SingRankSingapore Search Visibility Study 2026: What 59,570 Query-Months of Real Data Show
Contents
- Key findings at a glance
- Finding 1: position 1 now earns a lower CTR than position 2
- Finding 2: impressions barely move across page 1
- Finding 3: the visibility cliff starts at position 11
- Finding 4: zero-click queries, with names and numbers
- What does this mean if you run a Singapore business?
- Method
- Limitations — read these before quoting
- FAQ: about this study
Three findings from 59,570 query-months of real Search Console data across a Singapore and Indonesia SME portfolio: measured click-through at position 1 (1.4%) now averages LOWER than position 2 (3.0%); impressions are essentially flat across all of page 1; and the real visibility cliff starts at position 11 — not position 2. This page is the study, the method, and the honest limitations, published so anyone can cite or challenge it.
Why we published this: most SEO statistics circling the industry are either American, outdated, or unsourced. This is Southeast Asian, current, and first-party — and where a number has a caveat, the caveat is printed next to it. Cite freely with a link; the suggested citation sits at the end.
Key findings at a glance
| # | Finding | The number | What it changes |
|---|---|---|---|
| 1 | The CTR inversion | Position 1 averaged 1.4% CTR vs 3.0% at position 2 (17 months, portfolio-wide) | Chasing #1 at any cost can buy the worst-performing slot on an AI-answered query |
| 2 | Page 1 is visibility-flat | Position–impression correlation is just −0.05 within positions 1–6 | Being #7 gets you shown about as often as #1 — what differs on page 1 is clicks, not impressions |
| 3 | The cliff is at 11 | Relative visibility: positions 1–7 = 1.00 → 8–10 = 0.86 → 11–15 = 0.59 → 16–20 = 0.47 → 51+ = 0.30 | The expensive jump is page 2 → page 1, not #3 → #1 |
| 4 | Zero-click is not a rumour | e.g. one query with 1,564 impressions and 0 clicks at strong positions | Ranking #1 for the wrong query pays nothing |
Finding 1: position 1 now earns a lower CTR than position 2
Across 17 months of Search Console data in our client portfolio, the average measured click-through rate at position 1 was 1.4%, against 3.0% at position 2 — an inversion of the curve every SEO textbook still prints.
| Where you rank | Measured average CTR (portfolio, 17 months) |
|---|---|
| Position 1 | 1.4% |
| Position 2 | 3.0% |
Two forces produce this, and honesty requires naming both. The first is the AI answer layer: on a growing share of queries, an AI Overview or featured answer sits above the #1 result and absorbs the click. The second is query mix: sites accidentally rank #1 for informational queries that never click to anyone (see Finding 4), which drags the position-1 average down. Both halves teach the same lesson — a #1 ranking is no longer a promise of traffic, and buying one blindly is buying a number, not a customer.
Finding 2: impressions barely move across page 1
Within a query, moving between positions 1 and 6 barely changes how often the result is shown: the within-query correlation between position and impressions measured just −0.05 for positions under 3 and for positions 3–6. Impressions are counted when a result is rendered, and on most page-1 layouts, all ten results render.
The practical consequence is quietly important for anyone reading their own Search Console: for page-1 queries, your impression count is already a near-unbiased relative measure of demand. You do not need to correct it for position — the correction only starts to matter from position 11 down.
Finding 3: the visibility cliff starts at position 11
Where visibility actually collapses is the jump to page 2. Fitting a two-way fixed-effects model over 7,121 queries and 59,570 query-months, then applying isotonic regression so the curve cannot rise with depth, the relative visibility curve comes out as:
| Position band | Relative visibility (1.00 = page-1 top) |
|---|---|
| 1–7 | 1.00 |
| 8–10 | 0.86 |
| 11–15 | 0.59 |
| 16–20 | 0.47 |
| 21–30 | 0.41 |
| 31–50 | 0.36 |
| 51+ | 0.30 |
Read together with Finding 1, this reframes where SEO budgets should point: the move from position 14 to position 7 roughly doubles how often you are seen at all, while the move from position 3 to position 1 changes little in visibility — and on AI-answered queries may even cost clicks. The expensive, valuable climb is onto page 1. The vanity climb is to the very top of it.
