Google Search Console changed three times in October 2026, and if you only open it once a week, you probably missed all three. Google shipped a multi-select country filter, broke and then fixed the Performance report’s custom date picker twice in two days, and kept expanding the AI reporting that now includes a pilot that actually pays some publishers. None of these changes will transform your traffic overnight. But each one changes something small about how you read your own data — and the date picker bug, in particular, is a reminder not to treat Search Console as your only copy of your numbers.
Here is what changed, what it means in practice, and what to check this week.
1. The country filter finally lets you select more than one country
On October 6, 2026, Google announced that the Country filter in the Search Console Performance report now supports multi-select. Until then, your options were one country at a time, or a comparison of exactly two countries side by side. If you wanted a regional total — say, the United States plus Canada, or all of your English-speaking traffic — you exported the data and grouped it in a spreadsheet. Now you tick the checkboxes and Search Console returns one combined set of clicks, impressions, CTR, and average position for the whole group.
This sounds minor. It is not, if you run a site with a genuinely international audience or one where a handful of countries drive most of the revenue. Here is how I would actually use it.
Filter vs. Compare: two different tools
The Performance report now has two ways to work with countries, and they answer different questions:
- Filter tab: select several countries and see their combined traffic as one total. Use this when you want a regional number — “North America,” “English-speaking markets,” “everywhere I sell.”
- Compare tab: still handles two countries at a time and adds a Difference column. Use this when you want to see how one market behaves against another — “Is my CTR problem a US problem, or is Canada the same?”
Before this change, the Compare tab’s two-country limit was the only built-in way to look at multiple markets, which is why so many publishers ended up in Excel. The Filter tab removes that busywork for the common case.
The practical move: match it against your earnings
For a US-focused blog, the most useful version of this feature is not traffic for its own sake. It is traffic matched against money. Open the Performance report, filter to the United States plus Canada (or whatever your top two or three markets are), and note the clicks and CTR for the same date range. Then open your AdSense Countries report for the same dates and compare RPM by country.
That comparison answers a question most publishers get wrong: is a low RPM a geography problem or a monetization problem? If US visitors earn far more per thousand impressions than visitors from elsewhere, your “low RPM” may really be a traffic-mix issue, and the fix is attracting more US readers, not redesigning your ad layout. If RPM is low everywhere, the problem is on your pages. Search Console shows where your Google clicks come from. AdSense shows what visitors from each country earn you. Put the two side by side and the answer usually gets obvious fast.
Two honest caveats
First, Google’s announcement did not spell out whether every property gets the change at the same time, and the Search Console help page still describes the older “compare two countries” workflow. If you open your report and still see the old single-select behavior, you may be on the older version — check again in a few days rather than assuming you misread something.
Second, Google has not explained whether grouping countries changes how its privacy filtering works. Search Console already suppresses very small slices of data for privacy. A grouped view should be fine, but if a small-country combination returns less data than you expect, that is a plausible reason why. Test it on your own property before building a reporting process around it.
2. The custom date picker broke twice in two days
This one was the most visible of the three changes, because it broke something publishers use constantly.
On the morning of October 13, 2026, the custom date range picker in the Performance report stopped working. The end date got stuck in 1969 — specifically, trying to change it produced an error along the lines of “Date can’t be after 1969-12-31.” One user who flagged it publicly, Punit on X, saw the end date showing 1970-01-01 instead. The preset ranges (24 hours, 7 days, 28 days, 3 months, 6 months, 12 months, 16 months) kept working; only the custom range was broken.
Google fixed it the same morning, around 9:10 AM ET on October 13. Then, the next morning, it broke again. Search Engine Roundtable’s Barry Schwartz reported on October 14 that the date picker was broken a second time, and that it was fixed again by about 9:45 AM ET that day. Two outages in two days, each fixed within a few hours — annoying rather than catastrophic, but exactly the kind of thing that derails a reporting deadline if it lands on the wrong morning.
Why a date picker bug matters more than it sounds
Most real SEO questions need a specific window, not a rolling one. You want to compare the two weeks before and after a content update, or check how your site looked during a particular algorithm update. Presets cannot do that cleanly. Month-over-month and year-over-year checks also depend on exact start and end dates. When the picker locks up, those comparisons stall, and client reports or weekly reviews get delayed.
