Running ads · Updated October 4, 2026
Why is my conversion rate different in Google Ads and in Analytics?
Updated October 4, 2026
Because they divide by different things and credit conversions differently. Google Ads divides conversions by ad clicks, dates them by the click and credits the ad. Analytics divides by sessions or users, dates them by when they happened and can credit another channel. Compare conversion counts, not rates, and bid on the Google Ads count.
If the two rates differ, compare conversion counts instead
If Google Ads and Analytics show different conversion rates for the same campaign, stop comparing the rates. They divide by different things, so they would differ even if both counted exactly the same leads. Compare the number of conversions, in the order below, then decide which number to act on.
Google Ads divides by ad interactions. Google defines its conversion rate as "The average number of conversions per ad interaction, shown as a percentage." On a Search campaign that mostly means clicks. The top of the fraction is the Conversions column: every primary action, including calls from your ad. With more than one action, or a count of Every, Google says the rate "might be over 100%."
Analytics divides by sessions or by users. Session key event rate is the share of sessions in which a key event happened. User key event rate is the share of users who triggered one. Someone who sends your form twice in one visit is still one converting session.
Neither number is wrong. One says how often a paid click turned into a counted conversion. The other says how often a visit to your website did. Most of the gap between them is calls, dates and credit, and each of those can be lined up.
For example
In April, a roofer's Search campaign gets 500 clicks. Google Ads reports 30 conversions: 20 form fills and 10 calls from the call button in the ad, so a 6% conversion rate. Analytics shows 450 google / cpc sessions with 18 form fills, a 4% session key event rate, and 400 users, so a 4.5% user key event rate. The 12 conversions Analytics lacks: the 10 callers never opened the website, one person clicked on April 28 and sent the form on May 2, which Analytics files under May, and one came back through an unpaid Google result before sending it, so Analytics credits Organic Search. The same 20 form fills sit behind all three rates, once calls, dates and credit are separated out.
Bid on the Google Ads count, and judge it on booked jobs
If you have to pick one number to run the campaign on, use the Google Ads Conversions column, because it is the one bidding works from. Use Analytics to judge the website and your other channels, and your own records to judge the business.
- Google Ads conversion rate
- Conversions from your primary actions per ad interaction, filed under the day of the click and credited to your ads.
- It is what Smart Bidding reads. Google says the data in the Conversions column is used for your bidding strategies, so this is the count a fix has to make right.
- Analytics key event rate
- Sessions or users with a key event, filed under the day it happened, across every channel that brings visitors.
- It shows what the site does with visitors. Use it to compare landing pages and channels, not to second-guess what the campaign bids on.
Use it for the campaign
Use it for the website
Then hold both against your inbox and call log. Neither tool knows which form fill became a job. Count the leads from ads that booked work, and divide ad spend by them. If cost per booked job holds steady, a gap between the two rates is a reporting question, not a performance problem. That is a judgment, but it is the one that keeps you from rebuilding tracking that works.
If Google Ads counts more conversions than the leads you actually received, that is a different problem with its own fixes, covered in "Why does Google Ads show more conversions than I actually got?"
Line them up, in this order
If the dates are recent, compare a finished month instead
Compare a full month that ended at least a week ago, with the same dates set in both tools. The last few days of any range are still filling in, and not at the same speed.
Google says to allow 24 to 48 hours before comparing Google Ads with Analytics, and up to 7 days for conversion data to fully appear after big changes or after linking accounts. Analytics takes 24 to 48 hours to process its own data. A key event imported into Google Ads arrives up to 24 hours after it happened, and modeled conversions can take up to 5 days to settle.
Then check the time zones. Analytics reports in the property's time zone and Google Ads in the account's, and Google lists a mismatch as a cause of discrepancies. Find the Google Ads one under Admin, then Preferences. Google Ads lets you reset it only once, only eastward, and only through an Admin user of a manager account. The Analytics property's time zone can be changed in its settings, from that day forward, so if they differ, change the Analytics one to match. Over a whole month it only moves the conversions near midnight on the first and last day; it matters when you compare single days.
Switch Google Ads to the date each conversion happened
Add the Conversions (by conv. time) column before you compare anything else. Google's regular conversion columns file each conversion under the day of the click. Analytics files it under the day it happened.
Google's own example: a click on July 19 and a purchase on July 20 count on July 19 in Google Ads and on July 20 in Analytics. A lead that comes in a week after the click moves a week, and around the first of the month that moves it into another month.
There is no conversion rate column by conversion time. Google's by-conversion-time columns cover conversion counts and values, not a conversion rate. If you want a rate, divide the by-time count by clicks yourself, and treat it as rough, because those conversions came from clicks in more than one month.
Click-date months keep growing. A conversion is recorded until the action's window closes, 30 days after the click by default, so last month in the regular column can still rise. Compared too early, it looks low against Analytics.
Why Google's pages name the date three different ways
One Google Ads page says conversions are reported by the date of the click, another by "query date", a third by "the ad impression date". On a Search ad all three are normally the moment someone searched and clicked. What matters is that none of them is the day the lead came in.
If Google Ads counts calls, take them out of the comparison
Compare form fills with form fills. If the Conversions column includes calls from your ads, separate them before you hold it up against Analytics.
A call from the call button in your ad goes straight to your phone. The caller never loads your website, so Analytics never sees them. Once call reporting is on, Google includes its Calls from ads action in the Conversions column automatically. For a plumber or roofer, where many leads call, that alone can put the Google Ads rate well above the Analytics one.
- In Google Ads, click Campaigns, then Segment above the table, then Conversions, then Conversion action.
- Read the Conversions (by conv. time) figure on your form action's row, the column added above. That is the number to compare with Analytics.
- Read Clicks from the campaign's own row. Google blanks clicks, impressions and cost on the segmented rows, so a rate per action has to be worked out by hand.
