Last updated: February 12, 2026
You logged in this morning and something is wrong. Conversions have tanked. CPA has doubled overnight. Your campaigns that were printing money last week are now bleeding budget with nothing to show for it. You are staring at your Google Ads dashboard with that sinking feeling in your stomach, wondering what the hell happened.
Take a breath. You are not alone, and this is almost certainly fixable.
An analysis of over 17,000 Google Ads campaigns found that performance fluctuations affect the vast majority of advertisers at some point. In 2025, 87 percent of industries saw a 13 percent increase in CPC, while 91 percent of industries experienced a 14 percent decrease in conversion rates. Performance drops are not the exception. They are the norm. The question is whether you can diagnose and fix them before they drain your budget.
This guide gives you a structured 15-minute diagnostic framework to identify exactly what went wrong with your Google Ads campaigns. We will cover the five most common causes of sudden performance drops, walk you through the exact steps to pinpoint the problem, and explain why these issues keep recurring for advertisers who rely on manual campaign management.
Before we dive into each problem in detail, here is the step-by-step sequence you should follow the moment you notice a performance drop. Print this out. Bookmark it. You will use it more than once.
Open your Google Ads account and navigate to "Change history" in the left sidebar. Filter for the timeframe that aligns with when performance started declining. Look for any changes made to bid strategies, budgets, conversion actions, ad copy, keywords, targeting, or campaign settings. Also check for automated changes that Google may have applied through auto-applied recommendations. This single step identifies the root cause roughly 60 percent of the time.
Go to Goals, then Conversions, then Summary. Check the status column for each conversion action. Look for anything marked "Inactive," "No recent conversions," or showing setup warnings. If conversions went from a normal volume to zero overnight, the problem is almost always a broken tag, not a campaign issue. Also check whether enhanced conversions are showing any diagnostic warnings. We will cover this in detail shortly.
Navigate to the campaign or ad group experiencing the drop and open Auction Insights. Compare your impression share, overlap rate, and outranking share to the previous period. Look for new competitors entering the auction or existing competitors increasing their impression share. If your impression share dropped while your competitors' went up, you are being outbid.
Pull the search terms report for the affected campaigns and compare the current period to the previous one. Look for an influx of new, irrelevant search queries that are consuming budget. Check if any match type changes or keyword additions are pulling in broader traffic. Look for seasonal or trending searches that are matching to your keywords but converting at much lower rates.
Verify your landing pages are loading properly and your website is not experiencing downtime. Check if your payment method is active and has not been declined. Look at Google Ads status updates for any platform-wide issues. And compare your performance to industry trends to determine if the drop is market-wide or specific to your account.
If you have followed these steps, you should now have a strong hypothesis about what went wrong. The sections below will help you confirm and fix it.
This is by far the most common and least understood cause of sudden performance drops. And here is the frustrating irony: it is usually triggered by you trying to improve your campaigns.
Every time you make a significant change to a campaign running on Smart Bidding (which includes Target CPA, Target ROAS, Maximize Conversions, and Maximize Conversion Value), you trigger what Google calls the "learning phase." During this period, Google's algorithm essentially resets and begins testing different bid levels, audiences, and placements to recalibrate toward your new settings.
The learning phase typically lasts 7 to 14 days, but it can extend to 30 days for campaigns with lower conversion volume. Google's own documentation states that it takes up to 50 conversion events or 3 conversion cycles for the bid strategy to calibrate to a new objective. During this period, performance is unstable. CPAs can run 2 to 3 times higher than normal. Google may bid aggressively on some auctions and barely participate in others. Your cost per acquisition might spike to levels that feel alarming.
Here is what triggers a learning phase reset. Switching your bid strategy (for example, moving from Maximize Clicks to Target CPA). Changing your CPA or ROAS target by more than 15 to 20 percent. Increasing or decreasing your budget by more than 20 percent. Adding or removing conversion actions. Significantly altering your keyword list, ad groups, or targeting settings. Pausing and restarting a campaign.
