Most Facebook ads that "didn't work" were judged in the first week. That's when the numbers are smallest, the reporting is least complete and Meta's delivery system is still learning. A fair answer has three parts: ads start delivering within a day, lead numbers settle over a few weeks, and profit takes a couple of months to read.
- Day 1: ad review, then delivery. Meta says to allow up to 24 hours.
- Week 1: learning. It usually ends after about 50 results in a week.
- Weeks 2-6: first leads and a cost per lead you can start to trust.
- Around day 60: enough data to decide whether to scale.
- Don't edit daily. Every significant edit restarts learning.
The timeline, week by week
| When | What to judge | What's happening |
|---|---|---|
| First 24 hours | Only whether it's delivering | Ad review, then delivery starts |
| Week 1 | Nothing much. Leave it alone | Learning: Meta is finding who gives you results |
| Days 10-14 | Whether leads arrive and tracking records them | First leads, for a new account with tracking to set up |
| Around day 45 | Cost per lead against what you can afford | A cost per lead you can rely on |
| Around day 60 | Whether to raise the budget | Enough results to compare ads and audiences |
| Months 3-10 | Cost per customer | Creative testing compounds |
Delivery and learning: Meta Business Help Center, read October 2026. Days 10-60: the timeline on our Facebook ads service page.
Our own service page puts it plainly: "First leads typically land in 10-14 days, with reliable cost-per-lead by around day 45 and scaling decisions by day 60." That's for a new account where we set up tracking and creative first. An account that already tracks results well can move faster.
Why the first week misleads
Meta is still learning
Meta's delivery system tests who's likely to give you the result you asked for. Learning usually occurs after about 50 results in the week after the ad set's last significant edit. Until then, costs swing and results look random, because they partly are.
The report isn't finished
Meta can credit a sale within 1-day or 7-day after a link click on your ad. Someone who clicks today and buys on Saturday shows up later. The latest days always look worse than they'll end up. Why ROAS drops when nothing changed explains the lag.
Small numbers lie
Five leads one week and two the next isn't a trend. It's noise. Judge on at least a few dozen results before you compare ads.
What makes it take longer
- Constant edits. Meta counts changes to targeting, creative, the optimization event and more as significant edits, and each restarts learning.
- A budget too small for the result. Meta's delivery guide suggests a daily budget of at least 10 times the average cost of your optimization event. How much to spend on Facebook ads has a calculator.
- Too many ad sets. Four small ad sets learn four times slower than one. If Ads Manager shows Learning limited, here's how to fix it.
- A rare result. Optimizing for purchases on a store with few sales takes longer than optimizing for leads.
- A long buying decision. A $5,000 kitchen or a high-ticket program takes weeks from first click to sale. Judge leads early and sales later.
How long one account actually took
One climbing gym we manage is a fair example of the curve. In its first phase, from November 2025 to January 2026, the account cost £28.82 per booking. From July 2026 it was £5.13, on a similar budget. Ten months, 118 ads and 34 campaigns sit between those two numbers. Fixing tracking and the campaign objective did a lot of it, and so did testing creative month after month. None of it showed in week one.
When not to wait
Patience is for ads that are delivering and tracking properly. Some problems should be fixed straight away:
- Nothing delivers after 48 hours. Meta says to wait up to 24 hours to see delivery, and to delete and recreate the ad if it still won't run after 48.
- The ad is rejected. Read the reason, fix it, and resubmit.
- Tracking records nothing. If people book and Ads Manager shows zero, every number after this is wrong.
- Spend with no leads at all. Two to three times your target cost per lead with nothing to show usually means the form or page is broken. Nine checks for ads that aren't getting leads and clicks but no sales cover the diagnosis.
How to give it a fair test
- One campaign and one or two ad sets, each with a budget that can reach the result.
- Three to five ads, so Meta has something to choose between.
- No edits for seven days. Batch changes into one weekly session.
- A decision written down before you start: what cost per lead at day 45 means carry on, and what means stop.
- A proper test when you compare two approaches. Meta's A/B tests ensure nobody sees both versions, which turning ads off and on by hand can't.
After that, a steady creative testing cadence is what moves cost per customer down over months. It's also how we run Facebook ads for local businesses.
Questions people ask
How long do Facebook ads take to work?
Ads usually start delivering within 24 hours, and the learning phase ends after about 50 results in a week. For a new account with tracking and creative to set up, we see first leads in 10-14 days, a reliable cost per lead by around day 45 and scaling decisions by day 60. Profit takes longer to judge than leads.
How long should I run a Facebook ad before changing it?
At least a week, unless something is clearly broken. Every significant edit sends the ad set back into learning, so changing things every few days means it never finishes. Batch your changes and make them once a week.
Why are my Facebook ads not working after a week?
Often they are, and the report hasn't caught up: sales can be credited up to 7 days after a click. If there are clicks but no leads at all, the problem is usually the form, the page or the tracking rather than the ads.
How long is the Facebook learning phase?
Meta says learning usually ends after about 50 results in the week after the ad set's last significant edit. On a small budget that can take longer than a week, and Ads Manager labels the ad set Learning limited.
When should I stop a Facebook ad early?
If it isn't delivering after 48 hours, if it's rejected, if tracking isn't recording results, or if it has spent two to three times your target cost per lead with no lead at all. Fix the cause before you judge the ad.
When is it worth getting help?
When you've given the ads a fair test and still can't tell whether they work, or when every week brings a new change and no clear result. We're not the right fit if you can't commit a budget for at least two months.
Not sure whether your ads have had a fair test? Book a strategy call and we'll look at the account, the tracking and the timeline with you.
