How Long Does It Take for Paid Ads to Work?
The short answer: paid ads typically show first meaningful signals in 2–4 weeks, produce numbers you can trust by 60 days, and deserve a fair verdict at 90 days. Ads that are profitable in week one happen, but they're the exception. The most expensive mistake in small-business advertising isn't bad targeting — it's judging (and re-judging, and restarting) too early.
Nobody says this part out loud often enough, so we will: the first month of a brand-new ad account usually looks mediocre. That's not failure. That's the machine calibrating. Here's what each month is actually doing.
What happens in month one?
Month one is data collection wearing the costume of underperformance.
- Weeks 1–2: the platform's learning phase. Meta and Google are testing your ads against different slices of people, and costs swing wildly day to day. A $9 cost-per-result Monday and a $40 Tuesday is normal, not a crisis.
- Weeks 3–4: patterns emerge. One ad starts pulling ahead. Costs stop swinging quite so hard. You now know things you couldn't have known before: which message resonates, roughly what a click and a conversion cost for you — not for some industry average.
What "working" looks like in month one is not profit. It's leading indicators: click-through rates above ~1%, people reaching your product pages, add-to-carts, inquiries. Those tell you the engine is catching.
What happens in months two and three?
Month two is where optimization becomes honest. There's enough data to kill underperformers with confidence, feed winners, and refine audiences based on who actually bought (not who you guessed would). Costs usually drift down as waste gets trimmed.
Month three is verdict territory. By day 90 you have real answers: what a customer costs, which creative style wins, whether the unit economics work. From here, decisions get boring in the best way — scale what works, raise the budget deliberately, rotate creative before it fatigues.
| Timeline | What's happening | Fair expectation |
|---|---|---|
| Weeks 1–2 | Learning phase, costs swing | Ignore daily numbers entirely |
| Weeks 3–4 | Patterns emerge | Leading indicators, first sales |
| Month 2 | Real optimization begins | Costs stabilize and improve |
| Month 3 | Enough data for a verdict | Judge honestly, then scale or fix |
What resets the clock?
The learning phase restarts — fully or partially — every time you make a significant change. The most common self-sabotage we see in accounts founders ran themselves:
- Editing campaigns daily. Every big edit tells the platform to start re-learning. The account that gets touched every morning stays in week one forever.
- Pausing ads after a bad weekend. Two slow days is weather, not climate.
- Restarting from scratch. Turning everything off in frustration and relaunching a month later throws away every dollar of learning you already bought.
The discipline that separates accounts that mature from accounts that thrash: check daily if you must, judge weekly, decide monthly.
When is it actually not working?
Patience isn't a blank check, and 90 days of hope isn't a strategy either. Worry — legitimately — when:
- 60+ days in, there are no leading indicators. Decent budget, real impressions, but clicks and engagement stay flat: usually a message or offer problem, not a time problem.
- Clicks are plentiful but nobody buys, for weeks. The ads are fine; the landing page or offer needs the work.
- The math can't work even in theory. If your cost per sale would need to be a third of what the platform charges in your category, more patience won't fix arithmetic.
The 90-day verdict should be honest in both directions. Sometimes it says "scale." Sometimes it says "fix the website first." Both are wins compared to guessing — and a fair test is the only way to get either answer.
Frequently asked questions
How long does it take for Facebook or Google ads to start working?
Expect first meaningful signals in 2–4 weeks, trustworthy numbers by 60 days, and a fair verdict at 90 days. The first two weeks are a learning phase where costs swing widely and daily results mean very little.
Why are my new ads performing so badly?
New campaigns start in a learning phase where the platform tests your ads against different audiences, so early costs are high and erratic. Most accounts look mediocre in month one. Judge leading indicators (clicks, add-to-carts, inquiries) first, and hold profit judgments until 60–90 days.
Should I turn off ads that aren't working after a week?
Almost never. One week isn't enough data, and pausing or heavily editing campaigns restarts platform learning, which resets your progress. Check daily if you like, judge weekly, and only make structural decisions monthly.
When should I give up on paid ads?
If after 60+ days of properly funded testing there are no leading indicators at all, or clicks consistently arrive but never convert, the problem is usually the offer, message, or website rather than time. Fix that first — more ad spend won't out-run it.
Founder & CEO of Fiddle Leaf Marketing. A decade in performance marketing, now helping women-led brands grow with thoughtful paid ads.