The 1969 date is a tell, by the way. Dates in computing are often stored as seconds since January 1, 1970 (the Unix epoch). When a date field fails to load properly, it can fall back to zero — which displays as December 31, 1969, or January 1, 1970, depending on timezone. It is a classic symptom of a broken date control, not a sign that your data was damaged. Your underlying Search Console data was fine throughout both outages; only the picker was broken.
What to do when it happens again
Three workarounds kept working while the picker was down, and they are worth knowing for the next time any part of the interface glitches:
- Use the closest preset range (28 days or 3 months) and read the daily numbers from the chart. It is imprecise, but it gets you an answer the same day.
- Export the performance data and filter it to your exact dates in Google Sheets or Excel. The export path does not depend on the picker.
- Pull the data through the Search Console API, where you pass your own start and end dates in the request. Dashboards built on the API, such as Looker Studio reports, have their own date controls and were not affected.
The bigger lesson: keep your own copy
Search Console only keeps 16 months of data. A small interface bug can block access to it exactly when you need it. Publishers who already export their data on a regular schedule lost nothing here.
Set up a monthly export habit. Once a month, export your queries and pages, save the file with the month in its name, and keep a separate sheet for your key landing pages. It takes a few minutes, it protects you from short outages like this one, and it gives you history beyond the 16-month window. If the custom range still looks stuck for you after a fix, refresh the page or reopen Search Console in a new tab — a cached page can keep showing the old behavior for a while. If you have ever dealt with a Search Console performance report delay, you already know the interface is not the most reliable place to keep your only copy of your numbers.
3. The AI reports: what you can see, and what some publishers get paid
The third October change is really two related things, and publishers keep confusing them. Let me separate them clearly, because the difference matters: one is a report everyone can use, and the other is a payment pilot almost nobody is in.
The Generative AI performance report: impressions, not clicks
On August 31, 2026, Google finished rolling out a Generative AI performance report inside Search Console to every website worldwide, along with a Search generative AI control that lets a site remove itself from those AI features altogether. Google had announced the reports in June and tested them on a subset of sites before the global release.
The report is simple. It counts how often links to your site appeared inside AI Overviews and AI Mode, and separately inside the AI features in Google Discover. You can break those impressions down by page, by country, by device, and by date. If you run a small business site, you can now open Search Console and see which of your pages Google’s AI is choosing to cite, and whether that number is growing or shrinking.
What the report does not show matters just as much:
- No query dimension. You cannot see what people asked when your page turned up.
- No click data. It counts impressions only, so you cannot see how often an AI citation turned into a visit.
- No separation of AI Overviews from AI Mode. Both are combined into one impression count.
There is also no AI Overview or AI Mode filter in the main Performance report or in the Search Console API. The branded-queries filter, which splits branded from non-branded queries, sounds related but is not — it does not isolate AI Overview data. If your site has too few AI impressions, or if you opted out through the Search generative AI control, you may not see the report at all.
The practical way to use it is as a citation check. Look at which pages get AI impressions, compare them against your total impressions per page in the regular Performance report, and ask whether the pages Google’s AI likes are the ones you want it to like. If a thin page is getting cited while your best guide is not, that is a content-quality signal worth acting on. And if your AI impressions are climbing while your clicks are flat or falling, that is the AI-search traffic squeeze in your own numbers — worth knowing, even if there is nothing you can do about it directly.
The AI Contribution Pilot: real money, invitation-only, wildly uneven payouts
Separate from the report, Google is testing paying publishers when their content significantly shapes AI-generated responses in Gemini, AI Overviews, and AI Mode. The pilot, known as the AI Contribution Pilot, is invitation-only and reportedly covers about 100 publishers. Publishers are paid based on how significantly their content contributes to an AI-generated answer, rather than simply because their work is cited.
The payouts, according to reporting by The Information relayed by The Verge and TechRepublic, vary dramatically. One publisher that joined early has reportedly earned more than $1 million annually. Another newer participant has received about $50,000 to $60,000. Several smaller publishers have reportedly earned less than $1,000 over several months.
Read that spread carefully before you get excited or angry. This is not a royalty program with a published rate. It is a pilot, invitation-only, with opaque methodology, and the early numbers suggest it rewards publishers whose content already shapes a large number of AI answers — which tends to mean larger or more established sites. A small publisher earning under $1,000 over several months is not being cheated by a published formula; there is no published formula. The honest read is that Google is testing whether it can pay for AI training-and-answer value without committing to a model, and the precedent matters more than any individual payout.
