← Blog
SEO · November 22, 2026 · 8 min read

Reading Search Console After a Sale Event: Separating the Black Friday Spike From Real Ranking Changes, and the Pages That Earned a Refresh

D
Reading Search Console After a Sale Event: Separating the Black Friday Spike From Real Ranking Changes, and the Pages That Earned a Refresh

The week after Black Friday, every chart in Search Console is lying to you in a different way. Positions moved because the searchers changed, not the pages. CTR fell because the results pages filled with shopping units. A guide that "jumped to position three" was shown for deal queries it will never see again.

And underneath all of it there are three things worth knowing, which the spike hides unless you read the report with the event taken out.

This post covers why the week after a sale event is the worst time to judge any page, how to read the Performance report with the event excluded (the same-week-last-year comparison, the before-and-after windows, the query filters), the three signals that are real underneath the spike and how to find each, the January list the reading produces and in what order, and the one kind of change that doesn't wait.

The Search Console post covers the reports in general; this is the one week of the year when they need a different reading.

The short answer

Don't read rankings in the event week or the week after. Compare the event week with the same event week last year for the spike itself, and the four weeks after the event with the four weeks before for rankings, with deal and sale queries filtered out.

Look for three things: new queries a guide started appearing for, page-two guides that climbed and held, and guides that got the traffic but not the conversions. Write those into a January list, ranked by wasted traffic, and change nothing structural until then.

What the spike does to each number

MetricWhat the event does to itWhat it looks likeWhat it means
ImpressionsBuying queries double or triple; deal queries appear from nowhereA page's impressions up 150%Demand moved, not the page
Average positionNew deal queries pull it either way; more searchers on long-tail variantsA guide "rose" from 6 to 3, or "fell" from 3 to 7A different query mix, averaged; not a ranking change
CTRShopping units, ads and deal aggregators crowd the resultsCTR down a third at the same positionThe results page changed; see the CTR post for the normal reasons
ClicksUp, on the pages that rank for buying queriesThe best week of the yearReal traffic, temporary volume
Queries"Deal", "sale", "black friday", "cyber monday" variants of every keywordDozens of new queries per pageNoise, except the few that persist

Every row is a reason not to act on event-week numbers, and the rank tracking post's argument for 28-day windows applies with more force here: the event is the loudest weather of the year.

Reading the report with the event excluded

  1. The spike on its own: Performance → Search results → Compare → custom: event week this year against the same event week last year, aligned on the Thursday (Black Friday moves by up to six days). Clicks up year on year is the only honest "did the event go well" number for search.
  2. Rankings, before and after: the four weeks after the event against the four weeks before. Two clean windows with the spike between them; a page that's higher in the after window than the before window moved for real.
  3. Filter the deal queries out: Query → Custom (regex) → doesn't match deal|sale|black friday|cyber|discount|offer. Apply it to both windows. The averages become about the pages again.
  4. By page, not site: the site total hides which guides did what. Pages tab, sorted by clicks, in each window.
  5. Export both windows and put them side by side in a sheet; the three signals below come from the differences.

The three real signals

  1. New queries a guide started appearing for, and kept. In the after window, a guide shows impressions for a query it had none for in the before window, and the query isn't a deal variant. The event's traffic taught Google the page is relevant to it. Those queries are free targets: a sentence or a FAQ entry on the page, and they often stick.
  2. A page-two guide that climbed on event traffic and held. Position 14 before, position 8 after, with the deal queries filtered. The event's clicks gave Google a reason to test it higher, and the test held. It's the closest thing to free traffic on the site, and a January refresh (verdict, picks, a few internal links) usually finishes the job.
  3. A guide that got the traffic and didn't convert. Clicks from search tripled in the event week (Search Console), Amazon orders through that page's tracking ID didn't (Associates report). The traffic found the page and the page failed it: a stale pick, a weak verdict, a table that doesn't work on a phone, prices out of their band. The event is the stress test the ordinary weeks never run, and this is the list it produces.

The January list

Rank the third signal's pages by wasted traffic (event-week clicks minus what the normal conversion rate would have ordered) and refresh them first; they're the pages that earn most when fixed. Then the second signal's pages, with their page-two keywords as the target.

Then the first signal's queries, as additions to the pages that earned them. Everything on the list is a January job: December is the second peak, a refresh in it costs a sale week, and the data needs the four after-weeks to settle anyway.

The one change that doesn't wait

A dead pick, an unavailable top product, a price that's left its band: these get fixed when they're found, in any week, because every day they stay costs orders and the fix (replace the pick, relabel the tier) doesn't touch anything Google weighs.

The reading above is for rankings; stock and prices are a separate, continuous check that the event week makes more urgent, not less.

Doing it with AffSERP

The third signal is the one AffSERP's tools act on: a guide that had the traffic and not the orders is usually a guide whose picks or verdict have drifted, and the Refresh option regenerates it from current Amazon data on the same URL in January, with the "what changed" line as the record.

The first signal's queries go into the brief for that refresh as questions to answer, and the second signal's page-two keywords become the keyword for a rebrief through the X Best Review tool. The reading itself is yours, in Search Console and the Associates report side by side, and it's the most useful hour of December.

FAQ

Why does my Search Console data look strange after Black Friday?

Because the query mix changed for a week and then changed back. Deal and sale queries flood in, buying queries spike, shopping units and ads crowd the results pages so CTR falls at the same position, and impressions on 'best X' queries double while informational queries barely move. Every averaged number (position, CTR, clicks per page) is distorted by a different mix of searchers, not by anything that happened to your pages. Read the event week on its own, compare it with the same week last year, and judge rankings on the four weeks after it.

How do I compare traffic to the same period last year in Search Console?

Performance report → date picker → Compare → 'Compare custom' with the event week this year against the same event week last year (align on the Thursday, not the calendar date, since Black Friday moves). That's the only fair comparison for the spike itself. For rankings, compare the four weeks after the event with the four weeks before it, and filter out queries containing 'deal', 'sale', 'black friday' and 'cyber' so the deal traffic doesn't sit inside the average.

Should I update my content right after Black Friday?

Not on the basis of event-week data, and not structurally in December anyway. The week after the event is for reading, not changing: note which pages got the traffic, which new queries appeared, and which guides had the clicks but not the conversions. The changes come in January, when the data has settled and the traffic has dropped enough that a refresh can't cost you a sale week. The exception is a dead pick or a wrong price band, which gets fixed whenever it's found.

What's the most useful thing to learn from Black Friday traffic?

Which guides earned the traffic but not the money. A guide that got three times its normal clicks from search and the same number of Amazon orders has a page problem the ordinary weeks hide: a stale pick, a verdict that doesn't convince, a table that doesn't fit a phone, prices out of band. The event is a stress test, and the pages that failed it are the January refresh list, in order of how much traffic they wasted.

The stress-test failures, refreshed in January

A guide that had the traffic and not the orders has drifted; AffSERP's Refresh regenerates it from current data on the same URL, with the event's new queries folded into the brief.

Start with free words
D
Written by Daniel R. SEO Lead, AffSERP

Daniel leads SEO at AffSERP. He writes about keyword research, search intent and site structure, favouring methods you can run with Google and Search Console before reaching for a paid tool.

More about Daniel →