Last updated: 21 August 2026. Written by Adrian Tan, Singapore Digital Marketing (SDM).
There is a specific moment where content plans in Singapore go wrong, and it is not the writing. It is the morning someone exports 400 keywords from a research tool, sorts the column by volume, and starts at the top. Four months later the business has twelve published posts, no rankings, and a marketing lead who cannot explain to a director why the effort produced nothing.
The export was not the problem. Sorting by volume was. Content keyword research is a different job from the keyword research you do for a service page, and it uses different criteria: not “what has the most searches” but “what should we write next, given who we are, what we can prove, and what a post can still win in 2026”.
This guide covers that job. Why Singapore search volumes are far thinner than most plans assume — with real numbers from our own SEMrush Singapore data — where demand shows up outside a keyword tool, how to read the tools honestly, a scoring model for turning research into an ordered queue, and which keywords are no longer worth a post at all. It is part of our content marketing guide for Singapore businesses. For the mechanics of the tools themselves — match types, difficulty scores, competitor gap analysis — start with our keyword research guide for Singapore businesses; this piece assumes you can already run the tools and covers what to do with the output.
Two different jobs, often confused
Keyword research for a service page asks: what is the single commercial term this page should own, and can we beat the sites currently ranking for it? It is a one-off decision per page, weighted heavily towards commercial intent, and it usually points at a small number of high-competition terms.
Keyword research for content asks something different: given roughly forty publishing slots over the next twelve months, which forty topics, in which order, produce the most compounding value? That is a portfolio question. It cares about coverage, sequencing, and whether the set of posts together makes the site credible on a subject — not about any one term.
The distinction matters commercially, because writing a blog post that targets your service page’s head term does not help; it competes. In our own map, commercial head terms like “seo agency singapore” (KD 38) belong on the service page, and the blog takes the informational family around it. Getting that boundary wrong is the most common self-inflicted wound we see in Singapore content plans.
| Service-page keyword research | Content keyword research | |
|---|---|---|
| Question | What should this page own? | What should we write next, and in what order? |
| Unit | One term per page | A queue of 30–60 topics |
| Weighting | Commercial intent, conversion value | Coverage, proof available, sequencing, effort |
| Volume matters | Considerably | Far less than intent and provability |
| Failure mode | Targeting a term you cannot win | A list with no order and no depth |
The Singapore volume problem
Here is the single most important local fact, and it is one that imported playbooks get badly wrong: Singapore search volumes are small. A market of 5.88 million people does not generate US-scale query volumes, and a strategy that requires 1,000-a-month keywords to be viable is not a Singapore strategy.
These are real figures from our own SEMrush Keyword Magic data on the Singapore database, harvested for our content planning:
Read the last two rows together, because they carry the lesson. “Multilingual seo singapore” records 110 searches a month at difficulty 6 — a term a focused site can genuinely own, attached to a service with real deal value. Meanwhile a whole family of terms most agencies would assume are valuable, including “google ads cost singapore”, records no volume at all in the Singapore database. If you had planned by sorting on volume, you would have skipped the winnable term and written for a term the tool cannot even see.
Where demand actually shows up
Because Singapore volumes are thin and tool coverage is patchy, a keyword tool should be one of five inputs, not the only one. The other four are where the differentiated topics come from.
- Search Console query data. The most underused asset most sites own. Filter to queries where you already get impressions but a poor average position — those are topics Google has already decided you are relevant for, and the page currently answering them is failing. Setup is covered in our Search Console guide.
- Sales and support conversations. Ask the two people who speak to customers most for the questions they answer weekly. These map to real buying decisions and no tool surfaces them, because buyers ask them in conversation rather than in a search box.
- The SERP itself. People-also-ask boxes, related searches, and the subheadings of the pages currently ranking. This is free intent data specific to the exact query, and in Singapore it often reveals a local sub-question (grant eligibility, GST treatment, PDPA implications) that global pages ignore entirely.
- Community and comment threads. Local Reddit and Facebook communities, LinkedIn comment threads on your own posts, industry WhatsApp groups. This is where the objections live.
- AI assistant prompts. The questions people now ask a chatbot are longer and more conversational than search queries. Ask a model the questions your customers ask, look at what it answers with, and note where its answer is generic — that gap is a brief.
Reading the tools honestly
Every volume figure you will use is an estimate, and knowing how each one is wrong keeps you from over-trusting the decimal points.
Google Keyword Planner comes from Google, which is its advantage, but for accounts without meaningful active ad spend it returns bucketed ranges rather than precise numbers — you get a band, not a figure. Its data is also built for advertisers, so it groups semantically similar terms together, which inflates apparent volume for a specific phrasing.
Third-party tools (SEMrush, Ahrefs and the rest) model volumes from clickstream data and their own modelling. For a market Singapore’s size, small absolute numbers mean large relative error: the difference between a reported 40 and a reported 70 is not a signal you should plan around. What they are genuinely good for is relative comparison within the same tool and the same market, plus difficulty scoring and competitor gap analysis.
Search Console is the only source reporting your actual queries, and it is therefore the truest — but it only shows what you already rank for at all, so it cannot find topics you have never covered.