Finding 4: zero-click queries, with names and numbers
Zero-click search is usually discussed in the abstract. Here it is concretely, from our own portfolio — queries where a site held a strong position and earned nothing:
| Query | Impressions (window) | Clicks |
|---|---|---|
| "deepavali 2025" | 1,564 | 0 |
| "hdb hotline" | 1,275 | 0 |
| "biryani recipe" | 991 | 0 |
Nothing was wrong with the rankings. The searcher's question was answered on the results page itself — a date, a phone number, a recipe summary — and no visit was ever going to happen. The lesson for business owners is not despair; it is targeting: a query's value lives in its intent, not its volume, and part of an honest SEO programme is refusing to chase traffic that cannot become a customer.
What does this mean if you run a Singapore business?
Four working conclusions, each following directly from a finding above. First, judge SEO proposals that promise "#1 rankings" against Finding 1 — ask which queries, and what sits above position 1 on them. Second, when reading your own Search Console, trust impressions as a relative demand signal for page-1 queries (Finding 2). Third, prioritise pages stuck at positions 11–20: they sit just below the cliff edge, where the same effort buys the most new visibility (Finding 3). Fourth, before celebrating or buying any ranking, check whether the query can click at all (Finding 4).
These conclusions are also why our own practice targets position 3 rather than position 1 by default, and why the AI answer layer — being inside the answer rather than under it — is treated as a first-class surface. That method is documented separately in our AEO & GEO agency breakdown and why AI doesn't recommend your business.
Method
The data is first-party Google Search Console performance data across SingRank's client portfolio — SME websites in Singapore and Indonesia spanning renovation, automotive, F&B, medical equipment and logistics. The CTR comparison (Finding 1) covers 17 months of portfolio data. The visibility curve (Findings 2–3) is fitted on 7,121 queries and 59,570 query-months using a two-way fixed-effects model — log(impressions) explained by query fixed effects, position-band effects and month effects, so that each query is compared only with itself across time — followed by isotonic (PAVA) regression to enforce that fitted visibility cannot increase with depth. The curve refreshes weekly; figures on this page are from the August 2026 fit and will be re-stated, not silently edited, in future updates.
Limitations — read these before quoting
Every number above is real, and every number has a boundary. The CTR figures describe this portfolio — SME sites in Southeast Asia — and are contaminated by query mix in exactly the way Finding 4 describes, so they are valid as a description and invalid as a universal prediction table. The visibility curve is relative, not absolute: it recovers ratios between position bands, not the total search volume of anything, which no Search Console dataset can identify on its own. And portfolio composition changes over time; that is why the study carries a date, and why we would rather you cite it with its year attached.
FAQ: about this study
Can I cite or republish these numbers?
Yes — cite freely with a link to this page and the year (2026). Suggested citation: SingRank, "Singapore Search Visibility Study 2026", singrank.com, August 2026. If you challenge a number, we would genuinely like to hear it; the method section exists so the work can be checked, not just quoted.
Does position 1 being 1.4% CTR mean ranking #1 is worthless?
No — it means #1 is no longer automatically valuable. On queries with no AI answer and strong commercial intent, top positions still win. The finding is a warning against buying rankings blindly: on AI-answered and informational queries, the top organic slot now sits under the answer, and the average shows it.
Why does your data show a cliff at position 11 when CTR studies show decline from position 1?
Because they measure different things. CTR studies measure clicks per impression, which falls steeply from position 1. This study's curve measures impressions — how often you are shown at all — which stays flat across page 1 and collapses at page 2. Both are true at once, and together they say: get on page 1 for visibility, then win the click through relevance and the answer layer.
Is this data Singapore-specific?
The portfolio is Singapore- and Indonesia-weighted SME data, which is exactly why we published it — most public SEO statistics are US-weighted and enterprise-weighted. Treat the exact figures as Southeast Asian SME evidence, and the structural findings (flat page 1, cliff at 11, AI absorption at the top) as the parts most likely to generalise.
Will the study be updated?
Yes — the underlying curve refits weekly and the page will be re-stated periodically with the date changed and prior figures preserved in the text where they materially moved. A study that silently edits its own history is not one worth citing.
Who do I contact about the data?
The SingRank team, through the site's contact channel. Requests to verify a figure, methodological questions, and corrections are all welcome — that standing offer is part of what makes the numbers citable.
Published by the SingRank team — a Singapore SEO, AEO and GEO agency with a programming and data-science foundation (registered 2025; founder and team experienced since 2010). First-party data, stated method, printed limitations.
Disclaimer: figures describe SingRank's client portfolio for the stated windows and move over time; they are evidence, not guarantees. No agency, including SingRank, can guarantee specific rankings or AI citations.