The trap is this: when you see performance dip during the learning phase, your instinct is to make another change to fix it. But every new change resets the learning phase again. You end up in an endless loop of instability where the algorithm never gets enough stable data to optimize properly. One industry expert described it as "a pilot trying to land a plane while someone keeps grabbing the controls."
How to fix it. First, check your bid strategy status. If it says "Learning," do not make any additional changes. Wait. The algorithm needs 7 to 14 days of stable data to calibrate. If you absolutely must make changes during the learning phase, make them all at once rather than trickling them in over multiple days, because each individual change restarts the clock.
Going forward, adopt the rule of making no more than one significant campaign change per week per campaign. And when you do make changes, keep them incremental. Adjusting your Target CPA by 10 percent is fine. Cutting it in half will cause significant turbulence.
This one catches even experienced advertisers off guard. Your campaigns look like they stopped converting, but in reality, your conversion tracking broke and your campaigns are still performing fine. You just cannot see it.
Conversion tracking breaks more often than most people realize. Common causes include website updates that remove or alter the tracking code, CMS or plugin updates that interfere with tag firing, changes to your cookie consent implementation, Google Tag Manager container modifications, and SSL certificate issues that prevent tags from loading.
The telltale sign is a sudden, sharp drop in conversions to near zero, especially if your impressions, clicks, and CTR remain stable. If people are still clicking your ads at the same rate but nobody is "converting," the most likely explanation is that conversions are still happening but your tracking is not recording them.
How to fix it. Go to your Google Ads conversion summary and check the status of each conversion action. Google shows statuses like "Recording conversions," "No recent conversions," "Tag inactive," and "Unverified." Any status other than "Recording conversions" needs investigation.
Use Google Tag Assistant (the Chrome extension) to verify that your conversion tags are firing correctly on your conversion pages. Navigate through the conversion flow yourself and watch the Tag Assistant debugger to confirm each tag fires at the right moment with the correct parameters.
If you use Google Tag Manager, open the GTM preview mode and walk through a conversion. Verify that the correct triggers are firing and that the tags are executing in the right order.
One critical check that is often overlooked: make sure your conversion action has not been accidentally set to a "Secondary" optimization goal instead of "Primary." This change will not break tracking, but it will prevent Smart Bidding from optimizing toward that conversion, which has the same practical effect as turning off your tracking for bid optimization purposes.
This is one of the most actively searched Google Ads problems right now, and for good reason. Enhanced conversions have gone from a nice-to-have feature to an essential component of conversion tracking accuracy. As browser privacy restrictions tighten and third-party cookies continue to fade, enhanced conversions help recover conversion data that would otherwise be lost.
But the setup is notoriously finicky. The most common warning message advertisers see is "Enhanced conversions has setup issues impacting performance," which can appear even when you think everything is configured correctly.
The typical issues fall into a few categories. Incomplete implementation means enhanced conversions are enabled in your account settings but the tag is not actually sending hashed customer data. This happens when the Google tag or GTM configuration is missing the user-provided data variables. Empty data parameters occur when the enhanced conversion tag fires but sends empty values because the CSS selectors or data layer variables are not correctly mapped to the form fields on your conversion page. Consent configuration errors arise when enhanced conversions are not receiving proper consent signals, particularly common for businesses operating under GDPR or similar privacy regulations. Low match rates happen when the hashed data being sent does not match enough Google account data to be useful, often because email addresses are being captured with formatting issues or phone numbers are missing country codes.
How to fix it. In Google Ads, go to Goals, then Conversions, then Settings. Check the enhanced conversions diagnostic report. It will tell you specifically what is wrong. The most common fix is ensuring your Google tag is properly capturing and hashing the customer email address on your conversion page. Use Tag Assistant to verify the "em" parameter is present and contains a proper hash value (a long string, not an empty field).
If the em parameter shows as "tv.1~em." with nothing after it, your tag is configured to send data but the field is empty. This usually means your CSS selector is pointing to the wrong element on the page or your data layer variable is not populating. If the em parameter is entirely missing, your enhanced conversions tag configuration needs to be rebuilt.
