Google Ads starts showing ads within hours, but it takes about 90 days to know whether it works for your business. The first two to four weeks are the learning period, where Smart Bidding gathers data and costs are unstable. Weeks five to eight produce data worth acting on. Weeks nine to twelve tell you whether the changes held.
Anyone who promises results in the first fortnight is either selling something or has not run enough accounts to know better.
That said, 90 days is an average, and the real answer depends on two things about your business: how many conversions you generate per month, and how long your sales cycle is.
What happens in the first 30 days?
Very little that is worth judging. This is the learning period, and Google is openly experimenting with your money.
Smart Bidding works by predicting the value of each click from past conversions. A new campaign has no past, so it explores. It tries different searches, different times of day, different audiences, and watches what converts. That exploration costs money and produces noisy results.
Google usually shows a learning status for one to two weeks after a campaign launch or a significant change. Treat that as the minimum, not the answer.
What to do in month one: leave it alone. The single most common mistake is changing bids, budgets or targets in week two because the numbers look bad. Every change restarts the learning, so an account that gets adjusted weekly can stay in a permanent learning state for months.
The exceptions worth acting on immediately are genuine errors: a broken landing page, ads showing for obviously wrong searches, or conversion tracking that is not firing.
Why does conversion volume change the timeline?
Because the learning period is measured in conversions, not in days. Google publishes volume thresholds for its bidding strategies, and they tell you what the model needs.
For Target CPA, Google recommends evaluating performance over the last 30 days including at least 30 conversions. For Target ROAS on Search and Shopping, Google states you need at least 15 conversions in the past 30 days.
| Conversions per month | Realistic time to a reliable read | What to expect |
|---|---|---|
| 50 or more | About 60 days | Smart Bidding stabilizes quickly, decisions are well supported |
| 20 to 50 | About 90 days | The normal case. Monthly reviews, quarterly judgement |
| 5 to 20 | 120 to 180 days | Slow. Consider simpler bidding and fewer campaigns |
| Under 5 | Hard to read at all | Not enough signal. Fix volume or reconsider the channel |
What this means for you: if your account produces eight conversions a month, judging it at 30 days is meaningless. You will be looking at eight data points. Our post on bidding strategies covers which strategy suits which volume level.
How does your sales cycle affect it?
Add your sales cycle on top of the learning period. If it takes your customers six weeks to decide, then a click today does not become a sale until mid November, and Google Ads reports that conversion on the date of the click, not the date of the sale.
That creates a reporting lag that catches people out. Your last 30 days will always look worse than reality, because some of those clicks have not finished converting yet. Look at a period that ended at least one sales cycle ago if you want an accurate read.
For a plumber, the cycle is minutes and this barely matters. For a B2B software company with a three month cycle, the first meaningful read is six months out, and the account needs offline conversion tracking to see the sales at all.
What should you expect month by month?
A rough shape for a typical lead generation account.
- Weeks 1 to 2. Ads go live. Costs are high and inconsistent. Search terms report is full of things you did not want. Build the negative list.
- Weeks 3 to 4. Learning settles. First conversions arrive. Cost per lead is probably above target and that is normal.
- Weeks 5 to 8. Enough data to act on. Prune keywords, fix the landing page, adjust targets in small steps.
- Weeks 9 to 12. Performance should be near or at target. If it is nowhere close, the problem is usually the offer, the page, or the economics, not the account.
- Month 4 onward. Incremental improvement, expansion into new keywords or campaigns, and seasonal adjustments.
The uncomfortable version: if the account is still far from target at 90 days with reasonable volume, more time will not fix it. Something outside Google Ads is wrong.
What makes it take longer than it should?
Five things, and four of them are self inflicted.
- Constant changes. Every meaningful edit restarts learning. Batch changes and leave gaps between them.
- Too many campaigns. Conversion data pools at campaign level. Six thin campaigns learn slower than two healthy ones. See account structure.
- Broken or vague conversion tracking. If Google is learning from the wrong signal, it learns the wrong thing quickly. See conversion tracking.
- A budget too small for the industry. At an average cost per click of $5.42 across all industries, a $200 a month budget buys under 40 clicks. That is not enough to learn from.
- Genuinely low search demand. Sometimes nobody is searching for what you sell. No amount of patience creates demand.
When should you pull the plug?
When the economics do not work and you have proven it, rather than when you are frustrated.
The test is straightforward. Take your actual cost per lead, divide by your close rate to get cost per customer, and compare it to your profit per customer. If cost per customer is higher than profit per customer after 90 days of properly managed advertising, the channel does not fit your business at current margins.
The all industry average cost per lead was $66.69, with a confirmed range across industries from about $26.84 to $131.63. Our posts on cost per lead benchmarks and what Google Ads cost have the full figures.
Before you quit, check the two things that are not the account: whether the leads are being followed up quickly, and whether the landing page matches the ad. Both routinely account for more of the gap than bidding does.
Common questions
How long until my ads start showing?
Usually within a few hours. New ads go through a review process, and most are approved within one business day. If an ad is disapproved, our post on fixing disapproved ads covers what to do.
Is the learning period the same for every campaign type?
The principle is the same but the volume thresholds differ. Google publishes a higher bar for Demand Gen, at least 50 conversions in the past 35 days for Target ROAS, than for Search at 15.
Can I speed it up?
Somewhat. Fewer campaigns, accurate conversion tracking, a sensible budget and leaving the account alone all shorten the timeline. Nothing shortens it below the number of conversions the model needs.
Does pausing over the weekend help?
No. It slows learning and it removes days of data. If certain hours genuinely do not convert, use an ad schedule rather than pausing manually.
Why did performance drop after two good months?
Common causes are seasonality, a new competitor in the auction, a change someone made, or a landing page that broke. Check the auction insights report, though note it does not display at all when your impression share is under 10%.
Should I switch agencies if month one is bad?
Month one is always bad. If month three is bad and nobody can explain why in plain language, that is the signal. Our post on choosing a Google Ads agency covers what good communication looks like.
Where to go from here
Set the expectation before you start, not after. Budget for 90 days, agree what success looks like in numbers, and resist judging anything in the first month. That single decision does more for results than most optimizations.
If you are partway through a launch and unsure whether what you are seeing is normal, our Google Ads management page explains how we structure the first quarter, and you can get in touch for a straight read on where your account is. Our case studies show the kind of timelines real accounts actually follow.
