 a sale, 50 purchases a week is about A$286 a day for one ad set. Most small stores never get there optimising for purchases, and Meta marks those ad sets learning limited. That is not a reason to stop. It is a reason to judge on earlier signals, as in step 5.
4. Check what that budget buys in clicks
Before you commit, run the numbers the other way: daily budget ÷ cost per click = clicks a day, and clicks × your conversion rate = sales a day. If the answer is a fraction of a sale, the budget can't be judged on purchases.
What I found:
- A clothing brand was paying about A$2.90 per click to its site, which converted 0.96% of visitors. At that rate, clicks alone cost about A$300 per sale. Hitting a A$100 cost per sale needed the traffic to convert at 2.9%, three times the site average. Ten days at A$60 a day was expected to bring 2 to 6 purchases, too few to judge return on.
- A swimwear store's history suggested A$30 a day would buy about 17 landing-page views and one add to cart a day. Its first week, spending a little under A$30 a day, came in at or above that: A$169.53 spent, 139 link clicks, 11 add to carts and no purchases, which the store's own orders confirmed.
Neither of those was a failure. Both were the expected result of the budget, worked out in advance. The problem is when nobody does the maths and a founder switches the ads off after a week of no sales.
5. If you can't afford the floor, narrow the test
A small budget can still learn something. It just has to ask a smaller question.
- Test fewer ads. Two or three instead of six.
- Use one campaign and one ad set. Splitting a small budget across ad sets starves every one of them.
- Optimise for add to cart instead of purchase. On the A$30 a day swimwear account, I estimated purchase optimisation would get 0 to 3 conversions a month, too few for Meta to learn from. Add to cart is still a buying signal, and it builds up much faster.
- Judge on early signals. Click-through rate, cost per click, add-to-cart rate and cost per add to cart, until purchases are high enough to compare.
When you raise the budget
A higher budget doesn't buy sales at the same price. Meta spends the extra on people a little less likely to buy, so cost per sale usually rises. Judge every raise on cost per purchase in the week before and the weeks after.
What I found:
- In a women's fashion account I ran, I raised the budget four times in 12 days, from A$80 to A$132 a day. Cost per purchase rose from A$21 to A$37 and ROAS fell from 4.9x to 3.1x. Each read window had only 13 or 14 purchases, so the numbers are small, but cost per sale climbed with the raises, and the raises came too close together to judge any one of them.
I now make one change per campaign at a time and leave three or four days before reading it, so each change can be judged on its own.
What share of revenue should go on ads?
Treat the percentage as the answer to the maths, not the starting point. Here is a hypothetical store: from every dollar of sales before GST, 50 cents is left after product, packaging, shipping and payment costs. On A$10,000 of sales, that is A$5,000 before marketing. If the owner wants to keep A$2,000 as profit, A$3,000 is left for ads and any management fee.
That works out at 30% of sales for this store. A store with thinner margins gets a much lower number, and a store with strong repeat buying can justify a higher one. A percentage someone quotes you without knowing your margin is a guess.
Want a second pair of eyes?
If you run an Australian ecommerce store and want help setting a budget that can actually answer a question, I do it on a free 15-minute Meta ads audit. Your live campaigns on screen, the one fix to make first. No access needed, no obligation.
Method
Every "What I found" comes from an Australian ecommerce ad account I audited or managed between July and October 2026, in fashion, swimwear and jewellery. Names are removed. Figures are as Meta reported them, in Australian dollars, unless the text says otherwise. The worked examples and the revenue split are hypothetical. The stop rule and ad counts are how I run accounts, not Meta requirements, and these are examples from a small number of accounts, not benchmarks for every store.