Impressions But No Clicks: 8,782 of Ours, Explained

Your agency sends a report. Impressions are up. Clicks are flat. They call it progress and ask you to be patient. Here is the same chart from our own site, so you can judge that pattern with something real in front of you.

Between 30 August and 26 September 2026, singrank.com's 40 highest-impression search queries recorded 8,782 impressions and 0 clicks. Not one of the 40 reached the top 10. Three reached positions 11 to 20, and between them those three carried 327 impressions, 4% of the total. 27 of the 40 sat past position 50, and those 27 carried 6,385 impressions, so 73% of everything we were "seen" for was well beyond where anyone scrolls.

That is our own Google Search Console, and we are publishing it because the number is the argument. If your report looks like this, the honest reading is not that the work failed. It is that the work has not started producing yet, and anyone presenting those impressions as a result is showing you the wrong column.

The short version:

• An impression only means a link was seen, or potentially seen. At position 63 it means almost nothing.
• Sort your queries by position, not by impressions. If none sit in the top 10, you have exposure rather than visibility.
• Impressions without clicks is normal in months 1 to 4, worth questioning by month 5 to 8, and a finding by month 9 to 12.

What Google actually counts

An impression is a weaker thing than the word suggests. Google defines it as meaning that "a user has seen (or potentially seen) a link to your site" in Search, Discover or News. Note the parenthesis. Potentially seen.

Average position is defined as "a relative ranking of the position of your link on Google, where 1 is the topmost position, 2 is the next position, and so on", averaged across every query where your content appeared.

Put the two definitions together and the trap is plain. A result at position 63 counts an impression even though nobody scrolled to it. Impressions can climb steadily while clicks sit at zero, and nothing has gone right or wrong. The metric grew. The business did not.

Are AI answers mixed into these numbers?

Yes, and you cannot separate them. Google says appearances in its AI features are "included in the overall search traffic in Search Console", reported in the same performance report. So some share of any site's impressions comes from AI surfaces, and the report will not tell you which. That is one more reason to judge position bands and enquiries rather than a single total.

Our 40 queries, sorted by position instead of impressions

Same data an impressions chart is built from, grouped the way we think an owner should read it.

Average position band	Queries	Impressions	Share
1-10	0	0	0%
11-20	3	327	4%
21-30	1	86	1%
31-40	1	94	1%
41-50	8	1,890	22%
51-60	12	2,317	26%
61-70	9	2,668	30%
71+	6	1,400	16%
Total	40	8,782	0 clicks

Sorted by impressions, this reads like momentum. 29 of the 40 queries name Singapore. 17 of them are someone shopping for an agency, carrying 4,299 impressions between them, and 9 ask about price, cost or packages, carrying 2,043. Commercial intent, real exposure, a story you could tell for six months.

Sorted by position, it says one thing: on the phrases that would pay, we are not in the race yet.

The barbell every young site has

Look at the two ends of the same table. Our best-placed query sits at average position 11.2 and carries 99 impressions. Our biggest query by demand sits at position 63.4 and carries 887.

So we rank respectably for a phrase few people want, and we are on page 6 for the phrase everyone wants. That is not a contradiction and it is not failure. It is the shape of a site that has not yet earned enough trust to compete where the money is, and it is why one site-wide "average position" figure tells an owner almost nothing. Two sites with the same average can be in completely different businesses.

For context on age: the oldest article on this blog is dated 11 June 2026, so the domain you are reading is three and a half months into publishing. SingRank was registered in Singapore in 2025, and the people behind it have worked in software and digital marketing since 2010, but this site's numbers are a young site's numbers. We would rather show you ours than borrow someone else's.

What if your rankings held but your traffic fell?

That is the opposite problem and it needs a different answer. Zero clicks on deep positions means you were never in the race. Falling clicks on positions that did not move means something took the click in front of you, usually an answer above the results or a competitor with a better title.

The test is whether position and clicks moved together. If position is flat and clicks dropped, look at what now sits above you rather than at your own rankings, which is the case we worked through in why your website traffic dropped but rankings didn't. If position is deep and clicks are zero, as ours are, there is nothing to diagnose yet. There is only work to do.

Three questions to ask about any search report

Whoever does your search work, including us, these turn a report into a decision.

