
Short answer: Google's new Search Central page says that identifying Search traffic by the HTTP Referer header “isn't exact,” and that the data in your analytics or server logs “often doesn't match 1:1 with the clicks reported in Search Console.” It names four technical reasons: privacy settings and browser configuration, redirects on your own site, different counting methods, and paid traffic mixed in with organic. In plain terms, if Search Console says 500 people clicked and your analytics says 430 visited, Google is telling you that is expected, and the useful question is whether the two lines move together.
This matters to owners who get a monthly SEO report and hear a version of “the numbers are different, but both are right.” A dental office in Auburn, a heating and cooling company in Fort Wayne and a lawn care outfit in Kendallville can all end up in the same meeting: the agency shows Search Console clicks, the owner opens Google Analytics, and the totals do not agree. Until now, much of the explanation lived in forum threads and experience. Google has now put the plumbing in writing, which makes it easier to point at one source and move on.
A note on what this post is. We did not test anything on a client site for this article, and the Google page announces no new feature. It documents behavior that already existed. We read Google's page, Google's older guide on using Search Console and Analytics together, Google Analytics help, and three trade reports. Where a source is a secondary summary, we say so. We also sell SEO and analytics services, so weigh our advice with that in mind.
Key Takeaways
- Google added the page, titled “Identify Google Search traffic in your web analytics or server logs,” on November 12, 2026, according to its own Search Central changelog. The stated purpose is to document how site owners can identify and track traffic from Google Search in their web analytics systems.
- Google Search sends a Referer that contains only the scheme and host, such as https://www.google.com/, never the path and never the search query. Queries live in Search Console.
- Google Images traffic sends the same values, and Discover traffic uses the values of whatever surface the person was on, so the referrer alone cannot separate them from web search.
- Google says Search traffic that appears as referral or direct can point to custom channel rules, referral exclusions or filters in your analytics setup, which is the one cause you control directly.
- Google's older guide says small differences between clicks and sessions can be ignored, and that the general trends should have the same pattern in both tools.
What Did Google Publish on November 12?
Google's Search Central documentation changelog for November 12 reads: “Added a new page about identifying traffic from Google Search using HTTP Referer headers.” The stated reason is “To document how site owners can identify and track traffic from Google Search in their web analytics systems.” The page sits in the Monitoring and debugging section next to the existing guides on debugging traffic drops and using Search Console.
Search Engine Roundtable reported the page the same day and quoted its opening paragraph. Barry Schwartz added a dry aside: he wondered why Google posted this “when we don't see that much traffic from Google Search these days.” That is a joke from a trade reporter, not a finding, and we are not endorsing it, but it captures a mood in the industry. Whatever the reason for the timing, the content is practical.
ppc.land read the page closely and described it as short: four sections and one table, no screenshots, no figures, and no claim about how much Search traffic a typical site loses to the gaps it describes. That is worth remembering. Google gives you a list of reasons. It does not tell you how large any of them is for a business like yours.
What the Page Covers, in One Place
- How Google sets the Referer: the origin policy, so only scheme and host are sent.
- The values to match: a short table for web, Android and iOS.
- Why the Referer is sometimes blank: four default privacy scenarios, plus tracking scripts that do not always run.
- How to match it: in analytics tools or in server logs, using only the landing page requests.
- Why your data will not match Search Console: four common technical reasons.
Which Referrer Values Mean Google Search?
When someone clicks a result, their browser or app tells your website where the click came from. That message is the HTTP Referer header, and yes, the spelling is a historical typo that the web kept. Google Search tells browsers to send only the origin, so your site sees the scheme and host and nothing after the last slash. Google's page says this means “the Referer header contains only the scheme and host of the referring page, without the URL path or query parameters.”
The values Google lists are:
- Google Search on the web: https://www.google.com/ or a local Google domain, such as https://www.google.co.uk/ or https://www.google.co.jp/.
- Google app on Android: android-app://com.google.android.googlequicksearchbox/https/www.google.com when the app opens the link in a browser tab, or https://www.google.com/ (or a local Google domain).
- Google app on iOS: https://www.google.com/ or a local Google domain.

ppc.land pointed out a detail in the Android value: it is the only one that names the app, and a log parser that looks only for strings starting with https:// would miss it. If you or a vendor analyze raw server logs, that is the line to check. The page does not say when the Android app sends the plain web value instead of the app value, other than the browser-tab case.