For businesses using CRM integrations like HubSpot or Salesforce for enhanced conversions for leads, a common issue is low volume. Google needs a minimum number of conversions before it can effectively match and validate the data. If your lead volume is light, it may take longer for the system to show green status indicators.
Sometimes your campaigns are fine. Nothing changed on your end. Your tracking is working. Your bids are stable. But performance still dropped. When this happens, the cause is almost always external.
The most common external factor is a new competitor entering the auction or an existing competitor significantly increasing their spend. When someone new starts bidding aggressively on your keywords, two things happen simultaneously. First, CPCs rise because the auction is more competitive. Second, your impression share drops because you are losing more auctions at your current bid levels. The combined effect is fewer impressions, higher costs per click, and often lower conversion rates because the searcher now has more options to choose from.
Seasonal shifts can also explain performance drops that feel sudden but are actually predictable. If you sell B2B software, January performance often dips because buying committees are finalizing budgets, not making new purchases. If you sell consumer products, performance drops in January after the holiday buying frenzy. Industry events, regulatory changes, and even macroeconomic shifts in consumer confidence can move the needle.
How to diagnose it. Auction Insights is your primary tool here. Compare your impression share, overlap rate, and position above rate to the previous period. If you see new domain names appearing in the competitive landscape, or existing competitors with significantly higher impression share, you have your answer.
Google Trends can help you determine if overall search demand for your keywords has shifted. If search volume dropped across the board, your performance drop may be demand-driven rather than competition-driven.
How to fix it. If a new competitor entered, you have three options: increase your bids to maintain position (which costs more), improve your Quality Score to compete more efficiently at the same bid level, or pivot to less contested keywords where you can still win profitably. Often the best approach is a combination: hold your ground on your highest-converting keywords while finding new keyword opportunities the competitor has not discovered yet.
Quality Score is the metric that determines how much you pay per click and how often your ads show. It is built from three components: expected click-through rate, ad relevance, and landing page experience. Each component is rated as "Above average," "Average," or "Below average." When any of these components degrades, your costs go up and your impression share goes down.
Quality Score degradation is sneaky because it usually happens gradually, not overnight. Your landing page might slow down after a website update. Your ad copy might become less compelling as competitors improve theirs. Your CTR might decline as ad fatigue sets in after running the same creatives for months. Each of these factors individually might cause a small dip. Together, they can trigger a significant performance decline.
How to diagnose it. In your keywords report, add columns for Quality Score, Expected CTR, Ad Relevance, and Landing Page Experience. Sort by Quality Score and look for keywords where the score has declined. Compare current scores to historical data (you can view Quality Score history by segmenting by day).
Pay particular attention to your highest-spend keywords. A Quality Score drop on a keyword that accounts for 20 percent of your budget will have far more impact than the same drop on a keyword that accounts for 1 percent.
How to fix it. The fix depends on which component is underperforming. For Expected CTR, test new ad copy with stronger headlines and more compelling calls to action. For Ad Relevance, make sure your ad copy closely matches the intent behind the keyword and includes the keyword itself in the headline. For Landing Page Experience, check page load speed (aim for under 3 seconds), ensure mobile responsiveness, and verify that the landing page content directly addresses what the ad promised.
One of the most confusing situations advertisers encounter is seeing "Limited by budget" appear on their campaigns after changing their bid strategy, even when the budget has not changed and was previously sufficient. This happens all the time, and the reason is straightforward once you understand how Smart Bidding works.
When you switch to a more aggressive bid strategy (say, from Maximize Clicks to Maximize Conversions, or when you lower your Target CPA), the algorithm may start bidding higher on individual clicks because it believes those clicks are more likely to convert. Higher individual bids mean your daily budget gets consumed faster, which triggers the "Limited by budget" warning.
This does not necessarily mean you should increase your budget immediately. During the learning phase, the algorithm is still calibrating. It might be overbidding on some auctions that it will eventually learn to skip. Wait for the learning phase to complete (7 to 14 days) and then reassess. If you are still limited by budget after the learning phase, evaluate whether the higher CPCs are producing proportionally better conversion rates. If they are, the budget increase may be justified. If they are not, adjust your CPA or ROAS target to bring bids back in line.