1. How many of my queries are inside the top 10, and what share of my impressions do they carry? If the answer is "none" and "none", you have exposure rather than visibility. Ask which named queries are meant to reach the top 20, and by when.
2. Which way is each query's position moving? A query going from 63 to 41 is real progress with no clicks attached. A query sliding from 41 to 58 while the site's impressions rise is a decline hidden by growth. Both are invisible in a total.
3. Which of these queries would actually pay? 17 of our 40 are people shopping for an agency. Those are worth climbing for. A competitor's login page is not, whatever the number beside it does. We made that case at length in how to tell a real trend from a dead end.

A report that cannot answer all three is not reporting. It is reassuring.

When impressions without clicks becomes a real problem

Timing is the whole judgement, so here is where we would draw the lines.

Months 1 to 4. Impressions appearing at all is the signal. Zero clicks from positions 40 to 70 is ordinary, and we would not renegotiate anything on it. This is where our own site sits.

Months 5 to 8. Now you want movement in the bands rather than the totals. Some named queries should have crossed into the top 20 and a few clicks should exist. If everything is still past position 40, ask what changed in the plan and what evidence says it will work.

Months 9 to 12. By now impressions without clicks is a finding, not a phase. Either the targets were too competitive for the budget, which someone should have said in month one, or the pages are not good enough to rank. Our performance benchmarks piece sets out what to baseline and re-measure every 30 days.

One more reason clicks lag, and it is not an excuse we invented: an answer above the results absorbs the click. Our Singapore search visibility study measured that across 59,570 query-months of real data. Even a top-10 position earns less than it once did, which makes the position bands more important, not less.

What this costs us to say

It costs us the easiest sales line in the business. Every agency in Singapore can show a rising impressions chart, ours included, and we have just explained why you should not buy on one.

It also invites the obvious question: if your own site gets zero clicks on its main queries, why hire you? Fair enough. Our answer is that we would rather be judged on whether we told you the truth about a number than on whether the number flattered us. Our site is three and a half months into publishing, against agencies with ten-year-old domains defending these exact phrases for a decade. We will publish this table again as it changes, including if it does not improve.

And the harder concession: if you need enquiries this quarter, none of this is your answer. There are situations where search is the wrong thing to buy, and a report full of position-60 impressions is a fair sign you are in one of them.

One next step

Open your own Search Console, go to the performance report, switch on the average position column, and sort your top 40 queries by position instead of impressions. Count how many sit inside the top 20. That takes five minutes and it is the most useful thing an owner can do with the tool unaided.

If you would rather we read it with you, send us a screenshot or view-only access through our contact page and we will tell you which band your demand sits in and what would have to move. It starts as a free check, and if the honest answer is that your current agency is doing fine, we will say so.

Figures are singrank.com's own Google Search Console data for 30 August to 26 September 2026, covering the 40 queries with the most impressions in that window. No client data appears on this page.

FAQ

What does it mean if impressions rise but clicks stay at zero?

Usually that your pages appear deep in the results, where nobody scrolls. Google counts an impression when a link was seen or potentially seen, so a result at position 63 still registers one. Across our own top 40 queries, 8,782 impressions produced no clicks at all.

Is zero clicks normal for a new website?

In the first few months, yes, if positions sit past 40. Our site is three and a half months into publishing and none of its 40 highest-impression queries reached the top 10. The signal to watch is whether named queries climb into the top 20, not whether impression totals grow.

How should I read an SEO report from an agency?

Sort queries by average position rather than impressions, then ask three things: how many sit in the top 10, which way each position is moving, and which of those queries would pay if they ranked. A report that cannot answer all three is reassurance rather than reporting.

Why does a top-10 position earn fewer clicks than it used to?

Because an AI answer or summary above the results often satisfies the question first. We measured that effect across 59,570 query-months of Search Console data in our visibility study. It is a reason to watch position bands and enquiries together, rather than treating rank as the finish line.

Can I see AI Overview impressions separately in Search Console?

No. Google says appearances in its AI features are included in overall search traffic in the same performance report, so they are not broken out. You can watch the Web search type and compare trends, but you cannot attribute a given impression to an AI surface.

My rankings did not move but my clicks fell. Is that the same thing?

No, that is the opposite case. Zero clicks on deep positions means you were never visible. Falling clicks on steady positions means something above you took the click, so the useful check is what now appears above your result rather than your own ranking.

When should I stop paying for SEO?

Our line would be month 9 to 12 with no query having crossed into the top 20 and no explanation that survives the three questions above. Before that, judge the bands rather than the totals. After that, either the targets were mispriced from the start or the pages are not competitive.