Two sentences under the table matter more than the table. First: “Traffic from Discover uses the same values as the surface where the user viewed Discover, such as the Google app or Google Search on the web.” Second: “Other Google services that are served from the same origins (for example, Google Images) send the same Referer values, so this traffic is also included when you match these patterns.” For a local business this means a click from Google Images and a click from a regular search result look identical in your analytics unless you use Search Console's search type filter to separate them.
One more limit, stated in a boxed note on the page: the Referer “doesn't include the search query that the user entered.” If you want to know which searches brought people in, use the Performance report in Search Console. ppc.land added the history: queries stopped traveling in Google's referrer data in 2011, when Google moved searches to encrypted connections and analytics reports filled with the label “(not provided).” The new page confirms that position.
Why Does Some Google Traffic Show Up as Direct?
Google sets its policy, but the browser decides what to send. The page says that “the user's browser or operating system might still omit the Referer header (or block JavaScript access to document.referrer) for privacy and security reasons,” and lists four scenarios where this happens by default:
- Private browsing modes and anti-fingerprinting protections in the browser.
- Opening search results in a new tab, in some browsers.
- Moving from a native mobile search widget to an external browser.
- Visits from people with privacy extensions or strict browser settings turned on.
A visit with no referrer has no source to report. ppc.land notes that most analytics tools file such sessions under direct traffic, alongside typed addresses and bookmarks. So a customer who searched for your plumbing company on a phone, tapped your result and landed on your site may show up in your report as someone who typed your web address. Google's page does not estimate how often this happens, and it names no browsers.
A second loss sits on top. The page says that “JavaScript-based visitor tracking might not log all visits because of browser privacy settings, consent management, and ad blockers.” Here the visit is not mislabeled. It is missing. If a visitor declines your cookie banner, or runs an ad blocker, your analytics tag may never fire, and Search Console will still count the click. ppc.land makes the contrast with server logs: a server records every request it answers, whether or not a script ran. That does not make logs a better number, since they include bots and asset requests, but it explains why they can read higher than a tag-based tool.
This lines up with something we wrote about earlier. In our guide to the Google Analytics reports small businesses actually need, the point was that every report depends on a tag you cannot see running. The new page is Google's own acknowledgment of that dependency for Search traffic.
What Are Google's Four Reasons the Numbers Disagree?
The final section of the new page is the one most likely to be quoted in a client report. Here are Google's four reasons, with our notes on what each means for a small site.
1. Privacy settings and client configurations
Google says browser privacy settings, extensions, ad blockers or user consent choices “can block analytics scripts, strip URL query parameters, or suppress the HTTP Referer header.” You cannot fix this on your own site, apart from running a consent banner that does not block more than it needs to. For a local service company whose customers are mostly on phones, the practical answer is to accept some loss.
2. Redirects on your site
This is the cause most in an owner's control. The page says client-side redirects, such as JavaScript or meta refresh redirects, “may replace the Referer with your own page's URL,” and that redirects that do not preserve the query string “drop URL parameters such as gclid (which is commonly used for ad-tagging).” A redirect that sends a stricter Referrer-Policy header can also remove the Referer. If your site was rebuilt and old addresses forward to new ones, or if a marketing page forwards visitors through a script, this is where Google visits can quietly become “your own site” in the report.

3. Different measurement methods
Google puts it this way: “Search Console records a click when a user selects a link in Google Search, whereas web analytics tools record a visit or session only after the destination page loads and executes the tracking code.” ppc.land draws the obvious consequence: a visitor who clicks and closes the tab before the page finishes loading is a click in Search Console and nothing in analytics, and slow pages widen the gap. By that logic, a slow mobile site can create a mismatch on its own.
4. Paid and organic traffic
The page states that Search Console “reports only organic clicks from Google Search and doesn't include clicks on Google Ads, whereas your web analytics system might include paid traffic in its totals, depending on how you configure it.” If your analytics view is filtered to Google as a source without separating the medium, your sessions can include ad clicks that Search Console will never count. We have a longer treatment of that problem in our post on why Google Ads, GA4 and your CRM never agree.
Notice what is not on the list: nothing here suggests Google is hiding clicks or that one tool is broken. Each reason describes a normal difference between counting a click on Google's side and counting a page view on yours.
What Else Does Google's Older Guide Say About Mismatches?
The new page ends by linking to an existing Search Central guide, “Using Search Console and Google Analytics data for SEO,” which has a longer table of reasons for a big gap between clicks and sessions. We read it on November 16. It lists, among others:
- Implementation in Google Analytics. The tag has to be on every page. Google gives the example of pages where the Analytics tag is missing, and notes that Search Console data is processed uniformly for all properties, so your configuration affects it less.
