A Search Console impressions spike usually comes from one or two pages, not your whole site. Open the Pages tab, compare the spike weeks with the weeks before, and sort by the difference. Then find the day each page jumped and look one to two weeks earlier for what you changed.
That's what I did for supermarketers.ai this week. I pulled Search Console and Google Analytics, gave the data to Claude along with my commit history, and asked what changed. The answer was two pages and one kind of edit.
What happened to my impressions?
Weekly impressions sat around 1,300 through mid-October. Then they jumped to 2,320 in the week of Oct 26 and 3,121 in the week of Nov 2. That's 2.4 times as many in two weeks.
Clicks held at 7 to 8 a week. Google Analytics showed organic visits going from 5 to 10 a week. So Google was showing the site a lot more, and a few more people came.
What caused the spike?
Two pages made 78% of the extra 2,789 impressions:
- The AI visibility tools guide added 1,347.
- The B2B buyer research guide added 838.
The new Hiring Index page added another 15%. Every other page stayed flat at about 100 impressions a day, so this wasn't Google lifting the whole site.
Both pages got new titles, descriptions and FAQs on Oct 20. Both jumped on the same day, Oct 28. Eight days is a normal gap for Google to recrawl a page and show the new version.
Why did the title change work?
The new titles and FAQs used the exact questions buyers type. The new impressions came from queries that repeat that wording almost word for word:
- "Do B2B buyers use ChatGPT to research suppliers?" went from 0 to 150 impressions.
- "How are B2B buyers using AI to research software?" went from 26 to 171.
- "Who does AI visibility audits for B2B software companies?" went from 0 to 35, at position 1.2. It matches a FAQ question I added.
Matching the title tag, the description and the FAQs to the exact question I want to be shown for made the difference. The numbers are small, but they point the right way.
Why didn't clicks follow?
The new queries are long, full-sentence questions. They came from 34 countries instead of 11, mostly on desktop, and the two pages earned 4 clicks in three weeks.
That pattern fits Google's AI features, AI Overviews and AI Mode, which Search Console counts as impressions. It also fits AI visibility tools that run tracked prompts through Google. Search Console can't show who searched, so treat this as the likely reason, not a proven one. Its generative AI report, filtered to the pages, is the place to confirm it.
I don't expect organic clicks to grow much from here. Buyers do more of their research inside an AI answer and click through only at the end, once they have a shortlist. Being shown in that answer is the goal, and impressions are an early signal of it.
How do you trace your own spike?
Run these six checks in order:
- Search Console, Performance, Pages tab. Compare the spike weeks with the weeks before and sort by the difference. If one or two pages hold most of the gain, it's a page change.
- Filter to one page and look at the daily chart. A clean step on one day means Google recrawled a change.
- Check your commits or CMS history for changes to that page 3 to 14 days before the jump.
- Open the Queries tab for that page and compare the two windows. New queries that reuse your new wording are the link.
- Check Countries and Devices. Long questions from many countries with no clicks point to AI features.
- Google Analytics, Traffic acquisition, Organic Search. See whether visits followed.
You can export each report and give them to Claude or ChatGPT to read side by side. That's how I found the two pages in a few minutes.
What I ruled out: an IndexNow ping that went live the same day (Google doesn't use IndexNow), my own rank tracker (it checks short keywords, none of the spiking queries), and how many changes I shipped (the busiest week, Oct 5, had flat impressions).
Why track every change?
I could only answer this because every change to the site is logged with a date, a reason and the number it should move. The log is public on the Lab. When a number moves, the log tells you what to look at first.
The site is custom-coded with Claude Code, which makes small experiments fast. A CMS works too. The point is to write down what changed and when.
Can you see ChatGPT traffic in Google Analytics?
Partly. In Google Analytics, go to Reports, Acquisition, Traffic acquisition. Change the primary dimension to Session source / medium, then add Landing page + query string as a secondary dimension. Visits from ChatGPT show up with chatgpt.com as the source and the page they landed on. The question the person asked doesn't come through.
What are the limits?
- This is one small site, around 3,000 impressions a week.
- Search Console can't show who searched, so the AI-features reading is likely, not proven.
- The two pages were rewritten again on Nov 10. I'll read those changes on Dec 9.
How does SuperMarketers use this?
SuperMarketers is the AI marketing partner that gets B2B SaaS companies chosen by more buyers, and runs the marketing behind it. Every client change goes into a dated log like this one, and we read each change against Search Console and AI answers 28 days later. That's how we decide what to repeat. See how we track AI visibility.
FAQ
Why did my Search Console impressions suddenly go up?
Usually one or two pages changed. Open the Pages tab, compare the spike weeks with the weeks before, and sort by the difference. In my case two pages made 78% of the jump.
How do I find what caused a spike in Search Console?
Find the page, find the day it jumped, then look one to two weeks back for what you changed on it. Then check the Queries tab for new questions that reuse your new wording.
Why are my impressions up but my clicks are not?
Long, full-sentence questions from many countries with almost no clicks usually mean Google's AI features (AI Overviews and AI Mode) are showing your page. Search Console counts those as impressions.
Do title changes increase impressions?
They did here. After a title rewrite, the question "Do B2B buyers use ChatGPT to research suppliers?" went from 0 to 150 impressions on that page.
Are impressions a vanity metric?
On their own, they can be. They show that Google displays your page, not that anyone visited. Treat a jump as a leading signal and check clicks and Google Analytics visits before calling it growth.