How to compare date ranges in Google Search Console (and read the results)

Most Search Console questions are comparisons in disguise. Is this week worse than last week, is this spring worse than last spring, did the redesign help? Compare mode in the Performance report answers all of them, but only if you pick the right two periods and know which numbers to trust.

Quick answer: To compare date ranges in Google Search Console, open Performance > Search results, click the Date filter, choose the Compare tab, and pick a preset or set two custom ranges. The chart then shows both periods together, and the table lists each period side by side with a difference column. For most traffic questions, start with the last 3 months versus the same period last year, then sort the Pages table by Clicks Difference to see which pages drove the change.

#What Compare mode shows you

Compare mode lives in the Performance report. You choose two date ranges, and Search Console lines them up:

• The chart plots both periods on the same axis, so you can spot the day the lines start to separate.
• The metric totals for clicks, impressions, average CTR and average position cover both periods.
• The table gives one column per period for each metric you have switched on, plus a difference column you can sort by.

Any filter you add, such as a country, device, page or query, applies to both periods, which keeps the comparison fair.

#How to compare date ranges in Search Console, step by step

Google adjusts the interface from time to time, so labels may differ slightly, but the flow has stayed the same for years.

#1. Open the Performance report

Select your property and go to Performance > Search results. The search type defaults to Web, so change it first if you need image, video or news data.

#2. Open the date filter and choose Compare

Click the Date filter, then choose the Compare tab (the other tab filters to a single range).

#3. Pick a preset or set custom dates

The presets cover the common cases, such as the last 3 months compared with the previous period or with the same period last year. Use the custom option for exact dates, such as the four weeks before and after a site launch, then apply the filter.

#4. Switch on all four metrics

Click the metric boxes above the chart to toggle clicks, impressions, average CTR and average position. Seeing all four stops you from reading a whole story into clicks alone.

#5. Read the chart before the table

Look for where the two lines diverge. A gap that opens on a single day points to a single event: a deployment, an outage, a ranking update. A gap that widens slowly over weeks points to something gradual. Lines that move together with the same weekly rhythm usually mean the change is noise.

#6. Break the change down in the table

Work through the table tabs: Queries, Pages, Countries, Devices, Search appearance and Dates. On the Pages tab, sort by Clicks Difference so the pages that lost the most clicks come first. Google recommends exactly this in its guide to debugging drops in Google Search traffic.

#Which date ranges to compare for which question

The comparison you choose shapes the answer you get. Week over week can make a normal seasonal dip look like a disaster, and year over year can hide a problem that started last Tuesday.

Before any of these, Google suggests viewing the full 16 months as a single range to spot yearly patterns. If last year shows the same dip in the same month, it's probably seasonal.

#Lining up a comparison with a Google update

Rollout dates are public on the Google Search Status Dashboard. Say you want to know whether the May 2026 core update affected your site. It started on May 21, 2026 and took about 12 days to roll out, so compare a window that begins once the rollout finished in early June with an equal-length window that ends before May 21.

If the after window is clearly lower, the decline began during the rollout, and the same weeks last year look normal, the update is a likely suspect, though not a proven one. Our guide to Google core update recovery goes deeper.

#Rules for a fair comparison

Most misleading comparisons break one of these rules. Run through them before you trust a number.

1. Use equal-length ranges. A 31-day month against a 30-day month gives the longer side a free extra day of clicks.
2. Match weekdays. Most sites have a weekly rhythm, so compare whole weeks (7, 14 or 28 days). Calendar months contain different mixes of weekdays.
3. Leave out days that are still filling in. Google's help page on troubleshooting data discrepancies says performance data "is usually available in 2-3 days." If the newest days are incomplete, the current period looks artificially low, so end custom ranges a few days back.
4. Remember the time zone. Daily data is recorded in Pacific Time (the 24-hour view is the exception), so a Search Console "day" may not match your local day or your analytics tool.
5. Watch moving holidays. Year over year, the same dates fall on different weekdays, and holidays like Easter or Thanksgiving can land in different weeks.
6. Separate brand from non-brand. A brand campaign can hide a non-brand loss. If your property has the branded queries filter, check each segment separately (its AI-assisted classification can mislabel some queries).
7. Respect the 16-month limit. The interface keeps 16 months of data, so year-over-year comparisons only work for recent periods.