- Cookies or tracking. If your site asks people to accept tracking and they opt out, Google Analytics data can be skewed.
- Timezone. You can choose a timezone in Google Analytics, but Search Console's default timezone is Pacific Time and cannot be customized. Google flags this as most noticeable for sites with a big offset from Pacific. For a business in Northeast Indiana, the Eastern offset is only a few hours, so it mostly affects how a day is cut off, which matters when you compare a single day.
- Attribution. Google Analytics offers three attribution models, while Search Console counts every click in Google Search.
- Canonical URLs. Search Console reports only on the Google Search canonical URL, whereas Google Analytics reports on any URL that includes the tracking code, so you will likely see more URLs in Analytics.
- Traffic breakdowns. Search Console splits traffic by web, image, video, news and Discover, and those categories are different in Google Analytics.
- Non-HTML pages. PDFs shown or clicked in Search are included in Search Console by default, but your Analytics may not measure them.
- Bot traffic. Google Analytics automatically excludes traffic from known bots and spiders, while Search Console does not necessarily filter them out.
For a small site, the ones we would check first are the missing tag, the consent banner and the PDFs. A restaurant that links its menu as a PDF, or a contractor that posts a price sheet as a PDF, can have real Google clicks that analytics never sees, because a PDF cannot run a tracking script unless it is configured another way. That is our reading of Google's PDF note, not something we tested.
How Big a Gap Is Normal, and Which Number Do You Trust?
Google does not give a percentage, and neither will we. What the older guide does give is a rule of thumb. When comparing the tools, you will notice that even the most similar metrics, sessions and clicks, “don't match exactly,” and “what's important is that the general trends have the same pattern.” It sorts discrepancies into two groups: if the difference is small, “you can ignore” it, since the systems are different and a small gap is expected; if it is big, work through the reasons list.

Which tool wins? Google's guide is blunt: “The source of truth for Search performance will always be Search Console, while the source of truth for behavior inside your site will be Google Analytics.” Put differently, use Search Console to answer “did more people click us in Google?” and use Analytics to answer “what did they do once they arrived, and did they call or fill out the form?” Neither tool is built to answer the other's question.
Here is an illustration, and it is hypothetical, not a client result. Suppose a dental office in Auburn sees 500 Google clicks in Search Console in October and 430 organic sessions in analytics. In November, Search Console shows 560 and analytics shows 480. The totals disagree both months, but both lines went up, and the ratio is roughly stable. Per Google's guidance, that is not a problem to chase. Now suppose November analytics shows 300 while Search Console shows 560. A gap that opened suddenly points at something that changed on your side: a tag removed in a site update, a new redirect, a consent banner change. That is the sort of break worth an hour of investigation.
The trend test is also how we would approach a drop in either tool. Our post on SEO warning signs that show up before a traffic drop covers the pattern: look for a change that lines up with a date, and confirm it in two sources before you decide anything.
How Do You Check Your Own Setup in One Afternoon?
These steps are our suggestions based on Google's pages, not an official Google procedure. You need view access to Search Console and Google Analytics. If an agency or a former employee set them up, ask them for the answers in writing.
- Pick one window and apply it to both tools. Google's older guide suggests a default of the last 28 days in its template and warns that Search Console data can be delayed by a couple of days. Use the same dates, and the same country and device filters in both tools, which Google recommends for comparing apples to apples.
- Filter Analytics to the right slice. Google's Looker Studio template filters to sessions where the source is google and the medium is organic. If your report uses source alone, paid sessions may be included, which is Google's fourth reason.
- Write down the two totals and the ratio. Do it once, today, and save it. A ratio you recorded in November is the only way to tell in February whether a change is normal.
- Look at where Google visits are being filed. In Analytics, open the Traffic acquisition report and note the Organic Search and Google row, then check whether Google appears under referral or direct. Google's page says that if Search traffic is “missing or shows up as referral or direct traffic,” you should check custom channel rules, referral exclusions or filters.
- Review any custom channel rules. ppc.land observes that Google Analytics added an AI Assistant channel to its default channel group on May 13, 2026, and that every custom rule is evaluated against the same referrer data, so a pattern that is too broad could pull Google visits out of organic search. If someone built a custom rule for AI traffic, read its pattern. We covered measurement of AI visits in our GA4 AI Assistant channel guide.
