Two clocks run at once and almost nobody separates them. One is whether the bid strategy has adjusted. The other is whether the data you are reading is even complete yet. You can wait out the first and still be reading a number that is wrong.

On this page

• How long does Google Ads conversion data take to settle?
• How long does Smart Bidding take to adjust after a change?
• How do I measure settlement on my own account?
• Why did testing settlement on a 7-day window give the wrong answer?
• Can I use conversion lag data to correct a conversion-date figure?

How long does Google Ads conversion data take to settle?

A by-conversion-date figure takes about 4 days to finish being written. Measured against a saved snapshot, windows ending 4, 9, 16, 23, 30 and 35 days ago all backfilled 0.0% on re-read. A window ending 1 day ago backfilled +83.7%.

So: never read a window containing days younger than 4 days. Full stop. That rule alone prevents most premature reversals.

How long does Smart Bidding take to adjust after a change?

Separate question, and it depends on conversion volume, not on calendar time. A campaign getting 5 conversions a week is not going to resolve anything in 14 days regardless of how patient you are.

How do I measure settlement on my own account?

The method generalises. My numbers don't — settlement differs by account.

1. Pull your key metric by conversion date for a set of windows: ending 1, 4, 9, 16, 30 days ago.
2. Save the snapshot with a timestamp.
3. Wait 6 days. Pull the identical windows again.
4. Compare. Where backfill hits 0.0%, that's your settled boundary.

Now you know the youngest window you are allowed to read on your own account.

Why did testing settlement on a 7-day window give the wrong answer?

Settlement was once validated on a 7-day window whose last day or two were fresh. Dilution held the measured backfill to +3.5%, which read as "settles fast". That figure was then applied to a 3-day window sitting entirely inside the fresh zone.

Both "ended 1 day ago". Completely different exposure. The 3-day window read as a 55% collapse and nearly triggered a strategy reversal. Nothing had happened.

Test at day grain, or on a matched-length window. Never validate on a long window and apply to a short one. A 3-day window that is 1 day old is 100% unsettled.

Can I use conversion lag data to correct a conversion-date figure?

Do not use click-to-conversion lag data to "correct" a conversion-date figure. They are different measurements. Applying one to the other manufactures a recovery that is not there.

Related questions

How much does a plumbing or electrical company need to spend on Google Ads before it starts booking jobs, and how long should it take?

There is no universal floor in dollars — the honest floor is defined by your cost per booked job and the volume Smart Bidding needs to steer, which for most trades means enough budget to produce roughly 15 conversions a month at your local cost per lead. Read the full answer.

My average CPC doubled and I didn't change anything — what actually causes that?

The universal published answer is "a competitor entered the auction," which is unfalsifiable and useless. Before you accept it, separate a genuinely more expensive auction from a shifted traffic mix. Read the full answer.

My Google Ads conversions dropped and I didn't change anything. Is it broken, or is the data still filling in?

Before you diagnose anything, check whether the window you are reading is old enough to be complete. If any day in it is less than 4 days old, the number is wrong and there is nothing to diagnose yet. Read the full answer.

Written by Usama Khan, who runs Google Ads for e-commerce brands and home-services companies at AdventurAds. Figures are from accounts under active management and are anonymised. Last reviewed 2026-08-26.