Search Console is the one dashboard on this site that cannot flatter us. It counts how often Google showed a page. It counts how often someone clicked.

For 1–28 September 2026, BuzzRiding recorded 14 clicks from roughly 2,600 impressions. Click-through rate was 0.5%. Average position was 36.5.

Those four numbers hide the real story. Underneath them, the data is lopsided in ways that change what a new blog should do next.

🔬 How this was built

Every figure comes from the Search Console Performance report for 1–28 September 2026, Web search type. We copied the table rows into CSV files. A design note states what this report is and is not. Google documents that its tables omit anonymized queries and truncate rows, so every query list below is partial.

The month in four numbers

Search Console listed 38 pages with at least one impression and 135 queries. That is a small footprint. Small footprints make averages unstable, and the next six findings are all about that.

Finding 1: one page holds more than a third of the impressions

The Gamma vs Beautiful.ai vs Canva comparison logged 979 impressions in the month. It earned zero clicks.

That single page accounts for more than a third of all impressions on the site. The other 37 pages with impressions share the rest.

The effect on the headline number is large. Remove that page and about 1,600 impressions remain. The same 14 clicks then give a click-through rate near 0.9%, not 0.5%.

The site-wide rate is not a measure of how good our titles are. It measures how much of our visibility sits on one page nobody clicks.

Finding 2: seven of the ten top queries are two comparisons

The ten most-shown queries are not ten topics. Seven of them are phrasings of just two comparisons.

Four are Canva-versus-Gamma variants. Three are Surfer, Clearscope and Frase variants.

Together those seven queries show 302 impressions. None produced a click.

The Surfer SEO vs Clearscope vs Frase post shows the same pattern on its own. It logged 180 impressions across 25 queries at an average position of 37.2. It earned zero clicks.

The demand for these comparisons is real and specific. The ranking is not there yet, and position 37 is page four.

Finding 3: two page-one positions rest on almost nothing

The page-level data contains two pages with strong-looking positions. The AI skills post averaged position 5.0. The ChatGPT vs Claude vs Gemini post averaged 6.5.

The first had exactly one impression. The second had four.

An average position built on one impression is one search on one day. It says nothing about where the page ranks for anyone else. We nearly counted both as wins.

The rule that follows is simple. Never quote a position without the impression count beside it.

Finding 4: the 14 clicks are spread across nine URLs

No page earned more than four clicks. The Claude Projects template post led with four. The AI marketing job titles post and the AI marketing certifications post earned two each.

Six more URLs earned one click each. They are the interview questions post, the Perplexity Comet vs ChatGPT Atlas post and the Meta Advantage+ post. The best GEO tracker tools post, the portfolio guide and the blog index complete the list.

That is a long tail, not a hit. It also means one lucky click moves a page's rate by several points. Treat every per-page rate below as a rough guide.

Finding 5: same position band, very different click rates

Two pages sit in nearly the same spot. The Claude Projects post averaged position 5.9. It earned 4 clicks from 68 impressions, a 5.9% click-through rate.

The job titles post averaged position 6.7. It earned 2 clicks from 237 impressions, a 0.8% rate.

The gap is probably not the title, which already names its topic plainly. The likelier cause is who was searching. Search Console shows only three queries for the job titles page: “ai marketing roles”, “advertising” and “ba marketing”. Each has one impression. Two of them have little to do with the page.

We cannot prove that, because most of the 237 impressions sit behind hidden queries. It remains the likelier explanation than a weak headline.

Finding 6: the best-performing page shows no queries at all

For the Claude Projects post, Search Console listed no queries. It had four clicks, 68 impressions and an empty query table.

Google explains why. Anonymized queries are left out of the table but counted in the totals, as its Performance report documentation describes. Its deep dive on filtering and limits adds that the interface stores only top rows and exports at most 1,000.

So the pages with the least visible data are the ones where guessing a better title is most dangerous. We would be rewriting from a handful of rows. That is why we are not.

How to read your own data the same way

None of this needs a paid tool. The same four checks work on any small site.

  1. Sort pages by impressions. Find out what share the top page holds.
  2. Group queries by meaning. Ten rows may be two topics in different words.
  3. Put impressions next to position. A position with fewer than 20 impressions is noise.
  4. Check which pages show no queries. Those are the ones you cannot optimise from the table.

Do this monthly, on the same day, and keep the numbers. The trend across months is worth more than any single reading.

What we are changing because of this

💡 Key takeaway

The problem is not click-through rate. Most impressions sit on page four or later, where almost nobody clicks. Better titles cannot fix a ranking problem.

What this data does and does not tell you

It tells us where visibility sits. One comparison page and two comparison topics dominate our impressions. The pages with the most clicks are mostly not the pages with the most impressions.

It does not tell us why a searcher clicked or skipped. It does not show the hidden queries behind most page-level impressions. It cannot say whether a new title would lift click-through. No page here had enough clean data to test one.

It also cannot separate real demand from ranking luck. A page sitting 37th for a busy query can still show a handful of impressions on a few lucky days. Only more months of the same report can settle that.

That is why this post states its limits as plainly as its findings. A new blog's first Search Console data is a map of where to look next. It is not a verdict.

The obvious caveat

One property, one month, 28 days. The tables show only top rows and omit anonymized queries. We copied the rows into the CSVs by hand rather than exporting them, so a transcription slip is possible. The raw files sit in the repository so anyone can check our arithmetic and disagree with it.

Frequently Asked Questions

Why can a page show an average position of 5 and get no clicks?
Because the average may rest on one impression. In this data, one page averaged position 5.0 on exactly one impression and another averaged 6.5 on four. Always read position alongside the impression count.
What are anonymized queries in Search Console?
They are searches Google hides from the query table to protect the searcher's privacy. Google's documentation says they are omitted from tables but included in chart totals unless you apply a query filter.
Is a 0.5% click-through rate bad for a new site?
It depends on where the impressions land. An average position of 36.5 means most impressions appeared on page four or later, where few people click. Judge pages with 20 or more impressions individually instead.
How much Search Console data can you export?
Google's documentation says the interface exports at most 1,000 rows for query and page reports. Requests without query or URL dimensions, such as countries and devices, export everything. Small sites rarely hit the limit.
Should you rewrite titles when impressions are high but clicks are zero?
Only when the page already ranks well and the visible queries match its topic. In this data, the high-impression zero-click pages ranked on page four or later. A new title would not have helped.