The working rule: use third-party tools to size and compare, use Search Console to validate and to find near-misses, and never let a volume figure outrank an intent judgement.
Intent, funnel stage and format
Once you have candidates, classify them before you order them. Intent determines what the post has to do, which determines its format and its length — and, crucially, whether it belongs on the blog at all.
| Query pattern | Intent | Belongs on | What the post must do |
|---|---|---|---|
| “what is…”, “how does… work” | Informational, early | Blog | Explain completely; capture the reader for later, expect no conversion |
| “how to… singapore” | Informational, actionable | Blog | Be genuinely executable; local specifics are the differentiator |
| “… cost”, “… price” | Commercial investigation | Blog | Publish real market ranges and what drives them; this converts |
| “X vs Y” | Commercial comparison | Blog | Give an honest verdict with the case for each side |
| “best … singapore” | Commercial, late | Blog (listicle) | Rank others credibly; a self-serving list will not rank |
| “[service] agency singapore” | Commercial head term | Service page | Do not write a blog post for this — it cannibalises |
| “[brand] + term” | Navigational | Existing page | Nothing new needed |
Two Singapore-specific notes. “Cost” and “price” queries are unusually valuable here relative to their volume, because Singapore buyers research price hard before enquiring — but they only work if you publish real ranges. A cost post that says “it depends” is worse than no post. And local qualifiers matter more than the volume suggests: 320 searches for “local seo singapore” from people in Singapore is worth vastly more than a global term with ten times the volume and no local buying intent. Our post on local SEO in Singapore covers where that geography shows up.
Scoring the queue
This is the step that converts research into a plan. Score each candidate out of 5 on five criteria, sum, and publish in descending order. It takes about ninety minutes for forty topics and it is the highest-leverage ninety minutes in the whole programme.
| Criterion | What a 5 looks like | What a 1 looks like |
|---|---|---|
| Proof | We have first-party data, a case study or direct operating experience | We would be summarising other people’s opinions |
| Commercial proximity | A reader with this question is weeks from buying | Interesting, but the reader will never buy anything |
| Winnability | Low difficulty, weak incumbents, we have a clear angle | Page one is national publishers and Google’s own docs |
| Cluster fit | Strengthens a cluster we are already building | Isolated; nothing links to it and it links to nothing |
| Durability | Still true in two years with a light refresh | Obsolete next quarter |
Note what is absent: search volume is not one of the five. Volume is a tiebreaker between two topics that score the same, not a criterion. Note also that proof comes first. With the clicks available to generic informational pages falling, the only durable advantage is having something to say that a model cannot synthesise from other people’s pages — your own numbers, your own cases, your own operating experience. If a topic scores 1 on proof, it does not matter what it scores on volume.
A worked read: a post on the Singapore digital marketing grant landscape scores high on commercial proximity and winnability (1,900 searches at KD 17) but only moderately on durability, because grant terms change — Enterprise Singapore has said the EDGE framework will consolidate EDG, PSG and MRA from the second half of 2026, with the existing schemes accessible until it launches. That is not a reason to skip it. It is a reason to write it and diary a refresh, which is precisely the judgement a volume-sorted list cannot make.
Which keywords are no longer worth a post
This is new since 2024, and most keyword research advice has not caught up. Some queries that used to justify a post no longer justify one, because the click is not there to be won.
Pew Research Center analysed 68,879 Google searches by 900 US adults in March 2025. Where an AI summary appeared, users clicked a traditional search result in 8% of visits, against 15% where no summary appeared. They clicked a link inside the summary in only 1% of visits, and ended their browsing session entirely in 26% of visits versus 16%. That is US data and Singapore’s behaviour will differ — but the direction applies everywhere the feature is live.
The kinds of queries that get resolved in the answer, with no click:
- Single-fact lookups. “What is the GST rate in Singapore”, “how many public holidays does Singapore have”. The answer is one line, and it will be given.
- Pure definitions with no decision attached. A standalone “what is a landing page” post is now, for most businesses, a bad use of a slot — fold definitions into the pages where the decision is made.
- Generic how-tos with no local or first-hand element. If the answer is the same in Singapore as in Sydney and you have no operating detail to add, you are writing a page that already exists a thousand times.
What still earns clicks: anything requiring judgement or trade-offs, anything with Singapore-specific rules or prices, anything backed by data only you hold, and anything where the reader needs to trust the source before acting. Add a fourth: being the page the AI answer cites. Our guides to AI Overviews in Singapore and how to show up in AI Overviews cover what makes a page citable, and tracking AI search traffic covers the measurement gap.
Set against Ahrefs’ December 2023 finding that 96.55% of roughly 14 billion pages in its Content Explorer index get zero Google traffic — with Ahrefs noting its sample skews to the higher-quality end of the web — the case for ruthless topic selection is not a stylistic preference. It is the difference between publishing and publishing into a void.
Build clusters, not lists
The last structural point. Forty unrelated posts perform worse than twenty-five posts covering three subjects thoroughly, because coverage of a subject is itself a signal — to search engines choosing what to trust, and to a human deciding whether you know what you are talking about.
