> Quick answer: Wait at least 6 weeks and accumulate 100+ conversions before making a pause decision. Sustained underperformance on cost per conversion, zero-impression keywords, or low Quality Score are the signals that actually matter. Optimize first. Pause second. Remove last.
Killing a Google ad too early is one of the most expensive mistakes in paid search. The algorithm needs time to learn, and pausing resets that progress. Here's exactly when to act, and when to hold off.
---
Don't Kill Too Early: The Learning Period
Patience in the first weeks of a campaign is not passive. It's the actual strategy.
New campaigns need at least 6 weeks to stabilize
Per Google Ads Help Center documentation on Performance Max campaigns, new campaigns need at least 6 weeks to ramp up. The machine learning algorithm is still gathering signal during this window. Cutting a campaign short at week two gives you incomplete data and an incomplete picture.
Why pausing resets the algorithm (and costs you data)
When you pause a campaign running smart bidding (Target CPA or Target ROAS), the algorithm loses its accumulated training data. Performance may not recover immediately when you resume. Google Ads documentation advises against making frequent changes during the initial learning phase, because each disruption restarts the ramp-up clock.
Every pause is a restart, not a rest. That distinction matters when you're trying to evaluate whether a campaign actually works.
How long to wait before evaluating performance
Wait at least 6 weeks from launch before drawing firm conclusions. After any significant change (new ad copy, new bid strategy, new budget), allow another 1-2 weeks before assessing impact. The numbers you see during the learning phase are not the numbers the campaign will settle at.
---
What Data Do You Actually Need?
Good kill decisions require enough data to be statistically meaningful, not just recent data.
The 100-conversion threshold for cost-per-conversion metrics
Per Google's Ads Help Center guidance on evaluating performance based on your goal, you need at least 100 conversions before cost-per-conversion metrics stabilize. With fewer conversions, a single outlier skews the average dramatically. A campaign sitting at $80 CPA with 15 conversions might settle at $35 with 150. Don't make permanent decisions with thin data.
Accounting for conversion delay when analyzing results
Conversions can happen up to 30 days after a click. That means last week's numbers are still incomplete. A customer who clicked your ad on Monday may not convert until next month. If you're evaluating the last 7 days, you're missing a large portion of the revenue those clicks will eventually drive.
Why recent data is incomplete and shouldn't drive kill decisions
Pull 30-60 day windows for analysis. If your sales cycle is long, consider filtering out the most recent 14-30 days before drawing conclusions. Basing a pause decision on incomplete recent data produces false negatives. A campaign that looks like it's failing may just be mid-cycle.
---
Metrics That Signal It's Time to Pause
When the data is sufficient and the pattern is clear, these are the signals that justify action.
Cost per conversion above your target by a sustained margin
One bad week is noise. A sustained cost per conversion above your target, measured across 100 or more conversions and 30 or more days, is signal. Define your acceptable threshold before you start analyzing. Then measure against it objectively.
Keywords with zero impressions for extended periods
Google handles some of this automatically. Per the Google Ads Help Center documentation on auto-pause for low activity keywords, Google automatically pauses positive keywords that are over 13 months old with zero impressions in the past 13 months. You can manually unpause them if you expect conditions to change. But if a keyword has drawn no impressions, it is not competing in the auction.
Quality Score and CTR filters for identifying poor performers
Per the Google Ads Help Center guidance on improving keywords for the Search Network, you can filter your keyword list by Quality Score, CTR, and first-page bid estimate to surface underperformers. A low Quality Score signals a mismatch between the keyword, ad copy, and landing page. A CTR below 0.2% with 1,000 or more impressions is a consistent red flag for relevance problems.
Don't treat these filters as automatic kill criteria. Treat them as a list of candidates to investigate.
When automated rules can help (and what thresholds to set)
Automated rules remove emotion from pause decisions. Per Google Ads documentation on common ways to use automated rules, example thresholds include. pausing keywords where cost per conversion exceeds $20 with 100 or more conversions in 30 days, or pausing ads with CTR below 0.2% and 1,000 or more impressions. Automated rules typically execute within 2 hours of a condition being met.
Set your own thresholds based on your target CPA, not generic benchmarks. Generic benchmarks don't know your margins.
---
Try Optimization Before You Kill
A struggling campaign is often a fixable campaign. Optimize first, pause second.
