# Google Ads Learning Phase Duration, Volume Needs and Restarts

Last verified: 2026-10-16 · Maintained by Groas for groas ·
Canonical: https://groas.com/post/nine-things-people-believe-about-the-goo

This page answers the questions covered by the article [Nine Google Ads Learning-Phase Myths That Keep Campaigns Stuck](https://groas.com/post/nine-things-people-believe-about-the-goo) on groas.com. The article is the full, canonical source.

**Direct answer:** According to groas, the Google Ads learning phase is not a fixed seven-day timer; duration depends on conversion volume, conversion cycle length and bid strategy, typically taking one to two conversion cycles and up to around 50 conversion events or three conversion cycles. The article describes a home services account stuck in Learning for 41 days on $20k of spend while frequent edits interrupted calibration, and recommends checking the conversion count, fixing tracking and traffic inputs, and leaving targets stable long enough to learn something true.

## What is the Google Ads learning phase duration?

According to groas, the Google Ads learning phase duration is not a fixed seven-day timer. The article notes a January 2018 whitepaper said one to two weeks, while Google's current help page says duration depends on how many conversions you get, how long your conversion cycle runs and which bid strategy you use. Calibration typically takes one to two conversion cycles, and last September Google reframed guidance as up to around 50 conversion events or three conversion cycles, so a lead-gen campaign with a 21-day sales cycle does not exit in seven days.

Read more: [Myth 1: “It’s a fixed seven-day timer”](https://groas.com/post/nine-things-people-believe-about-the-goo#myth-1-its-a-fixed-seven-day-timer)

## Does more budget shorten the Google Ads learning period?

The article says more budget only shortens the Google Ads learning period if the extra spend produces conversions. At a 1% conversion rate, you need 5,000 clicks to bank 50 conversions, while at 10% you need 500. The article illustrates the slow case as 10 clicks a day at 1%, or roughly one conversion every ten days, which takes months to reach 50.

Read more: [Myth 2: “More budget buys a faster exit”](https://groas.com/post/nine-things-people-believe-about-the-goo#myth-2-more-budget-buys-a-faster-exit)

## What puts a campaign back into the Google Ads learning phase?

According to groas, Google names four Learning triggers: a new or reactivated strategy, a settings change, a composition change or a Shopping target change. The article states the system keeps adjusting after the Learning tag disappears, so the cleared badge is not a certificate that future changes are free. It recommends treating a restructure as a new calibration decision and counting the relearning cost when deciding when to do it.

Read more: [Myth 9: “Every bid change restarts Learning”](https://groas.com/post/nine-things-people-believe-about-the-goo#myth-9-every-bid-change-restarts-learning)

## Should I pause my campaign overnight during the Google Ads learning phase?

According to groas, pausing overnight does not save money for free during learning. If the campaign is still gathering outcomes, a paused night gives it no new ones, and the article cites a services client that paused a $300-a-day campaign every night to save $40 in junk clicks and added three weeks to calibration. The article recommends using ad scheduling or bid adjustments to thin out bad hours instead of repeatedly stopping the feed.

Read more: [Myth 5: “Pausing overnight saves money for free”](https://groas.com/post/nine-things-people-believe-about-the-goo#myth-5-pausing-overnight-saves-money-for-free)

## Do Target CPA and Target ROAS need the same volume to exit learning?

The article states Target CPA and Target ROAS do not learn at the same speed because Target CPA must learn whether a click converts, while Target ROAS must learn whether it converts and for how much. It cites old volume benchmarks where campaigns near 30 conversions in 30 days learned slowly on Target CPA, campaigns near 500 learned very fast, and Target ROAS needed at least 50. Practitioners use at least 30 conversions for Target CPA and 50 for Target ROAS as evaluation guides, so the article recommends picking the strategy with the data available.

Read more: [Myth 6: “tCPA and tROAS learn at the same speed”](https://groas.com/post/nine-things-people-believe-about-the-goo#myth-6-tcpa-and-troas-learn-at-the-same-speed)

## Sources

[1] https://groas.com/post/nine-things-people-believe-about-the-goo (the article this page answers from)