Practically: pick two or three subjects tied to what you actually sell. For each, identify one pillar (the comprehensive overview), two or three cornerstones (cost, how-to, comparison) and six to ten spokes (specific questions). Sequence pillar first, then cornerstones, then spokes, and link them deliberately — every spoke up to the pillar, the pillar down to every spoke, and sideways between siblings. The SEO guide for Singapore businesses covers the structural side, and the on-page SEO checklist covers what each page then needs.
This is also where cadence and research meet. A cluster of twelve posts at two a month is a six-month commitment; at four a month it is a quarter. Deciding cluster size without deciding cadence produces a plan that stalls halfway — which is why our guide to how often to publish should be read alongside this one.
Keeping the queue alive
A keyword research exercise is not an annual event. Three habits keep the queue current at almost no cost:
- A monthly Search Console pass. Twenty minutes. Look for queries gaining impressions that no page targets — those are briefs the market handed you.
- A standing capture channel. One shared document where sales, support and delivery drop questions as they hear them. Most of the best posts we publish start here.
- A quarterly cull. Delete queue items that no longer score. A topic that has sat unwritten for four quarters is telling you something.
And keep the refresh candidates in the same queue as the new topics, competing on the same score. An existing page that ranks at position 12 for a term with real intent will usually outscore a new post on a fresh topic — the audit method is in how to do an SEO audit, and the economics are in measuring content marketing ROI.
Frequently asked questions
How is content keyword research different from normal keyword research?
Normal keyword research asks which single term a page should own. Content keyword research is a portfolio decision: given roughly forty publishing slots over a year, which topics in which order produce the most compounding value. It weights coverage, available proof, sequencing and effort far more heavily than volume, and it deliberately routes commercial head terms to service pages rather than blog posts, because a blog post targeting your service page’s term competes with it.
Why are Singapore keyword volumes so low?
Because the market is 5.88 million people, and volume scales with population. Real figures from our SEMrush Singapore data: “seo cost singapore” 720 searches a month at difficulty 17, “local seo singapore” 320 at 13, “multilingual seo singapore” 110 at 6 — while a whole family of terms including “google ads cost singapore” records no volume at all in the Singapore database. A term with 110 monthly searches and low difficulty attached to a high-value service is a good target here, not a rounding error.
Should I write a post for a keyword with zero search volume?
Often yes. Zero recorded volume usually means the tool cannot see it, not that nobody asks it — tools under-report small markets, long conversational queries and newly emerging terms, and questions asked of AI assistants do not appear in keyword tools at all. Write it if sales hears the question regularly, if it maps to a real buying decision, and if you can answer it with something specific. Do not write it if the only reason is that it looks easy.
Which keywords are not worth writing about any more?
Single-fact lookups, standalone definitions with no decision attached, and generic how-tos with no local or first-hand element. Pew Research Center found that across 68,879 searches by 900 US adults in March 2025, users clicked a search result in 8% of visits where an AI summary appeared against 15% where none did, and clicked a link inside the summary in just 1%. Queries whose answer fits in a summary no longer reliably produce a visit. Judgement, trade-offs, Singapore-specific rules and first-party data still do.
How many keywords should be in a content plan?
Enough for twelve months of your actual cadence, plus about 30% slack — so roughly 30 to 60 scored topics for most Singapore SMEs. More important than the count is the shape: two or three clusters covered thoroughly beats forty unrelated posts, because coverage of a subject is itself a credibility signal to both search engines and human readers.
Can I trust the search volumes in SEMrush or Ahrefs for Singapore?
Trust them for relative comparison within the same tool and market; do not treat the absolute numbers as precise. Third-party volumes are modelled from clickstream data, and in a market Singapore’s size small absolute numbers carry large relative error — the gap between a reported 40 and 70 is not a planning signal. Google Keyword Planner comes from Google but returns bucketed ranges without meaningful active ad spend, and groups similar terms. Search Console is the only source showing your actual queries, but only for terms you already rank for.
Where to go next
Once a topic is chosen, how to write a blog post that ranks in Singapore covers turning it into a finished article, and topical authority and content clusters covers the order to publish them in.
With a scored queue in hand, the next decision is cadence: how often should you publish turns available hours into a schedule you can hold. Once posts are live, repurposing content is how you get more than one impression surface out of each. If you are still weighing whether the blog leg earns its budget, does blogging still work for SEO in Singapore is the direct answer, and for the tool-level mechanics, our keyword research guide covers what this piece assumes.
If you would rather have the research, the queue and the writing handled together, that is what our content marketing services in Singapore do — and you can see what that has produced for other Singapore businesses in our case studies.
Keyword figures cited are from our own SEMrush Keyword Magic harvest on the Singapore database and are estimates, as all third-party volumes are. On grants: ongoing marketing retainers and ad spend are generally not claimable; only pre-approved solutions under the Productivity Solutions Grant are, SDM is a pre-approved PSG vendor, and clients apply for and manage the grant themselves. Enterprise Singapore has said the EDGE framework will consolidate EDG, PSG and MRA from the second half of 2026; existing schemes remain accessible until it launches. Check current terms before planning against them.
