How long does it take for Google Ads to work? Every answer you will find is a range, and the ranges contradict each other. One page says twenty four hours, the next says one to three months, a third says a year before you see full potential. They are describing different outcomes, and none of them is describing your account, which is the only one that matters to you.
A range is close to useless anyway. It tells you to wait without telling you what you are waiting for, so you discover the money was wasted at the end of the wait rather than in the second week. Replace it with checkpoints. Something should be true at day seven, something else at day thirty, something else again at day ninety, and every missed checkpoint points at a specific cause.
What should be true after the first week?
Data, not results. By day seven the campaign should be spending close to its daily budget, appearing for queries you recognise as your own, and recording conversions that match what actually arrived in your inbox or phone. Nothing about cost per enquiry means anything yet. You are confirming the plumbing works.
Three checks and nothing else. Open the search terms report and read the actual queries, because that is the only honest statement of what you are buying. Compare recorded conversions against real enquiries, one by one, for the week. Check whether the budget is being spent or whether delivery is stuck.
If spend is well under budget, you are almost always below the auction floor for your keywords, or the targeting is so narrow that there is nothing to buy. If the queries look nothing like your business, the match types are doing something you did not intend. If conversions are recorded but no enquiries arrived, the tracking is counting the wrong event and every number after this point is fiction.
What should be true after thirty days?
Enough conversions to compare things. By day thirty a working account has produced a few dozen conversions, a clear split between keywords that convert and keywords that only spend, and a cost per enquiry that has stopped swinging wildly week to week. The number may still be too high. It should no longer be random.
Thirty days is also the earliest point at which judging the account is fair. Automated bidding needs a body of conversion data before it can do anything clever, and until that exists the system is effectively guessing on your behalf. The same principle applies on the other platform, and I have written about why on the Facebook ads learning phase: the mechanics differ, the patience requirement does not.
The common failure at this checkpoint is not a bad cost per lead. It is too little data to have a cost per lead at all, which is nearly always a budget problem rather than a performance problem. An account that generates four conversions a month cannot learn, cannot be optimised and cannot be judged. If that is where you are, the honest options are to raise the budget, narrow the account to one tight campaign so the same money concentrates, or accept that search volume for what you sell is simply small. Working out what that floor looks like is the same arithmetic I use for setting a starting ad budget.
What should be true after ninety days?
A cost per sale you can plan around. By day ninety you should know what a customer costs to acquire through search, which keywords produce them, which ones never will, and whether the number works against your margin. This is the first point at which scaling or stopping is a decision rather than a guess.
Ninety days is also long enough to see the slower effects. Quality Score settles once the account has history, so clicks that cost a fortune in week two often cost noticeably less by month three on identical keywords. Seasonality shows up. Your conversion rate from enquiry to sale becomes visible, which is usually more informative than anything inside the ad account.
For a local business the picture arrives faster, because the query set is smaller and intent is sharper. The structural decisions behind that are in running Google Ads for a local business.
Why are my Google Ads not working yet?
Because one of four things is wrong, and each one announces itself at a different checkpoint. Tracking that counts the wrong event fails week one. Budget below the auction floor fails week four. Query mismatch shows up in the search terms report immediately. A landing page that does not convert shows up only once traffic is genuinely relevant.
Read the misses in order, because fixing them out of order wastes money:
- Tracking. If recorded conversions do not match real enquiries, stop and fix this before touching anything else. Every optimisation decision downstream depends on it.
- Budget. Too few conversions to learn from is not poor performance, it is an unreadable account. Concentrate the spend before you judge it.
- Query match. If the search terms report is full of queries you would never bid on, the account is buying the wrong traffic and no bid adjustment fixes that.
- The landing page. Relevant clicks that do not convert are a page problem, not an ads problem. This is also where most accounts actually lose, long after the ads stopped being the issue.
There is a fifth cause that sits outside the account entirely, and it is the one people reach for last. The ads can be working while nothing happens, because enquiries arrive and nobody calls them back quickly or follows up more than once. That is a different problem with a different fix, and I have written about why leads do not convert separately. If the ads are producing enquiries and the pipeline behind them is the weak link, you can build that follow up system yourself over about thirty days rather than spending more on traffic.
How do you tell slow apart from broken?
Slow improves in a direction. Broken repeats. If cost per conversion is falling month over month, even unevenly, the account is learning and the right action is to leave it alone. If the same number sits flat across sixty days of spend, waiting longer is not patience, it is a decision to keep paying for the same result.
The test I use is whether anything about the account is different from thirty days ago. Fewer wasted queries, better Quality Score, more conversions for the same spend, a tighter keyword list. If none of those moved, nobody is managing the account and time will not fix that.
When should you stop?
When you have given it ninety days, fixed the four checkpoint failures, and the cost per sale still does not fit your margin. That is a genuine answer rather than a failure, and it is worth far more than another three months of hoping. Some businesses are simply not bought through search.
Before you conclude that, check one thing: whether anybody was ever searching for what you sell. Google Ads capture existing demand, so if the query volume is not there, no amount of time or budget creates it. That is not an account you need to optimise. It is a channel that was never the right one, and the money belongs somewhere people can be shown the offer rather than asked to search for it.
Frequently Asked Questions
How long does it take for Google Ads to start working?
Clicks begin within hours of approval, usually the same day. Useful data takes about thirty days, and a cost per sale you can plan around takes about ninety. Anything you measure in the first week is a plumbing check rather than a performance verdict, so treat early numbers as confirmation that the setup works.
Why are my Google Ads getting clicks but no conversions?
Three usual causes, in order of likelihood. The tracking is counting the wrong action, so the conversions exist but are not being recorded. The search terms do not match what you sell, so the clicks were never buyers. Or the landing page asks for too much. Check the search terms report first, because it is the fastest to read.
How much should I spend before judging Google Ads?
Enough to produce roughly thirty conversions, because that is the point where differences between keywords stop being noise. Work backwards from your expected cost per conversion rather than picking a round monthly figure. If that total is more than you can commit, narrow the account to one campaign so the same money concentrates.
Do Google Ads get cheaper over time?
Often, though not automatically. Accounts with history tend to earn better Quality Scores on the same keywords, which lowers what you pay per click, and better conversion data lets automated bidding place smarter bids. The improvement comes from accumulated relevance and data, so an account nobody manages will not drift cheaper on its own.
Is three months long enough to judge Google Ads?
Yes, provided the account was actually managed during it. Ninety days of monitored spend with tracking verified, wasted queries removed and the landing page tested is a real verdict. Ninety days of an account nobody opened is not three months of evidence, it is one week of evidence repeated twelve times.



