When Should You Start Running Meta Ads?

Meta ads are usually the first paid channel an ecommerce store tries, and often the first place it loses meaningful money.

Not because the platform doesn't work — it does — but because most stores start before the conditions that make it work are in place.

Here's what needs to be true first.

1. Your margin has to support it

The gating condition, and it's arithmetic rather than opinion.

Your break-even ROAS is 1 ÷ contribution margin %. At 25% contribution margin you need 4x just to break even. Sustaining 4x on cold traffic is genuinely difficult.

If your contribution margin is under about 30%, fix that before you advertise. Raise prices, reduce landed cost, or charge properly for shipping. No amount of creative testing rescues a model that needs 5x ROAS.

This is the step almost everyone skips, and it's the one that decides the outcome.

2. Your product has to be proven

Meta amplifies whatever is already happening. If your product converts poorly organically, paid traffic converts worse — it's colder.

Before spending, you want:

Some organic or referral sales — evidence people buy this without being interrupted
A conversion rate you know. Under ~1.5% and paid traffic will struggle
Some reviews. Social proof matters more for cold traffic than warm

If you've had no sales at all, ads are an expensive way to test whether people want your product. Friends, markets, communities and organic social answer that question for free.

3. Your tracking has to work

Meta optimises toward whatever you tell it to. Broken tracking means it optimises toward nothing.

Minimum:

Pixel installed and firing — verify with Meta's Pixel Helper, don't assume
Purchase events working, with values attached
Conversions API ideally, since browser tracking loses an estimated 30–50% of events to Safari, iOS and ad blockers

Running ads with broken tracking wastes the budget and the learning. You end up with no result and no explanation.

4. You need enough budget for it to learn

Meta's delivery system needs a volume of conversion events before it optimises well — commonly cited as around 50 per week per ad set.

If your budget produces five conversions a week, the algorithm never leaves its learning phase. Delivery stays erratic, costs stay high, and results are too noisy to interpret.

The practical implication: a budget too small doesn't produce a small result, it produces an unreadable one. You spend the money and learn nothing, which is the worst of both outcomes.

Better to run a properly-funded test for two weeks than a starved one for three months.

So when is the right time?

When all four are true:

Contribution margin above ~30%
Product proven with organic sales and a known conversion rate
Pixel and purchase events verified working
Budget sufficient for meaningful weekly conversion volume

Miss one and you're likely to conclude "Meta ads don't work for us" when what actually happened is that the conditions weren't met.

What does a sensible first test look like?

Run it for at least two weeks. Meta needs time, and results in the first few days are noise.

Change nothing while it runs. Every meaningful edit resets the learning phase. This is genuinely hard to resist and genuinely important.

Have more than one creative. Meta's current system rewards genuinely different concepts — different hook, format and visual treatment — not the same image with different text. Near-identical variations get grouped and compete with each other rather than expanding your reach.

Judge it on blended ROAS, not Meta's number. Meta will report a figure flattering to Meta. See blended ROAS and MER.

Decide your stopping rule before you start. Write down the cost per purchase at which you'll turn it off. Deciding that in advance, while you're calm, is worth more than any targeting setting.

What if I can't afford a proper test?

Then don't run a small one. Genuinely.

Better uses of a limited budget, roughly in order:

Fixing margin, so ads become viable later
Organic content and community, which compounds
Email, which is close to free and converts existing interest
Improving conversion rate, which makes every future channel cheaper

A store with 45% contribution margin and a 3% conversion rate can advertise profitably. One with 22% margin and 1.2% conversion cannot, and no budget fixes that ordering.

Knowing whether it's working

The honest answer to "are my Meta ads profitable" needs contribution margin and blended ROAS in the same place — Shopify revenue, product costs, shipping, and spend across every platform.

Meta's dashboard can't tell you this; it doesn't know your costs. That gap is why we built Rev Room for NZ and Australian Shopify stores — blended ROAS across Meta, Google and TikTok against real contribution margin, so "is this working" has an actual answer.

The short answer

Margin first. Under ~30% contribution margin, fix that before advertising
Prove the product organically before paying for traffic
Verify tracking rather than assuming it works
Fund the test properly — a starved budget produces unreadable results, not small ones
Set your stopping rule in advance, and judge on blended ROAS

General information only, current as at 20 August 2026. Platform behaviour and best practice change frequently — verify current guidance before making decisions.