Another common scenario is when your Target ROAS is set too aggressively, causing the algorithm to only bid on a narrow subset of highly qualified auctions. This can make your campaigns appear "limited by budget" when in reality they are limited by the volume of auctions that meet your ROAS threshold. In this case, the fix is to relax your ROAS target slightly (try 10 to 15 percent increments) and see if volume increases without significantly impacting efficiency.
This is where many advertisers waste time and money. They see a performance drop, assume something is broken in their account, and start making changes. Those changes trigger a learning phase reset, which causes further performance instability, which prompts more changes, which triggers another reset. It is a vicious cycle, and it often starts because the original "problem" was not a problem at all. It was a normal market fluctuation.
If all or most of your campaigns declined at the same time and by a similar percentage, it is more likely a market-wide shift than an account-specific issue. If your Auction Insights show your impression share remained stable but overall impression volume dropped, demand decreased across the market. If Google Trends shows declining search interest for your core keywords during the same period, the market contracted temporarily.
Seasonal patterns are remarkably consistent year over year. If you have historical data, compare your current performance to the same period last year. Many industries see predictable dips in January (post-holiday), mid-summer, and early September. B2B companies often see slowdowns around major holidays and end-of-quarter budget cycles.
What to do when it is a market shift. Resist the urge to make dramatic campaign changes. Instead, adjust your expectations and possibly reduce budgets temporarily to maintain efficiency. The worst thing you can do is overreact to a seasonal dip by restructuring your campaigns, because you will trigger learning phase resets right when you need stability most.
If only specific campaigns declined while others remained stable, the issue is localized and likely account-specific. If your click volume held steady but conversions dropped, check your conversion tracking and landing pages. If your impression share dropped while competitors' shares increased, you are losing the auction. If the decline coincided precisely with a change in your change history, the cause is almost certainly that change.
Every diagnostic step we have covered so far follows the same pattern. Something went wrong. You noticed. You investigated. You fixed it. Then you wait for the next thing to go wrong. This is reactive management, and it is how 99 percent of Google Ads accounts are managed.
Think about what reactive management actually means in practice. A learning phase resets and performance dips for two weeks before you notice and respond. A competitor enters your auction and outbids you for several days before you check Auction Insights. Your conversion tracking breaks on a Friday evening and you do not discover it until Monday morning, losing an entire weekend of data. An irrelevant search term starts consuming 10 percent of your budget, and you do not catch it until your weekly or monthly search terms review.
Every one of these delays costs money. Not theoretical money. Real, measurable wasted spend that accumulates in the gap between when a problem occurs and when you detect and fix it.
Here is the uncomfortable truth. The five problems we covered in this article are not one-time events. They are recurring patterns that happen to every Google Ads account, constantly.
The learning phase resets every time someone adjusts a campaign. Conversion tracking can break after any website update, CMS change, or plugin update. Competitors enter and exit your auction constantly. Quality Scores shift as competitors improve their ads and as Google updates its algorithms. Search query patterns evolve as consumer language changes, trends emerge, and seasonal patterns cycle.
If you are managing your campaigns manually, or relying on an agency that reviews your account weekly or biweekly, you are always going to be playing catch-up. You will always be diagnosing problems after they have already cost you money. You will always be reacting to yesterday's performance data while today's problems compound unseen.
This is not a criticism of the people managing these campaigns. It is a structural limitation of human-paced management applied to a system that operates in real time.
Imagine a scenario where learning phase impacts are minimized because bid adjustments are made continuously in small increments rather than in large, disruptive chunks. Where conversion tracking anomalies are detected within minutes, not days. Where new competitor activity is identified and responded to in the same auction cycle it appears. Where irrelevant search terms are caught and excluded in real time, before they waste a single dollar. Where Quality Score components are monitored continuously with proactive adjustments to ad copy and landing page alignment.
