How long should we wait after a change before judging a Google campaign?
Our answer
After a budget or target edit, let Google bidding settle 1 to 2 conversion cycles. Then judge over 2 cycles or 2 weeks, whichever is longer. Give a new campaign or bid strategy a month or 50 conversions.
- Confidence
- Medium
- Strongest proof
- 13 Google rules
- 3 datasets, 5 practitioners, 4 opinions or claims
- Sources
- 10 sources
- 25 statements
- G
GPL
WL
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentBudget and target edits never trigger Learning, which lasts at most 7 days, but performance converges over 1 to 2 conversion cycles (Google). Then evaluate over 2 conversion cycles or 2 weeks, whichever is longer, allowing for conversion delay.
Practitioners wait about two weeks or more. Elevarus puts learning at 7 to 14 days, WordStream gives a new bid strategy 14 days and Jetfuel Agency gives a first Target ROAS two to three weeks.
Low volume stretches the wait. groas's log of 40 lead gen, home services and SaaS launches found a median 34 days to a stable exit for campaigns under $5,000 a month. Amsive found two weeks too early to read Google's August change.
Our judgment
- After a budget or target edit on a running Smart Bidding campaign, wait 1 to 2 conversion cycles, then judge over 2 cycles or 2 weeks, whichever is longer. Leave out recent days still missing delayed conversions. Google's learning, measurement and evaluation pages all set the wait in conversion cycles, and its product manager says these edits cause no learning period.
- After a new campaign, a new or reactivated bid strategy or a new conversion goal, make no edits for 2 weeks and decide at about a month or 50 conversions. We combine Google's own windows, including 15 days for a Shopping Target ROAS and 30 days with 30 conversions for Target CPA, with the waits Elevarus, WordStream, WebFX, Jetfuel and groas give.
- Space target changes at least one conversion cycle apart and change one thing at a time. Google advises against several target changes in one conversion cycle, groas gives each change two weeks, Amsive advises against changing everything together, and Ana Kostic's all-at-once rebuild cost a month of traffic and sales.
- Don't judge when the Learning label clears. Google says it clears within 7 days while performance takes longer to converge. groas says an Eligible status does not mean the cost per conversion is dependable, and Menachem Ani says the label going away doesn't mean the campaign is stable.
- The same rule applies at every spend level and to lead gen and local businesses. Conversion volume and delay set the wait, not spend. Google states its rules in conversion cycles and conversions, and groas's log covers small lead gen and home services campaigns.
Our judgment, built from Google, Elevarus, WordStream, WebFX, Jetfuel Agency, groas, Amsive, Ana Kostic (PPC Live) and Menachem Ani: our reading of these sources, not a finding from any one of them.
What the sources say
25 statements from 10 sourcesMenachem Ani relays that Google says its Learning label lasts 7 days at most, but says a new Performance Max campaign still needs 3 to 6 weeks before you judge it.
“The label going away doesn't mean the campaign is stable.”
Google says judge Smart Bidding over at least 2 full conversion cycles, better a month or 50 conversions, allow for conversion delays, and wait 1 to 2 cycles after major changes before adjusting again.
“A good first step is to set a time frame that includes at least 2 full conversion cycles.”
Google says after a target change the bidder reacts at once but needs time to hit it, so wait 1 to 2 full conversion cycles before judging, and big target changes move spend by a similar amount.
“It’s recommended that you wait 1-2 full conversion cycles before comparing average target ROAS to the actual ROAS performance using the Bid Strategy Report.”
Google's bid strategy report shows conversion delay; Google says to wait until all conversions are reported before judging a bid strategy, and use a longer date range if gaps are not significant.
“Wait until all your conversions have been reported before evaluating the performance of your bid strategy.”
Google's peak-season advice: adjust budgets and targets in real time without a learning reset, judge results over one to two conversion cycles, not 24 hours, and uncap budgets with targets set to holiday goals.
“Look beyond a 24-hour snapshot and review performance over 1 to 2 conversion cycles”
WordStream's Michelle Morgan says to give a new Google bid strategy 14 days and change targets only 20% at a time. A Google product manager has reportedly said target changes cause no learning period.
“only adjust things 20% at a time”
Google says budget or target changes take effect at the next auction without triggering learning, but advises against several target changes in one conversion cycle and judging over 2 cycles or 2 weeks.
“Avoid making multiple target changes within a single conversion cycle unless there is a clear business need.”
Google says Smart Bidding's Learning status lasts up to 7 days while ads keep serving, budget or target changes never trigger it, and results should be judged over 2 conversion cycles or 2 weeks.