- Check the referral exclusion list. In Google Analytics, referral exclusions live in the data stream settings under the tag settings, in a list called unwanted referrals. Google's help page says you can configure a maximum of 50 per data stream and that conditions use OR logic. Make sure google.com is not in that list. Google's example uses are payment processors and similar cross-domain flows, not search engines.
- Test your own redirects. Search for your business, click your result from a phone, and watch the address you land on. If your old page forwards through a script, or a redirect drops the query string, ask your web developer to switch it to a server-side redirect that preserves parameters. This is our recommendation based on Google's second reason.
- Confirm the tag is on every page, including landing pages built outside your main site. Online booking pages, quote forms on a separate service and landing pages from a third-party builder are common places for a missing tag.
- If you read server logs, count only landing page requests. Google notes that the browser sends a Referer on requests for images, CSS and scripts too, but for those the value is your own page's URL. To count Google visits, “only look at requests for the landing pages themselves.”
If you are comfortable in both tools, this takes an afternoon. If not, it is the sort of review an owner can hand to someone who is, with the list above as the brief. Our analytics services include this kind of tracking check.
What Does the New Page Leave Out?
The page is useful, and it has gaps. ppc.land went through them, and we agree with its reading.
- No mention of AI Overviews or AI Mode. Both run on Google's own search domains, and the page does not say whether a click on a link inside an AI Mode answer carries the same referrer as a standard result. ppc.land recalls that in May 2025 links in AI Mode carried noreferrer attributes that stripped referrer data, and that Google confirmed on May 23, 2025 that this was an unintended bug. We did not re-verify that history, and we cite it as ppc.land's account.
- No mention of the google.com/goto links. ppc.land reports that Google wraps redirects around some result links, first spotted on June 23, 2026, with Google confirming the rollout on August 26, 2026. The new page does not say whether a click through that redirect arrives in a different form. That is a question for your analytics vendor, not something we can answer.
- No mention of Gemini. ppc.land says Gemini stopped appending utm_source=gemini to cited links around September 28 and 29, 2026, and that visits from its iOS app have been reported to arrive with no referrer. The new page is scoped to Google Search, so the omission is consistent with its title, but it means AI referral traffic from Google's own assistant remains undocumented.
- No numbers. Google does not say what share of Search clicks reach sites without a referrer, how often the Android app sends its app-specific value, or how large the typical gap between Search Console and analytics is.
The lesson for a local business: this page makes reporting on classic blue-link traffic more defensible. It does not settle how to count the newer surfaces. If AI-driven visits matter to you, keep reading as Google documents more, and compare our earlier post on AI search referral traffic.
Why Does the Timing Matter Before Black Friday?
Reporting gets noisy at the exact moment you will want clean numbers. Google's November 2026 core update began rolling out on November 10, and ppc.land reported it may run until roughly three days before Black Friday, which falls on November 27. We went through the timing in our post on the core update landing before Black Friday.
On top of that, Search Engine Roundtable reported on November 13 that Google confirmed a data logging bug in the Search Console Generative AI performance report between November 9 and November 11, which resulted in a decrease in impressions. That affects one report, not the standard Performance report, and Google called it a logging issue. But it is a reminder that Search Console is a source of truth with its own occasional errors. Our earlier post on why rank reports can show drops that are not real makes the same point about third-party tools.
So the practical advice for the next two weeks is the unglamorous kind. Save the ratio from step three above this week, before the holiday traffic changes the mix. Do not rewrite your site or tear out a redirect in the middle of a core update rollout unless it is broken. And if the report you receive in early December shows a gap that looks different from your November baseline, you will know what to check and in what order.
What Do We Still Not Know?
Plenty, and it is better to say so.
- How large the gaps are. Google gives four reasons and no figures. We did not measure the gap on any site for this post.
- How AI Mode and AI Overviews clicks behave. The page does not address them, so we cannot tell you how they appear in your analytics.
- Whether the page will grow. The changelog entry does not say whether Google plans to add guidance for newer surfaces.
- What your own setup is doing. Everything above is general. Whether your analytics tags, consent banner or redirects are causing a gap is something only a look at your site can answer.
- Whether the ppc.land and Relevant Audience readings are complete. We read most of ppc.land's article and the opening of Relevant Audience's summary, not every section, and both are secondary to Google's own page.
So the advice stays modest. Record both numbers this week, check where your Google visits are being filed, test your redirects, and judge the trend, not the total. If Google extends the page, you will already have a baseline.
