For a Singapore business, non-branded impression diversity measures how many different buyer questions, outside searches for your own name, put your pages in front of searchers, and a business whose impressions come mostly from its name stays invisible to buyers who do not know it yet. Google added a branded queries filter to Search Console in November 2025, so you check the split directly in your own data. The questions you appear for also shape what Google and the AI chats learn about the problems you solve. Aemorph measures the spread by question cluster, then fills the clusters where Singapore buyers ask and you do not appear.
Low Diversity Shows as Healthy Impressions That Come From Your Own Name
On a Singapore site, low non-branded impression diversity shows as healthy total impressions that come mostly from searches for your own name, with only a handful of non-branded queries carrying the rest. The total looks like growth, while the buyers who search your category without your name never meet you.
| Symptom | What you see in the data | Likely cause to test | Fix | Where it gets fixed |
|---|---|---|---|---|
| Total impressions up, new buyers flat | Branded queries carry most of the impressions | Your pages answer few of the questions buyers ask about your category | List the category questions and answer each one | Your site's content plan |
| Non-branded impressions from a few queries | One or two pages carry your non-branded visibility | Coverage concentrated on a single topic | Spread answers across the question clusters | New and updated pages |
| Impressions without clicks on category questions | You appear, but low or below an AI summary | The page answers the question only in part | Lead each page with a clear, claim-first answer | The page itself |
| Missing from AI answers on category questions | AI chats name other providers | Little on your site states the problems you solve | Publish approved answers to the buyer questions | Your site and your LLM info page |
Symptoms, likely causes and fixes for low non-branded impression diversity on a Singapore site.
Low diversity is often the visible trace of topical thinness, where a site covers too few of its category's questions to rank for them.
What is non-branded impression diversity?
Non-branded impression diversity is the number of distinct non-branded queries and question clusters that earn your Singapore site impressions in Google, and how evenly those impressions spread across them. High diversity means new buyers meet you across many of the questions they ask; low diversity means a few queries, or your own name, carry your visibility.
What are brand keywords?
Brand keywords are searches that include your business name, or a product or service unique to you, such as an acronym or Chinese name buyers in Singapore use for you. Search Console classes as branded a query with your brand name in any language, its typos and queries that refer to a unique product or service of your site.
Branded Impressions Mask the Gap Because They Grow With Your Reputation
Branded impressions mask the gap because Singapore buyers who already know you keep searching your name, so total impressions rise with your reputation while your visibility to new buyers stays flat. The questions you rank for are much of what Google and the AI chats learn about the problems you solve, and searches for your name teach them only that you exist.
Google draws the same line. In announcing the branded queries filter in November 2025, Google wrote that branded queries typically lead to higher-ranking pages and higher click-through rates, while non-branded queries show how new users find your content without any initial intent to visit your site.
Source: Google Search Central, November 2025
AI summaries widen the gap on category questions. In Pew Research's March 2025 study of 900 US adults, 18% of Google searches produced an AI summary, and users clicked a traditional result in 8% of visits with a summary against 15% without one. A non-branded impression earns fewer clicks than it used to, so the number of questions you appear for carries more weight.
Source: Pew Research Center, July 2025
Why do branded impressions rise while new buyers never see you?
Branded impressions rise with each campaign, referral and repeat customer, while new Singapore buyers never see you because they search the problem, not your name, and your pages do not answer that problem. The two numbers move independently, which is why the total hides the gap.
Singapore Impressions Come Almost Entirely From Google
For a Singapore site, Search Console covers most of the picture, because Google carries more than nine in ten Singapore searches. Google held 92.16% of search engine market share in Singapore in September 2026 and Bing 4.53% (StatCounter).
| Measure | Figure | Market and date | Source |
|---|---|---|---|
| Google search engine market share | 92.16% | Singapore, September 2026 | StatCounter |
| Bing search engine market share | 4.53% | Singapore, September 2026 | StatCounter |
| Internet users | 5.78 million, 98.4% of the population | Singapore, as of October 2025 | DataReportal |
| Surveyed workers using AI tools at work | 73.8% | Singapore, May and June 2025 pulse survey | IMDA |
| AI adoption among non-SMEs | 62.5% | Singapore, 2024 | IMDA |
| AI adoption among SMEs | 14.5%, up from 4.2% in 2023 | Singapore, 2024 | IMDA |
Sources: StatCounter, September 2026 and others
A brand-term list for a Singapore business includes each variant buyers type, such as an acronym, a Chinese name or a name shared with another company, and Aemorph checks it against your Search Console queries before the first report. A grant-funded EDGE package that tracks keyword positions gains the same split when its keywords are checked against the brand-term list.
Measuring Non-Branded Impressions by Question Cluster Shows Where Buyers Miss You
The diagnostic splits your Singapore impressions into branded and non-branded, groups the non-branded queries into question clusters, and marks each cluster on your buyers' question list that shows no observed impressions in the available report. Aemorph runs it in six steps:
- Splits your Search Console impressions into branded and non-branded queries with the branded queries filter, or with your list of brand terms where the filter is not available.
- Counts the distinct non-branded queries that earned impressions over the period.
- Groups those queries into question clusters, using the question list built with the EAV Exhaustion Methodology, and shows each cluster's share of your non-branded impressions.