For most publishers reading this, the practical takeaway is threefold. First, do not wait for an invitation that may never come — the Generative AI performance report is the tool you actually have, and it is free. Second, the same work that improves your AI impressions (clear, citable, genuinely useful content with named facts and direct answers) is the work that would make you a plausible pilot candidate anyway. Third, keep your expectations honest: if you want the full picture on how the pilot works and what publishers are actually seeing, our earlier deep dive on the Google AI Contribution Pilot covers the payout reports in more detail.
What to check this week
If you have ten minutes, here is the October Search Console checklist:
- Open the Performance report and tick two or three countries in the Country filter. Confirm multi-select works on your property, and save a regional view you will actually use — your top markets grouped together.
- Test the custom date picker. Pick an arbitrary window, like September 1–15, and confirm it loads. If it still looks stuck, hard-refresh or open Search Console in a fresh tab before assuming it is broken again.
- Open the Generative AI performance report. If you do not see it, check whether your site has enough AI impressions to qualify, or whether the Search generative AI control is set to exclude your site. Note which pages get cited.
- Export this month’s queries and pages and save the file with the month in its name. Start the monthly-export habit now, while the date picker bug is still fresh in your memory.
- Do not chase the AI Contribution Pilot. There is no application. If Google invites you, evaluate the terms then. Until then, the report is your tool.
What did not change
A few things are worth stating plainly, because October’s news cycle blurred them. Search Console did not gain an AI Overview clicks filter. The API did not gain a generative-AI dimension. The date picker bug did not damage anyone’s data. And the AI Contribution Pilot did not become open to everyone. October’s changes were a useful filter upgrade, a two-day interface bug, and continued expansion of AI reporting that still shows impressions, not clicks. That is a solid month of tooling news — just not the revolution some headlines implied.
The bigger pattern is the one to watch. Google is slowly giving publishers visibility into how AI features use their content — first impressions, now (for a few) payments. The publishers who benefit most from that visibility will be the ones whose content is worth citing in the first place: specific, verifiable, and written to answer a real question. That part has not changed at all, and it is still the highest-leverage thing you can do.
Frequently Asked Questions
When did the Search Console multi-select country filter launch?
Google announced the multi-select Country filter on October 6, 2026. The Performance report’s Country filter now lets you tick several countries and see their combined clicks, impressions, CTR, and average position as one total. The older Compare tab still handles two countries at a time with a Difference column.
Why did the Search Console date picker show 1969?
The custom date picker broke on October 13, 2026, with the end date stuck at 1969-12-31 (or 1970-01-01 for some users) and an error saying the date could not be after 1969-12-31. That date is a classic software fallback: dates are often stored as seconds since January 1, 1970, so a failed date control falls back to zero, which displays as late 1969 or early 1970 depending on timezone. Your underlying data was not damaged.
How many times did the Search Console date picker break in October 2026?
Twice. It broke on the morning of October 13, 2026, was fixed around 9:10 AM ET that day, broke again on the morning of October 14, and was fixed again by about 9:45 AM ET on October 14. While it was broken, the preset ranges, data exports, and the Search Console API all kept working.
Does Search Console show AI Overview clicks?
No. As of October 2026, there is no AI Overview or AI Mode filter in the Performance report or the API. The Generative AI performance report shows impressions only — how often your links appeared inside AI Overviews, AI Mode, and AI features in Discover — broken down by page, country, device, and date. It does not show queries or clicks, and it does not separate AI Overviews from AI Mode.
What is the Google AI Contribution Pilot, and can I join?
The AI Contribution Pilot is an invitation-only test, reportedly covering about 100 publishers, that pays when a publisher’s content significantly shapes AI-generated answers in Gemini, AI Overviews, and AI Mode. There is no public application. Reported payouts vary enormously — from over $1 million a year for one early participant to under $1,000 over several months for several smaller ones — and Google has published no rate formula.
How long does Search Console keep my data?
Search Console keeps 16 months of data. That is one more reason to export your queries and pages monthly and keep your own copies: an interface bug, a reporting delay, or the 16-month cutoff can otherwise leave you without the history you need.