#How to read the results without fooling yourself

#Look at clicks and impressions together

If impressions held steady but clicks fell, you are still being shown, but fewer people are clicking. Google's advice for that pattern is to review your titles and snippets and to check whether competitors now have richer results. If clicks and impressions fell together, you have lost visibility, through rankings, indexing or lower search demand.

#Give average position some context

Small moves are normal. In the same debugging guide, Google describes a slip from position 2 to 4 as fluctuation that often recovers without any action, and advises against radical changes to pages that already perform well. A large drop across many terms, such as falling from the top 10 to around position 29, is different. Google suggests a site-wide self-assessment of how helpful and reliable your content is.

Average position is averaged across every impression. If your site starts appearing for lots of new queries at position 40, the average gets worse even though nothing you already ranked for moved. Check position query by query before you worry.

#Expect the table not to add up to the chart

The Performance table shows at most 1,000 rows, and rare queries are anonymized: left out of the query table but still counted in the chart totals. So the differences in the Queries tab will rarely sum to the change in total clicks. We explain why in why Search Console data looks wrong.

#Use annotations to remember what changed

A comparison is only as useful as your memory of what happened between the two periods. Since November 17, 2025, Search Console has supported custom annotations on the Performance chart, as PPC Land reported at launch. Right-click the chart and choose "Add annotation" to pin a note to a date.

The rules worth knowing:

• Each note can be up to 120 characters, and a property can hold up to 200.
• Annotations can't be edited, so write them carefully.
• They're shared with everyone who has access to the property.
• They're removed after 500 days, roughly the same span as the 16 months of data you can see.
• They are not shown in comparison mode or in the 24-hour view.

That last point shapes the workflow. Add an annotation whenever something changes: a template update, a migration, a batch of title rewrites, the start of a Google update. When you investigate later, open the full 16 months in the normal view, read the annotations to find the dates that matter, then switch to Compare and build custom ranges around them.

#From comparison to action

A comparison tells you where and when traffic changed. It doesn't tell you why. Once you know which pages or queries moved, Google's guidance points to the next checks:

• A site-wide drop: open the Page indexing report.
• A group of pages: run URL Inspection on a sample of them.
• Anything sudden: check the Security issues and Manual actions reports.
• A slow, steady slide on individual pages: treat it as content decay and plan a refresh.

For the full process, see why did my Google traffic drop, or get a fast first look with the free traffic drop checker. And when you change something, give it time. Google notes that some changes show within days while others can take months, so wait a few weeks before comparing again.

If you'd rather not build these comparisons by hand every week, RankSlide runs one for you daily. It compares each page's last 7 days with the average week of the previous 28 days, flags drops of 15%, 30% and 50% on pages with at least 20 weekly clicks, and explains each one in plain English. You can try the demo with sample data.

#FAQ

#Can you compare more than two date ranges in Search Console?

No. Compare mode shows two ranges at a time. For a longer view, look at the full 16 months as a single range, or export the data and compare several periods in a spreadsheet.

#Why can't I see my annotations in compare mode?

Search Console doesn't show custom annotations in comparison mode or in the 24-hour view. Switch back to a single date range to read them, note the dates, then return to Compare.

#What's the best comparison for a sudden traffic drop?

Start with the last 7 days against the previous 7 days to confirm the drop and find the day it began. Then check the last 3 months against the same period last year to rule out seasonality.

#How far back can I compare in Search Console?

The interface keeps 16 months of data, so year-over-year comparisons only work for fairly recent periods. To keep a longer history, export your data regularly or pull it through the Search Console API.