- Flags each cluster on the list with no observed impressions in the available report for the period: the questions buyers ask where you do not appear in the data Search Console shows.
- Tests the same clusters in ChatGPT, Claude, Gemini and Perplexity, so the gaps in Google and in the AI answers sit side by side through cluster-level measurement.
- Ranks the gaps by funnel stage, bottom-of-funnel questions first.
The split rests on Google's own definitions. Search Console Help defines an impression as a user seeing, or potentially seeing, a link to your site in Search, Discover or News. The branded queries filter segments that data into a Branded view, for queries that include your brand name or closely associated products, and a Non-branded view for the other queries. Position tracking adds the detail the filter leaves out, as the guide to tracking keyword rankings shows.
Sources: Google Search Central and Search Console Help, October 2026
How do you separate branded and non-branded queries in Search Console?
You separate branded and non-branded queries by opening the Search results Performance report in Search Console and applying the branded queries filter, which splits your query data into a Branded view and a Non-branded view. Where the filter is not available for your property, a filter built from your own brand terms does the same job less completely.
Aemorph's guide to tracking keyword rankings shows how to check positions for the non-branded queries this filter isolates.
Narrow Impressions Cost More as Buyers Ask AI
Low non-branded diversity costs a Singapore business more as buyers ask AI, because an AI answer to a category question draws on pages that answer that question, and a site that answers few of them gives the answer little to draw on. Google now reports this surface separately: its generative AI performance reports, announced on 3 June 2026, show how often URLs from your site appeared in AI Overviews and AI Mode, rolling out to a subset of websites first.
| Situation | Buyer impact | Aemorph's recommended recheck | Next step | What to retest |
|---|---|---|---|---|
| Brand-heavy, few non-branded queries | Buyers who do not know your name do not meet you | Before the next content plan | List the category questions and answer the bottom-of-funnel ones first | Non-branded impressions by cluster |
| Non-branded impressions in one cluster | Buyers find you for one problem and miss the rest | Monthly, with the report | Answer the empty clusters next to the one that works | Each cluster's share of non-branded impressions |
| Spread across the listed clusters | The target: new buyers meet you across the questions they ask | Monthly, and after a change to your offer | Keep adding answers as buyer questions change | Each cluster in Google and the AI chats |
Buyer impact and recommended recheck by impression spread, as illustrative scenarios.
Aemorph sets no fixed recheck interval. Your first working day with Aemorph is the first check date, and pages are then assigned for update in order of importance: bottom-of-funnel, middle-of-funnel, then top-of-funnel.
Aemorph's EAV Exhaustion Methodology widens that coverage by listing every question a buyer asks about your category before any page is written.
How often do you check impression diversity?
You check impression diversity monthly and after each batch of new pages goes live on your Singapore site, because new pages earn impressions only once Google has crawled them. Aemorph reports monthly, with results grouped by topic and prompt cluster and ranking reports alongside.
Search Console Shows Visibility, Not Revenue, and Its Brand Split Has Limits
Search Console shows the impressions Google counted for your Singapore site, and its branded filter is an AI-assisted classification with limits on where it is available, so you read the split as a strong signal rather than an exact count. The numbers describe your visibility; they do not describe your revenue or enquiries.
Google states the filter's limits in its announcement: an internal, AI-assisted system decides the classification, the filter works only for top-level properties, not URL path or subdomain properties, and only for sites with enough queries and impressions. Search Console Help adds that impressions are counted either by URL or by property depending on the report's grouping, so totals differ between views of the same data. It also omits anonymised rare queries from the table, though they count in the chart totals, and the table shows at most 1,000 rows, so a cluster with no observed impressions is a gap in the visible data rather than proof of zero impressions.
Your data stays yours. You grant Aemorph access to your Search Console and keep ownership of it, and Aemorph works under an access practice aligned with ISO 27001. Aemorph tracks rankings, traffic and visibility; it does not collect your sales, revenue or enquiry figures.
Does the branded queries filter change your rankings?
No: Google states that the branded queries filter has no effect on how Google Search ranking works; it only segments the data you see in Search Console. Rankings change through the pages you publish, not through how you filter the report.
SEO Built From the Question List Widens the Questions You Appear For
Aemorph's SEO widens a Singapore site's non-branded visibility by answering the clusters on your question list that show no observed impressions in the available report, bottom-of-funnel questions first, and reports the spread by cluster each month. The free Money on the Table audit puts a number on the revenue lost at your current visibility in search.
The Showstopper Sensor (Sprint 1), a 30-day showstopper-answer sprint for ChatGPT, Claude, Gemini and Perplexity, publishes your approved answers to your showstopper questions, the buyer questions that stop a purchase when AI answers them badly, on your site. Many of those questions never include your name, so the answers add non-branded coverage in the AI chats as well as in Google.
Does Aemorph need access to your Search Console?
Aemorph needs read access to your Search Console to run the branded split and the cluster count for your Singapore site, and you keep ownership of the property. Aemorph reports the result monthly by topic and prompt cluster.
Begin with a free strategic conversation, which includes prompting the AI chats for your current brand answers about your Singapore business. SEO is scoped from your question list and set out in your quote; starting points are on the pricing page.