Check whether Google Ads bids on its own tag or an Analytics import
Click Goals, then the Conversions dropdown, then Summary, and check where the primary form action comes from. A website tag means Google Ads counts the form itself. Google Analytics means the count is an Analytics key event imported into Google Ads. The two cases explain different gaps.
An import starts from the event Analytics counts. Google says counting method and conversion window match by default. What separates the counts afterward is Google Ads' side: it keeps only conversions that follow a click from that account, drops what its invalid click systems reject, may add modeled conversions, and never brings in history from before the import.
A Google Ads tag counts on its own. It has its own trigger, its own count setting (website actions default to Every) and its own window, 30 days by default. If it fires on a thank-you page while Analytics counts a form submit event, the two can drift apart in ways neither report shows. And Google says Google Ads conversions don't appear in Analytics' standard reports, only in its Advertising section, so in Traffic acquisition what you compare it with is Analytics' own key event.
If both are primary, one form fill counts twice in Google Ads and its rate roughly doubles against Analytics. Set one of them to Secondary; the conversions page named above walks through choosing which.
Why an action you created in Analytics is missing from the Conversions column
Google sets conversion actions created from the Analytics side, under Advertising, then Conversion management, to Secondary. They record in All conv. but not in Conversions, and bidding ignores them until you change the action's optimization to Primary in Google Ads. Google Ads also controls that setting alone; Analytics cannot change it.
In Analytics, read only google / cpc sessions in Traffic acquisition
Open Reports, then Acquisition, then Traffic acquisition, switch the table's first column to Session source / medium and search for google / cpc. Then set the Key events column to your lead event instead of all events, and compare that count. Google offers this filter, or Analytics' Advertising section, for a fair comparison with Google Ads. The same Google passage also says to use event-scoped reports, which this one is not, so pick one of the two and keep using it.
This report credits the visit, not the ad. Its channels are session-scoped, and Google says session attribution stays non-direct last click whatever attribution model you pick. Someone who clicks your ad, leaves, and types in your address within the lookback window, 90 days by default, still counts as google / cpc. Someone who comes back through an unpaid Google result counts as organic, while Google Ads still credits the ad.
The Advertising section can show a third number. Its reports use the property's reporting attribution model, which can split one key event across several channels, and Google says a change to that model rewrites past data as well as future data. Pick one Analytics report for the comparison and keep using it.
If the counts still differ by more than about 20%, check credit and counting
Google says discrepancies of "often up to 20%" are expected from attribution differences alone, and another Google page says they can occur even with a correct setup. Stopping at 20% is a rule of thumb, not a Google threshold: if your form counts sit within it after the steps above, stop there. If not, check these settings one at a time.
Typical gap, per Google
20%
Google says discrepancies of often up to 20% are expected from attribution differences alone.
- Which channels can take credit. Google Ads reports default to Google paid channels, Analytics reports to Paid and organic. Under Google paid channels, a lead whose last click was an unpaid search still goes to the ad. Analytics shows its setting under Admin, then Data display, then Events, then Attribution settings.
- Window against lookback. A Google Ads tag stops counting 30 days after the click by default. Analytics looks back 90 days for most key events. A lead that comes in 45 days after the click is not recorded by the Google Ads tag, while Analytics can still credit it to that click.
- Counting method. Google Ads website actions default to Every, so a lead action should be set to One. Analytics recommends Once per event, but key events carried over from old Universal Analytics goals default to Once per session.
- A lost click ID. Google says the click ID must not be altered or dropped by your website, for example by a redirect. Without it, an imported key event cannot be tied back to the ad.
- A cookie banner, on a busy account. With a banner that blocks Google's tags until someone accepts, a visitor who declines sends nothing at all. Both tools can estimate those missing conversions, but only above a volume threshold: Google Ads needs 700 ad clicks over 7 days per country, and Analytics needs the advanced setup and at least 1,000 daily users who accept, among other conditions. If your account clears one threshold and not the other, the one that models will count higher. A small local campaign usually clears neither, so declined visitors are missing from both counts.
Why the click count and the session count never match
Google lists the reasons. Someone who clicks your ad twice in one visit is two clicks and one session. Someone who clicks once and returns later from a bookmark is one click and two sessions. Someone who clicks and leaves before the page loads is a click with no session. And Google Ads removes invalid clicks from its reports, while Analytics still shows those visits.
They are why the rates can differ even when the conversion counts agree.
Why older guides, and some of Google's own pages, contradict this
Much of what ranks for this question describes Universal Analytics, which Google replaced with Google Analytics 4: goals counted once per session and last non-direct click credit. Even the current Google Ads help page on Analytics conversions links to a comparison page that Google labels "[UA] ... [Legacy]".
Google's current pages also disagree with each other. One Google Ads page says Analytics "often uses last-click" and, further down, that it "defaults to data-driven attribution." Read the setting in your own property under Attribution settings instead of trusting either line.
Before you change a setting
If Google Ads matches your own lead count, leave the tracking alone
Count the form fills and calls from ads in your own records for the same month. If the Google Ads form count is close to it, the gap with Analytics is definitions, and the tracking that steers bidding is doing its job. Change nothing.
Each change costs bidding time. Google says frequent changes to budgets, targets or conversion goals reset the standard 7 to 14 day learning period. Changing a count setting, a window or a Google Ads attribution model applies only going forward, and changing Analytics' reporting model rewrites its past data, so either way the comparison you just made stops lining up with next month's.
Fix it if Google Ads counts more than you received, or if Analytics shows form fills from google / cpc that Google Ads lacks entirely. The first is double counting, covered on the conversions page. The second points at the Google Ads tag or a lost click ID, and is worth fixing before anything else, because those are leads bidding never learns from.