This is not hypothetical. This is exactly what groas does. Because groas operates autonomously within the Google Ads ecosystem, with deep integration into Google's tools and APIs, it processes the same real-time signals that determine your campaign performance and responds to them at the speed they actually occur. Not tomorrow. Not at the next weekly optimization session. Right now.
The difference between reactive and proactive management is not just about catching problems faster. It is about preventing many of those problems from happening in the first place. groas makes gradual, continuous adjustments that avoid the kind of disruptive changes that trigger learning phase resets. It monitors conversion tracking health as a core function, not an afterthought. It tracks competitive dynamics in the auction continuously rather than checking Auction Insights once a week.
For advertisers who are tired of the diagnostic cycle we described in this article, who are exhausted from constantly firefighting yesterday's problems while today's issues pile up, groas represents a fundamentally different approach. Instead of managing campaigns in bursts of human attention, groas applies consistent, continuous optimization that keeps your account healthy around the clock. It does not eliminate every possible issue (external market shifts will always happen), but it dramatically reduces the time between problem occurrence and resolution, and it prevents the self-inflicted problems like learning phase resets and missed negative keywords entirely.
If CPCs jumped without any changes on your end, check three things. First, look at Auction Insights for new competitor activity. Second, check if Google auto-applied any recommendations to your account (this is increasingly common and can include broad match keyword additions, budget increases, or targeting expansions that change your competitive position). Third, verify that no match type changes occurred. Google's continued expansion of broad match behavior means your exact match keywords may be matching to queries that previously would not have triggered your ads, putting you into more expensive auctions.
To check for auto-applied recommendations, go to Recommendations in the left navigation, then click "Auto-apply" at the top right. Review the history of changes Google made automatically and reverse any that do not align with your strategy.
This often indicates that your conversion tracking is recording fewer conversions (even if all tags are technically firing), which causes Smart Bidding to become more conservative. Common causes include a website redesign that changed the conversion funnel, a form update that introduced more fields (increasing abandonment), a change in your product pricing that reduced purchase intent, or a landing page speed regression that increased bounce rates.
Run a page speed test on your key landing pages using Google's PageSpeed Insights. If your mobile score dropped below 50 or your load time exceeds 3 seconds, that alone could explain a significant conversion rate decrease. Research consistently shows that each additional second of load time reduces conversions by 7 to 12 percent.
Google's in-platform recommendations are designed to increase spending, not necessarily to improve your performance. Some recommendations are genuinely helpful. Others, like adding broad match keywords, enabling Search Partners, or raising budgets, can flood your account with low-quality traffic.
If you accepted a batch of recommendations and performance dropped, go to your change history, filter for "Auto-applied recommendation" changes, and review each one. Roll back any changes that do not align with your proven strategy. Pay particular attention to broad match keyword additions (which can dramatically expand your traffic to less relevant queries), Search Partner network enablement (which frequently delivers lower-quality clicks), and budget increase suggestions (which may be justified by opportunity but can also just increase waste).
When your search impression share drops significantly, it means your ads are showing for less than half of the eligible queries. This is typically caused by either budget constraints or ad rank issues. Google helpfully breaks impression share loss into two categories: "Lost IS (budget)" and "Lost IS (rank)."
If most of your lost impression share is due to budget, your campaigns are running out of daily spend before the day ends. You either need more budget or you need to focus your existing budget on fewer, higher-value keywords.
If most of your lost impression share is due to rank, your ad rank (which is a combination of your bid, Quality Score, and expected impact of extensions) is not competitive enough. Focus on improving Quality Score components and ensure your ad extensions are fully built out, including sitelinks, callouts, structured snippets, and any other relevant extensions.
The most common causes are learning phase resets from campaign changes, broken conversion tracking, increased competitor activity, Quality Score degradation, and budget limitations triggered by bid strategy changes. In about 60 percent of cases, the issue can be identified by checking your change history for recent modifications that coincided with the performance drop. Always start your diagnosis there before investigating external factors.