Want Someone to Check Why Your Two Reports Disagree?
Button Block reviews analytics and Search Console setups for Northeast Indiana businesses. We can look at your tag coverage, channel rules, referral exclusions and redirects, save a baseline ratio and explain in plain language what each report can and cannot tell you. We sell these services, so weigh our advice with that in mind.
Frequently Asked Questions
- Why does Search Console show more clicks than Google Analytics shows organic visits?
- Google says the two tools count different things. Search Console records a click when someone selects your link in Google Search. Analytics records a visit only after your page loads and the tracking code runs. A person who clicks and leaves before the page loads, a visitor who declines tracking, and a browser with an ad blocker all produce a Search Console click and no analytics session. Google calls a small gap normal and says the trend should look similar in both tools.
- Can I see the search words people typed in my analytics?
- Not from the referrer. Google's new page states that the Referer header does not include the search query the user entered, and it points to the Performance report in Search Console for queries. That has been the practice for years, and the new page confirms it in writing.
- Why is some of my Google traffic filed under Direct?
- A visit with no referrer has no source to report, and most analytics tools file those under direct traffic. Google lists several ordinary reasons the header goes missing: private browsing and anti-fingerprinting protections, opening a result in a new tab in some browsers, moving from a native mobile search widget to an external browser, and privacy extensions or strict browser settings. Nothing on your site necessarily caused it.
- Can I tell Google Search clicks apart from Google Images and Discover clicks by the referrer?
- No. Google's page says other services served from the same origins, such as Google Images, send the same Referer values, and that Discover traffic uses the same values as the surface where the person saw it, such as the Google app or Google Search on the web. Search Console is the only place Google itself breaks those apart.
- Do I need to change anything on my website because of this page?
- The page documents existing behavior and announces no new feature, so nothing is required. It is worth checking three things: that no custom channel rule, referral exclusion or filter is reclassifying Google visits, that your site does not use client-side redirects that overwrite the referrer, and that redirects keep the query string. We cover how to check each below.
- Which number should I report to the owner or the board: Search Console clicks or analytics sessions?
- Google's own guidance says the source of truth for Search performance is Search Console and the source of truth for behavior inside your site is Google Analytics. For a monthly report, we would show both, say which is which, and watch the trend lines. If the two lines move in the same direction, the gap in the totals is rarely worth a meeting.
Sources & Further Reading
- Google Search Central, “Identify Google Search traffic in your web analytics or server logs” (published November 12, 2026): developers.google.com (read November 16, 2026): the Referer values, blank-referrer scenarios, matching guidance and the four reasons for a mismatch, all quoted from the full page.
- Google Search Central, “Latest documentation updates” (November 12, 2026 entry): developers.google.com (read November 16, 2026): the changelog wording and stated purpose.
- Google Search Central, “Using Search Console and Google Analytics data for SEO”: developers.google.com (read November 16, 2026): the clicks versus sessions comparison, the source-of-truth statement, the Looker Studio filter, and the longer discrepancy table. We read most of the page, not every section.
- Google Analytics Help, “Identify unwanted referrals”: support.google.com (read November 16, 2026): how the referral exclusion list works, the 50-per-stream limit and OR logic. The page returned incomplete content, so we cite only what loaded.
- Search Engine Roundtable, “New Google Help Doc Explains How To Identifying Traffic From Google Search” (November 12, 2026): seroundtable.com (read November 16, 2026): the same-day report with the changelog quote and Barry Schwartz's aside.
- Search Engine Roundtable, “Daily Search Forum Recap: November 13, 2026”: seroundtable.com (read November 16, 2026): the summary of the Search Console Generative AI performance report logging bug for November 9 to 11. We read the recap, not the linked story.
- ppc.land, “Google explains why analytics won't match Search Console click counts” (November 12, 2026): ppc.land (read November 16, 2026): the detailed reading of the page, the Android value, the omissions about AI Mode, goto links and Gemini, and the 2011 history. Secondary source; we read most of the article.
- Relevant Audience, “Identify Google Search Traffic in Analytics and Logs” (November 13, 2026): relevantaudience.com (read November 16, 2026): a secondary summary that matched Google's page on the Referer values, the Images and Discover point and the four reasons. We read the opening portion.
- ppc.land, “Google's core update may run until three days before Black Friday” (November 11, 2026): ppc.land (surfaced in search results November 16, 2026): the rollout window used in the timing section. We relied on the search snippet and our earlier post, which read the primary sources.