Improve keyword relevance and Quality Score
Before pausing a keyword, try making it more specific. Tighter alignment between the keyword, the ad copy, and the landing page raises Quality Score and lowers cost per click. Per Google's Ads Help Center, improving relevance is the recommended first step before deleting underperforming keywords.
Refine match types and negative keywords
Broad match captures traffic you didn't intend to pay for. Adding negative keywords reduces wasted spend on irrelevant searches. Switching to phrase or exact match gives you more control over intent. These changes often recover a campaign that looks like it should be killed.
Update ad copy and landing page alignment
Ad copy that doesn't match search intent produces low CTR. A landing page that doesn't deliver on the ad's promise hurts conversion rate. Fix the message before you write off the campaign. Many "poor performers" are poor communicators, not poor audiences.
Audit targeting and audience signals
Check device, location, and time-of-day performance before making a campaign-level decision. One underperforming segment dragging down the average doesn't mean the whole campaign should stop. Bid adjustments and exclusions can isolate the problem without destroying what's working.
---
How to Pause vs. Remove a Campaign or Keyword
These are two different actions with very different consequences.
Pause to preserve data and allow future resumption
Per Google's Ads Help Center documentation on enabling, pausing, and removing campaigns, pausing a campaign stops ads from showing but keeps all historical performance data intact. You can resume at any time. Pausing is reversible, which makes it the right default for any decision you're not fully certain about.
Remove only when you're certain (it's permanent)
Removing a campaign is permanent. Google's documentation is explicit. there is no undo function. Remove a campaign only when you have no intention of running it again and no need for the data. That threshold should be high.
Using automated rules to pause systematically
Automated rules are best for conditions you check repeatedly. Set a rule to pause keywords above a cost-per-conversion threshold with sufficient conversion volume. This takes the daily emotional calculation out of the process and keeps decisions consistent.
Manual pausing for one-off decisions
For campaigns you're actively testing, manual pausing gives you more deliberate control. Review the full data window, confirm the numbers are stable over enough time, and pause with a clear rationale you can document. That documentation matters when you revisit the decision later.
---
What Happens After You Pause. And How to Monitor
Pausing isn't the end. It's a checkpoint.
Algorithm learning resets when you pause campaigns with smart bidding
Resuming a paused campaign with Target CPA or Target ROAS triggers a new learning phase. Budget for a ramp-up period again, just as you did at launch. Expect some performance volatility in the first 1-2 weeks after resumption.
How to track performance in Advertise reporting
Having a single place to see performance across campaigns matters when you're making pause decisions. Coinis's Advertise page gives you real-time visibility into spend, conversions, and cost per conversion. You can identify the sustained underperformance that justifies a pause, rather than reacting to a single bad day. Live reporting currently covers Meta campaigns. Google Ads reporting is on the roadmap.
When to give a paused campaign another chance
Revisit paused campaigns every quarter. Ad costs shift. Competition changes. A keyword that didn't convert six months ago might perform with updated copy, a new landing page, or a shifted bidding strategy. Paused doesn't mean dead.
---
Or let Coinis do it.
From a product URL to a live Meta campaign. AI-generated creatives. On-brand copy. Direct publish to Facebook and Instagram. Real performance reporting. All in one platform.
Start free. Upgrade when you're ready.
15 AI tokens a month. No credit card.
Frequently Asked Questions
How long should I run a Google Ads campaign before pausing it?
Google recommends at least 6 weeks for new campaigns. The machine learning algorithm needs this time to gather enough data to optimize. Pausing before the learning phase completes resets the clock and wastes whatever data you've already collected.
What's the difference between pausing and removing a Google Ad?
Pausing stops your ads from showing but keeps all historical data intact and allows you to resume at any time. Removing is permanent. there is no undo. Per Google's Ads Help Center, remove a campaign only when you're certain you'll never run it again and don't need the data.
How many conversions do I need before evaluating cost per conversion?
Per Google Ads Help Center guidance, you need at least 100 conversions before cost-per-conversion metrics are statistically reliable. With fewer conversions, a single outlier can skew the average significantly and lead to a bad pause decision.
Can automated rules replace manual campaign monitoring?
Automated rules handle repetitive threshold-based decisions well, like pausing keywords that exceed your target CPA over 100+ conversions. But they don't replace judgment calls on creative quality, audience fit, or landing page alignment. Use both together.