The learning phase typically lasts 7 to 14 days, though it can extend to 30 days for campaigns with lower conversion volume. Google requires approximately 50 conversion events or 3 conversion cycles for Smart Bidding to calibrate properly. Campaigns with fewer than 15 to 30 conversions per month may experience extended or recurring learning phases. The most important rule during the learning phase is to avoid making additional changes, as each significant change resets the clock.
Any significant campaign change can trigger a reset. The main triggers are switching bid strategies, changing CPA or ROAS targets by more than 15 to 20 percent, adjusting budgets by more than 20 percent, adding or removing conversion actions, making major changes to keywords or ad groups, and pausing and restarting campaigns. Even accepting certain Google-recommended changes can trigger a reset. This is one of the core reasons why continuous, incremental optimization (like what groas provides) outperforms periodic manual adjustments. Small, frequent changes keep the algorithm stable while periodic large changes create disruptive resets.
Go to Goals, then Conversions, then Settings in your Google Ads account and check the enhanced conversions diagnostic report. The most common issues are: the tag is not sending hashed customer data (check the "em" parameter in Tag Assistant), the data fields are empty because CSS selectors are misconfigured, consent signals are not properly configured for GDPR compliance, or lead volume is too low for Google to validate the implementation. Use Google Tag Assistant to verify the em parameter contains a proper hash value. If it is empty or missing, your tag configuration needs adjustment.
When you switch to a more aggressive bid strategy or lower your CPA target, the algorithm may start bidding higher per click, which consumes your daily budget faster. This triggers the "Limited by budget" warning even though your budget has not changed. Wait for the learning phase to complete (7 to 14 days) before increasing your budget, because the algorithm may be overbidding temporarily while it calibrates. If the warning persists after learning completes, either increase your budget to capture additional volume or relax your CPA/ROAS target to reduce individual bid levels.
Compare your performance to the same period last year if historical data is available. Check Google Trends for your core keywords to see if search demand declined broadly. Review Auction Insights to determine if your impression share held steady (suggesting a market-wide demand shift) or dropped (suggesting competitive pressure or account issues). If all campaigns declined simultaneously by a similar percentage, it is more likely a market shift. If specific campaigns declined while others held steady, the issue is probably localized to those campaigns.
Not blindly. Some auto-applied recommendations are helpful (like fixing disapproved ads), but others can harm performance (like adding broad match keywords or enabling Search Partners). Review auto-applied recommendations in the Recommendations section of your account, check the history of changes Google made, and reverse any that conflict with your strategy. Consider disabling auto-apply for categories that have historically caused problems in your account. This is another area where groas provides significant value, as it continuously evaluates recommendation quality and only applies changes that align with your actual performance goals.
Go to Goals, then Conversions, then Summary. Check the status column for each conversion action. "Recording conversions" is healthy; anything else needs investigation. Compare your daily conversion count to recent history. If conversions dropped to zero while clicks remained stable, your tracking is almost certainly broken. Use Google Tag Assistant to verify tags are firing on conversion pages. Walk through your own conversion flow with Tag Assistant active and confirm each tag fires at the correct moment. Also check if any conversion actions were accidentally changed from "Primary" to "Secondary" optimization goals.
Because manual campaign management is inherently reactive. You cannot continuously monitor every signal that affects performance: auction dynamics change by the hour, search term patterns shift daily, conversion tracking can break at any moment, and competitors adjust their strategies constantly. Human managers check these signals periodically, in weekly or biweekly reviews, which means problems always accumulate between review cycles. This is the fundamental limitation that autonomous AI management solves. groas monitors all of these signals continuously and responds in real time, which prevents recurring problems like learning phase resets, negative keyword gaps, and conversion tracking issues from persisting long enough to impact your results.
Follow the 15-minute diagnostic framework outlined in this article: check your change history first (minutes 0 to 3), verify conversion tracking (minutes 3 to 6), review Auction Insights for competitive shifts (minutes 6 to 9), examine your search terms report for irrelevant traffic (minutes 9 to 12), and check external factors like landing page uptime and payment methods (minutes 12 to 15). This sequence is ordered by likelihood: change history and conversion tracking issues account for the majority of sudden performance drops, so checking them first gives you the fastest path to a diagnosis.
