Google caps Smart Bidding Learning status at 7 days

Google has rewritten its Google Ads help page on Smart Bidding learning, and the new version puts a hard ceiling on the Learning status: it never lasts more than seven days. The page, now titled About Smart Bidding learning, appears to replace an earlier article on the duration of the learning period and spells out which account changes trigger the status and which never do.

For an Australian business paying for Google Ads, the practical point is simple. Learning is a label about structural changes, seeing it is no reason to freeze the account, and a budget or target tweak does not send the account back into Learning, even though results still shift as the strategy moves to the new setting.

Hover over Learning status

Google's page lists three reasons a bid strategy shows Learning, and the account shows which one applies when you hover over the status.

• New strategy. The bid strategy was recently created or reactivated.
• Composition change. Campaigns, ad groups or keywords were added to or removed from the strategy.
• Goal change. The conversion goals or actions the strategy optimises towards changed significantly.

Applying a new portfolio bid strategy counts as a structural change as well. Google says it can display Learning for up to seven days, though it expects a fast ramp-up where the campaigns already carry conversion history. Bid strategy statuses are not available on Hotel campaigns.

Budget and target edits apply next auction, never Learning

The clearest new line in the document is a list of what never triggers Learning: budget and target edits, new or edited ads, and targeting tweaks to location, audience or language. Google says these never remove historical data either.

Budget and target updates apply from the next auction. Google's own example is a Target ROAS change from 500% to 400%. The strategy uses the new target immediately, performance converges on it over the conversion cycle, and no Learning status appears.

Creative edits work the same way, with one caveat. New or edited ads must pass the normal policy review before they serve, so a rewrite can still hold up delivery for a different reason.

This matters because Google's earlier version of this page listed a setting change as a Learning trigger, and in our experience plenty of agencies and in-house teams avoided routine budget moves for fear of resetting the algorithm. On Google's current account of it, a budget move does not reset learning, though performance still adjusts over the conversion cycle. Raising a daily budget takes effect immediately and lets a campaign scale without a Learning flag.

Smart Bidding settles within one to two conversion cycles

The document separates two things that usually get lumped together. The Learning status in the interface disappears within seven days. Performance settling to a new goal typically takes one to two conversion cycles, where a conversion cycle is the time from a click to the conversion it produces.

Google names three factors that govern how long that takes: conversion volume, the length of the conversion cycle, and campaign settings. Very tight targets, low daily budgets or narrow targeting can all slow the ramp-up.

So a plumber whose leads convert on the day should typically see Smart Bidding settle much faster than a solar installer whose enquiries take three weeks to become a signed quote, other things being equal, even though both watch the Learning label drop off within a week.

Google also warns that new campaigns or strategies with no history may see temporary swings while the system ramps up, and that ads keep entering auctions and serving as normal throughout. Learning does not pause delivery.

Reactivating 300 paused keywords beats a rebuild from scratch

The new FAQ section answers a question we hear often: does the campaign start from scratch when you switch bid strategies? Google says no. If the account has past conversion data, Smart Bidding builds its initial bidding models from query-level signals, historical conversion patterns and similar past campaigns in the same account.

We saw the value of that record auditing an account with 300 paused exact-match keywords and nine years of conversion history sitting idle. A rebuild with fresh keywords ignored the account's own record of what converted: the paused keywords had delivered leads at $175 each, against $442 to $689 for the new campaigns in their first nine days. The history is an asset, so reactivate and consolidate instead.

Google's FAQ also sets expectations on secondary metrics. Smart Bidding optimises only towards the primary conversion goal, so clicks, CPC and click-through rate may move around and are not meant to improve once Learning ends.

Campaign-per-keyword builds leave Target CPA under 30 conversions

The best practices section reads like a checklist for the way many small-business accounts are built wrong.

• One primary conversion goal. Optimise to a single, consistent goal set as the account default, and feed it through a tag or regular offline conversion uploads.
• Enough budget. Daily budgets need room for the system to capture conversions.
• Consolidated structure. Merge ad groups and campaigns where you can so conversion data pools into larger sets.
• Fewer target changes. Avoid multiple target changes inside a single conversion cycle unless there is a clear business reason.

The consolidation point is the one with teeth. Google's main Smart Bidding page recommends judging performance over periods with at least 30 conversions, or 50 for Target ROAS, such as a month or longer. Accounts split into a campaign per keyword theme, as single-keyword ad groups often are, rarely reach that volume per campaign, which in our experience is why Target CPA struggles in them.

The same page flags a label change from June 2026. "Maximize conversions with a Target CPA" becomes "Target CPA" and "Maximize conversion value with a Target ROAS" becomes "Target ROAS". Google says the bidding behaviour is unchanged and no action is needed in the account.

Judge results over two conversion cycles rather than one monthly report

Google's evaluation advice is specific. Allow one to two conversion cycles for the strategy to converge, then assess over two conversion cycles or a two-week window, whichever is longer, and account for conversion delay because conversions often arrive days after the click.

For businesses with long conversion cycles, that is a longer runway than a monthly report allows, and a change made mid-month will not be readable in that month's numbers. It is a cousin of the sample-size problem that makes proving a 25% lift take 66 days even at 420 clicks a day. Two weeks of data on a campaign with 15 conversions tells you very little either way.

For a business owner reviewing an agency's work, the questions are now concrete. If a campaign is in Learning, which of the three triggers caused it, and was the structural change deliberate? If performance dipped after a budget or target edit, the Learning label is not the cause, so what is? And has the account had time to settle and then two weeks or two conversion cycles of data, whichever is longer, before anyone called the result?

The next step is to open the Campaigns table and check the Bid strategy type column (add it under Modify columns, Attributes if missing), or the Bid strategies page (Tools, Budgets and bidding, Bid strategies) for portfolio strategies, hover over any Learning status to read the reason, and write down the date of the last structural change. Then allow one to two conversion cycles for performance to settle, set a review date at least two weeks out from there (or two conversion cycles, if longer), and leave the targets alone until it arrives.