Ads can bring clicks within a day or two of going live, but it usually takes about three months of steady running to judge them fairly. Google and Meta each describe a short learning period, measured in days or weeks. The rest of the time goes on collecting enough real enquiries and sales to tell a pattern from a lucky week. The 90 days is our rule of thumb rather than a platform rule, and nobody can promise what your ads will do, us included.
How long does Google Ads take to work?
Most Google ads are reviewed within one business day, so you can be live quickly. Some reviews take longer.
If your campaign uses Smart Bidding, which is Google’s automated way of setting a price for each auction, you may see a “Learning” status. Google says it lasts up to seven days and that ads keep running as normal. It appears after setup changes, such as a new bid strategy.
The figure that matters more is the conversion cycle: the time it takes for a click to turn into a conversion, such as an enquiry or a sale. Google says results typically settle within one to two conversion cycles, and suggests judging them over two cycles or two weeks, whichever is longer.
So if your customers usually take three weeks to go from click to enquiry, that’s six weeks before Google’s own guidance says you should judge the results. A takeaway will get there far sooner than a kitchen fitter.
How long do Facebook ads take to work?
Meta, the company behind Facebook and Instagram, says most ads are reviewed within 24 hours, though some take longer.
After that comes the learning phase. Meta says an ad set (a group of ads sharing an audience and ad spend) usually leaves it after about 50 results in the week following its last significant edit. During this phase, Meta says, costs per result are usually higher, performance is less stable, and results aren’t necessarily a guide to what comes next.
Fifty results a week is a lot for a small business. If each enquiry cost £20, as an example, that’s £1,000 a week. So you may see the status “Learning limited”. Meta says that isn’t a penalty. It means the setup isn’t getting enough results to settle, and it suggests options such as combining ad sets, widening the audience or raising ad spend.
Edits matter too. Changing the audience or the ad, adding a new ad, or pausing for seven days or more can send an ad set back into learning. Large changes in ad spend can as well. Meta advises waiting until learning has finished before you edit, unless you’ve a good reason.
What happens month by month?
This is a typical shape, and yours may differ.
| Period | What’s going on | What you should expect | What to decide |
|---|---|---|---|
| Before day one | Tracking and ads are set up | A test enquiry appears in your reports | Is everything being counted? |
| Month one | The platforms and your team are learning | Clicks, some enquiries, uneven costs, some wasted clicks | Are these the right people? |
| Month two | Waste comes out, and what’s working gets more of the ad spend | Fewer odd searches, steadier costs, first patterns | What should we do more of, and what should go? |
| Month three | There’s enough data to judge | A rough cost per enquiry and, ideally, per customer | Carry on, change or stop? |
If your sales take months to close, the third row moves later. A long sale can’t be judged in 90 days on sales alone, so you’d look at the quality of enquiries as well.
Why do early results wobble?
Wobble is normal, and it has several causes.
- Learning. Both platforms say performance can be unstable at the start, while their systems work out what to do.
- Small numbers. Four enquiries one week and nine the next looks like a doubling. It’s more likely to be chance.
- Delay. Google notes that conversions often take several days to come in, so last week usually looks worse than it’ll end up.
- Auctions. You’re bidding against other advertisers, and they change. Meta also says its system goes for lower-cost opportunities first, so costs can rise as those run out.
- Seasons. Your income changes with the time of year, and so does demand for your ads.
- Short time frames. Meta warns that hour-by-hour figures can swing and then average out. It suggests looking at a full week at least.
What are the healthy signs?
In the first 90 days, the direction of travel matters most. These are good signs:
- A test enquiry or sale shows up in your reports.
- The searches or audiences your ads reach look like your customers.
- Enquiries are from real people asking for things you sell, even if there aren’t many.
- Your cost per enquiry is level or falling, month on month.
- Someone can tell you what changed this month and why.
- Your ad spend matches what you agreed.
Cheap clicks alone don’t count. Google says clicks and click prices may move about while its system works towards your main goal, so judge the goal.
When should you stop or change your ads?
Most problems need a change. A few mean you should stop.
| What you see | When | What to do |
|---|---|---|
| Tracking fails, or enquiries aren’t being counted | Any time | Pause if you can’t judge anything, and fix the tracking first |
| Wrong people, wrong searches or wrong services | Month one | Change the search terms, audience or wording |
| A few hundred clicks and no enquiries | Months one to two | Look at the page people land on, your offer and your phone answering |
| Enquiries, but few turn into sales | Months two to three | Check how fast you reply, your prices and who’s enquiring |
| Each customer costs more than they earn you, with no improvement | Around day 90 | Change the approach or stop |
| Cash is tight, or you can’t handle the work | Any time | Reduce or pause. Protect the business first |
The best protection is deciding before you start:
- Write down what success looks like. Pick one main measure, such as the number of enquiries or cost per customer.
- Set your limit. Decide what you can afford to lose. As an example, £1,500 of ad spend a month for three months is £4,500. If losing that would hurt, start smaller.
- Put a review date in your diary. Then leave the settings alone until then, unless something on the list above happens.
Questions people ask
Can I turn my ads off after a month?
Yes, you can pause ads on either platform at any time. A month is rarely a fair test, unless something’s clearly broken or you can’t afford to continue.
Why did results drop after a great first week?
Often the first week was the lucky one. With small numbers, a good start and a dip can both be chance. Look at the whole month before you change anything.
Do Google ads work faster than Facebook ads?
They can, because people on Google are already searching for what you sell. On Facebook and Instagram, you’re reaching people who may not be looking yet, so it can take longer. It varies by business, so treat that as a tendency.
What to do next
Whoever runs your ads, ask for three things: a plain view of what success means, access to your own accounts, and a monthly note of what changed. If you’d like to see how we do it, How it works walks through it. On a monthly plan, we agree what success looks like, share a 30-day plan in your account and send a plain-English report each month. Plans start with three months and then need 30 days’ notice. That matches the period above, so weigh it up for yourself. Prices has the details, and we never charge a percentage of ad spend. If you’ve got ads running now and want a second opinion, there’s the Ads health check.
Related guides
- What an always-on campaign looks like, week by week
- How much does Google Ads cost, and how much should you spend?
- Which part of your marketing funnel isn’t working?
This guide is general information based on publicly available sources, correct when written in November 2026. It isn’t legal, financial or tax advice, and ad platforms change their rules and menus often, so check the latest guidance before you act.