Why did my conversions drop? A method to find the real cause

Why did my conversions drop suddenly?

A sudden conversion drop has a cause you can usually find within a few hours, as long as you look in the right order. Almost every case falls into one of four buckets, and each bucket leaves a different fingerprint in your data.

The trap is jumping to the most familiar explanation ("the algorithm", "seasonality") before checking the cheaper ones. The method below forces the order: data first, then changes, then isolation.

Is the drop real, or is it a tracking and reporting issue?

Check measurement before anything else, because a drop that isn't real can't be fixed by changing campaigns. Three reporting effects regularly make healthy accounts look broken.

1. Conversion lag. Google Ads reports conversions on the ad impression date, while other tools attribute them to the conversion date, and conversions can be reported up to 90 days after the click. The last few days of any Google Ads report will fill in over time, so a "drop" that only covers the last 2–3 days is often just lag.
2. Processing delays. For standard GA4 properties, Google says daily data takes about 12 hours to process, data processing can take 24–48 hours, and attribution credit for key events can change for up to 12 days. Don't compare yesterday with a fully processed week.
3. Invalid traffic cleanup. Google Ads removes some conversions it detects as spam or invalid traffic, which can show up as a small dip after the fact.

Then check that tracking still works. In Google Ads, go to Goals › Conversions › Summary and read the status of each conversion action: "Misconfigured" means conversions have stopped recording because of a setup error or a broken tag, and "Needs attention" means there are issues but recording hasn't stopped entirely. Finally, compare the same period in three places: the ad platform, GA4 and your CRM or backend. If sales or leads in the CRM are flat while the ad platform shows a drop, you have a measurement problem, not a performance problem. Our GA4 annotations guide explains how to mark these incidents so the next person doesn't re-investigate them.

What changed in the days before the drop?

Once the drop is confirmed as real, list every change made in the 7–14 days before it started, from every source, on one timeline. Most teams only look at the ad account, but the cause often sits somewhere else.

1. Your team's edits in the ad platforms

In Google Ads, Campaigns › Change history shows edits to accounts, campaigns and ad groups for up to two years, with filters by date, change type and campaign. The User column shows who made each change. Google also warns that some account-level settings and changes made by Google representatives during consultations may not be listed. Our Google Ads change history guide and Meta Ads activity history guide show how to read both tools.

2. Changes the platforms made for you

Automated entries in Google Ads change history appear under names such as "Google Ads system", and API-driven changes show the tool name. Auto-applied recommendations are a common blind spot, covered in our post on auto-applied recommendations.

3. Website, tracking and CRM changes

• Site releases, new landing pages, checkout or form changes, page speed regressions.
• New Google Tag Manager container versions, consent banner updates, removed or renamed events.
• CRM changes: new lead stages, required fields, routing rules or a different definition of a qualified lead.

4. Outside events

Holidays, a competitor's promotion, a price change, stock-outs, press coverage or a platform outage. Write down the date of each one, even if you think it's irrelevant.

How do I isolate which change caused the drop?

Isolate the cause by matching the exact shape of the drop (when, where, how big) against each change, and testing one hypothesis at a time. A change can only be the cause if its timing and scope fit the data.

1. Define the shift precisely. Which metric, from which date, by how much, compared with what baseline. "Leads down 35% since Tuesday vs the previous four Tuesdays" is testable; "results are bad" is not.
2. Segment before you guess. Split by channel, campaign, device, country, landing page and conversion action. A drop limited to one campaign points to the ad account; a drop across every channel points to the site, tracking or CRM.
3. Line up the timeline. Keep only the changes whose date is on or just before the start of the drop, and whose scope matches the segments that fell.
4. Write a prediction for each hypothesis. If the new form caused it, form-start rate should hold while submission rate falls, on all channels.
5. Check the prediction, then revert or test. Where you can, undo one change at a time or split traffic, and watch whether the metric recovers.

What can Google Ads tell me about a conversion drop?

Google Ads has a built-in feature called Explanations that suggests reasons for large performance changes. It works for Search, App, Performance Max, Demand Gen, Display and Video campaigns, and can point to bid, budget, conversion tracking, targeting, ad eligibility and auction changes.

Two limits matter: Explanations can't cover date ranges that include today or go back more than 90 days, and a change has to be large enough to trigger one. Treat it as a lead to verify, not a verdict.

Google's own troubleshooting article also lists what commonly moves performance: edits to bids, budgets, keywords, audiences and schedules; limited budgets; very narrow targeting; competitors' bidding; and shifts in market demand. One point deserves attention: a sudden drop in conversion tracking data can itself cause a drop in serving and spend, because Google Ads optimizes towards a conversion action it can no longer see. A tracking problem can turn into a real performance problem within days.

What does a conversion drop investigation look like in practice?

Here is an illustrative example (not a real client) of the method applied to a B2B lead generation account.

Example. Leads fall 35% starting on a Tuesday. The last three days are excluded to avoid conversion lag, and the drop still holds. Google Ads shows fewer conversions, but CRM leads fell too, so tracking is working. Segmenting shows the fall on Google Ads, Meta and organic traffic at the same rate, and clicks are flat. The timeline for the prior week lists a Google Ads budget increase (Friday), a new Meta creative set (Monday) and a website release that added a phone number field to the form (Tuesday morning). Only the release matches both the date and the "every channel" scope. The team makes the field optional, and the conversion rate returns to its baseline within a week.

Notice that the two ad changes were real and recent, and either would have been an easy scapegoat. The timeline plus segmentation ruled them out in minutes. When you need to present the result, our guide on explaining a conversion drop to a client covers the communication side.

How do I avoid the next drop becoming a mystery?

Keep a single change log that covers every source, so that the next time you ask "what changed?" the answer takes minutes instead of days. Platform histories only cover their own platform, have gaps, and don't capture site, CRM or offline changes.

• Log every change with date, owner, channel, what changed and why.
• Include automations and outside events, not only your team's edits.
• Review the log whenever a metric moves more than your normal weekly range.

You can start with a spreadsheet using our marketing change log template, or read the full marketing change log guide. Changeline is a free (in early access) tool built for this: a to-scale timeline per client where the whole team logs Google Ads, Meta, website, tracking and CRM changes. You can start a change log in Changeline in a couple of minutes.