Google Ads Now Says Smart Bidding Learning Takes Around 50 Conversions

Google updated its Smart Bidding documentation to describe the learning period in terms of conversions instead of days. The new wording says learning can take up to around 50 conversions or three conversion cycles. If you run Google Ads on a small budget, this single sentence explains why your campaigns feel like they never settle down.

What Google Actually Changed

Search Engine Land reported on September 22, 2026 that Google reframed how it describes the Smart Bidding learning period in its help documentation. The old framing leaned on elapsed time and conversion cycles. The new framing leads with conversion volume.

The updated wording is that Smart Bidding learning can take up to around 50 conversions or three conversion cycles to calibrate after a change. Google applies this to Search, Shopping, and Performance Max campaigns.

Nothing changed inside your account. No setting moved. This is a documentation change, which is exactly the kind of update that gets ignored and then costs people money six weeks later.

Google is also clear that 50 is an approximate ceiling, not a requirement. Campaigns with a lot of historical conversion data can calibrate faster.

Why The Number Matters More Than The Calendar

For years the advice floating around was to give a bidding strategy two weeks and then judge it. That advice was always a little sloppy, and this update is Google saying so out loud.

Two weeks means something completely different depending on volume. A campaign pulling 40 conversions a week clears the bar in under two weeks. A campaign pulling 4 conversions a week needs closer to three months to reach the same place. Same calendar, completely different amount of information going into the model.

If you have been evaluating a low-volume account on a two-week clock, you have been making decisions on a model that was still guessing.

What Puts A Campaign Back Into Learning

Google lists three triggers that push a campaign back into a learning state.

• Creating or reactivating a bidding strategy. Turning a paused strategy back on counts. It does not pick up where it left off.
• Changing bidding strategy settings. This is the one that catches people. Moving a target CPA or target ROAS is a settings change.
• Certain campaign composition changes. Google is vague here on purpose, but large structural edits belong in this bucket.

Google also notes that Performance Max takes longer to reach its performance goals when most of the traffic comes from outside Search and Shopping. If your PMax campaign is leaning on Display and YouTube inventory, expect a slower ramp than the 50-conversion figure suggests on its own.

What To Do With This

Here is the process. Stop measuring learning in days and start measuring it in conversions.

1. Pull your conversion count since the last bid change. Not the date. The count. That is the number that tells you whether the campaign is calibrated.
2. Work out your conversion cycle. If people research for three weeks before converting, your cycle is three weeks, and three cycles is over two months.
3. Set a real evaluation date from those two numbers. Low-volume accounts should be looking at 6 to 12 weeks, not 14 days.
4. Batch your changes. If you are going to adjust a target, adjust it once and leave it alone. Three small tweaks across three weeks is three restarts.
5. Check that your conversion tracking is actually firing. Fifty conversions only helps if they are real ones. If you are counting button clicks as leads, you are feeding the model 50 pieces of bad information.

That last point is the one I would start with. Before you worry about hitting a conversion threshold, make sure the conversions you are sending are worth counting. My guide to tracking phone calls and lead forms walks through how to set that up properly.

My Take

This quietly confirms something a lot of small advertisers already suspected. Smart Bidding is built for volume. If you do not have volume, it takes a long time to work, and there is no setting that speeds that up.

One of the biggest mistakes I see is a business owner switching bidding strategies every few weeks because performance looks bad. Every switch restarts the clock. The account never gets past the messy part, and then the conclusion is that Smart Bidding does not work, when it was never allowed to finish learning.

Keep in mind, if your account genuinely cannot generate 50 conversions in a reasonable window, that is a signal in itself. It may be worth counting a softer action as a secondary conversion while you build volume. Picking the right approach starts with understanding what each of the bidding strategies actually optimizes for.

Question to Answer:

How many conversions has your campaign recorded since the last time you changed a bid setting, and is that number anywhere close to 50?

Read the source