“If you have past conversion data in your account, Smart Bidding never starts from scratch.”
Google advises running experiments one at a time for at least 4 to 6 weeks, discards the first 7 days, and calls a spend gap under 20% between arms normal. Avoid reactive changes in the first 2 weeks.
“It’s recommended that you run the experiment for at least 4-6 weeks or longer if you have a long conversion delay.”
Google says Performance Max should use conversion tracking, and automated bidding typically needs 1 to 2 weeks, sometimes up to 6, plus 1 to 2 conversion cycles after big changes.
“Campaigns that use automated bidding strategies need time to optimize, typically 1-2 weeks, but sometimes up to 6 weeks.”
Google suggests starting new Search campaigns, or those with little conversion data, on Maximize Clicks, moving to Maximize Conversions or Target CPA once enough data builds. Learning typically lasts 5-7 days.
“For new campaigns or those with limited conversion data, consider starting with a "Maximize Clicks" bidding strategy to gather performance data.”
Google says Shopping Target ROAS needs at least 15 conversions per Merchant Center ID in 30 days, should start from the last 4 weeks' value per cost, and needs about 15 days before judging.
“Consider giving the bid strategy 15 days to run before you evaluate its performance”
Google says Target CPA can start with no conversion history, recommends a target equal to the last 30 days' average CPA, and warns that a target set too low can cost you conversions.
“For evaluation, we recommend you measure performance for the last 30 days, including at least 30 conversions.”
Google's Tal Akabas says changing the budget or target on a steady campaign does not start a learning period; only new campaigns learn, so launch them a few days before a promotion.
“You can change your bids or your targets confidently on an existing campaign and know that nothing's going to reset in the backend”
A review of 40 Google Ads launches in lead gen, home services and B2B SaaS found a median 34 days to stable results for campaigns under $5,000 a month; 28 of 40 were interrupted early, mostly on days 4 to 7.
“Among the sub-$5,000-monthly-budget campaigns that reached a stable exit, the median was 34 days.”
WebFX says to wait two to three weeks after switching to value-based bidding or changing conversion signals before judging results.
“give it about two to three weeks before you judge the results.”
groas, which sells AI-run Google Ads management, warns a high ROAS target can shrink volume while ROAS looks better; lower it in 10 to 20% steps, two weeks each, and judge profit dollars.
“The cost is easy to miss because the ROAS column improves as volume falls.”
Disputes that target edits no longer reset learning: PPC Hero says a target change over 20% triggers a new learning period, so move in smaller steps and wait a conversion cycle between changes. No data.
“A target change bigger than 20% triggers a fresh learning period”
Amsive's agency cohorts after Google's August 2026 target change: tCPA Search CPA ran 14% above target in week one, 62% in week two. It says two weeks is too early to call. Account counts not given.
“two weeks is not enough to know whether conversion volume is actually the reason.”
Amsive, a media agency, saw ROAS fall roughly 29% to 35% in its retail Target ROAS campaigns the week after Google's August 17, 2026 bidding change. Small group, one week, no control.
“A target that was once treated as a loose guardrail is now a more direct instruction.”
Since August 17, 2026, Google bids budget-limited campaigns on target-based strategies closer to their target, so a $10 target CPA now delivers nearer a $10 actual CPA. It won't change targets or budgets for you.
“If your campaign's Target CPA is $10 USD, but your recent actual CPA performance is $5 USD, your campaign now delivers more closely to a $10 USD actual CPA.”
Jetfuel, an agency, picks Google bidding by monthly account conversions: Maximize Clicks under 30, Maximize Conversions at 30 to 80, Target ROAS from 80. It starts Target ROAS at 80% of actual ROAS.
“set your Target ROAS at 80% of your current actual ROAS when first switching to smart bidding. Give it two to three weeks to stabilize.”
Ana Kostic says restructuring a Google Ads account all at once disrupts learning. One such rebuild cost about 40% of traffic and sales.
“radical restructuring is theoretically good for the practitioner, but it's not a good idea for the for the business.”
Elevarus, an agency selling this migration work, says Shopping campaigns with 30+ conversions in 30 days can move to Maximize Conversion Value without a target; fewer stay on Maximize Clicks. No data.
“Flag any campaign with 30+ conversions in the last 30 days as a Maximize Conversion Value candidate.”
Simran Harichand changed a target cost per acquisition mid-month without watching it, and an 80,000 euro monthly budget was heading for a 30,000 euro underspend. Watch any change for at least a week.
“I relearned how important it is to monitor a change that you make in the account and then look at it for at least a week.”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- Where each source standsA chart placing each source on the question
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer