Preschool Fees Singapore: 2 of 11 Sites Show a Price

Preschool fees in Singapore are the one thing a parent researches before anything else, and the sector's reference price is already public. ECDA publishes the monthly fee caps that Anchor and Partner Operator centres agree to keep to, so a parent knows roughly what childcare should cost before opening a single centre's homepage.

We went looking for the other side of that conversation. On 1 October 2026 we tried 12 Singapore preschool, childcare and infant care websites that appeared for real fee and location searches. Two of the 11 we could read showed any fee figure on the homepage, and none of the 11 put a price or the word "fee" in the title tag.

That is the gap this study is about. Not a missing price list. A missing reply to a question the parent has already answered for themselves, using someone else's page.

Key findings on preschool fees and Singapore centre websites

Every figure comes from 12 websites sampled on 1 October 2026 through our own metasearch, and each count is given with its base.

What we checked	Result
Homepages showing any fee figure	2 of 11
Titles containing a price or the word "fee"	0 of 11
Homepages mentioning fees, price or cost in words anywhere	8 of 11
First paragraph that names what the centre does	3 of 11
First paragraph naming Singapore or a town	5 of 11
Question heading or an FAQ block	9 of 11
Business-entity structured data	6 of 11
A clear way to enquire	11 of 11
WhatsApp as one of those routes	8 of 11
robots.txt served at all	10 of 11
robots.txt blocking any of 9 AI crawlers	0 of 10
Sites returning a bot challenge instead of a page	1 of 12

If you read one row, read the first two together: parents are asking what it costs, and almost nobody answers in the place they look first.

What we measured, and why these 12 sites

We ran five searches through SingRank's self-hosted SearXNG metasearch on 1 October 2026: "preschool singapore fees", "infant care singapore price", "childcare singapore near me", "best preschool singapore" and "preschool singapore". All five are real Google autocomplete phrases, confirmed in the same run with our keyword tool.

From those results we kept the websites that belonged to a centre or school operating in Singapore. We dropped directories, parenting and money publishers, listicles, government pages and anything about another country. That exclusion mattered more here than in most trades, and we come back to it below.

Then we fetched each homepage with plain HTTP and no JavaScript, because that is roughly what a crawler sees. We checked the first substantive paragraph, the title tag, whether any fee figure appeared, the headings, the structured data, the enquiry routes and robots.txt. We also ran our own audit crawler on 5 of them for speed and link health.

Eleven sites returned a readable page. One returned a 202 response of 169 bytes, a bot challenge rather than a page, so it is counted as attempted and not measured.

Did any homepage show a fee? Two did

Two of the 11 readable homepages carried a fee figure. Eight more used the words fee, price or cost without a number next to them, usually as a link to a page further in, or a line saying subsidies are available.

That last pattern is worth naming, because it looks like transparency and is not. "Subsidies available" tells a parent nothing they did not already assume. It answers no question. It is the childcare equivalent of "competitive rates".

Our wider study of 211 Singapore service websites found 99 showing a price, a little under half. Against that, 2 of 11 is poor even by the standards of trades that hate publishing numbers. We would not read a strong trend into 11 sites. We would read a direction.

Not one of the 11 titles mentioned price or fees

Zero of 11. The title tag is the line a parent reads in the results before deciding whose page to open, and on a fee search it is competing against pages built for exactly that search.

Look at what held those slots. For "infant care singapore price", 15 results came back and 3 were a centre's own website. The rest were parenting publishers, personal-finance blogs, directories and one aggregator page. For "preschool singapore fees", 18 results came back; the majority were publishers, directories, a government fees page and, in one case, a 2013 article about licensing fees for news websites.

So on the searches where a parent is asking about money, the answer they get is mostly written by someone with no child at your centre. We have seen the same shape in other trades, and it has the same consequence: a stranger's estimate of your fees becomes the first number a parent associates with your name.

One result deserves a separate mention. A page in the "infant care singapore price" set was a scraped copy of another centre's fee page, hosted on an unrelated domain, with spam text injected into the title. We name nobody, and we are not alleging anything about the original site. We are pointing out that an unanswered fee search is a slot somebody else will fill, and not always somebody careful.

The number parents already have

Here is why this matters more in childcare than in most trades. The reference price is not a secret anyone is guarding. It is published by the regulator.

ECDA's Anchor Operator page states that participating operators agree to "keep to a monthly fee cap of $610, $1,235 and $150 (excluding GST) for full-day child care, full-day infant care and kindergarten respectively". The Partner Operator page sets caps of "$650 and $1,290 (excluding GST) for full-day child care and infant care programme respectively for Singapore Citizen (SC) children", and says ECDA has appointed 380 childcare centres under the term running from 1 January 2026 to 31 December 2030.

Most of the centres in our sample are not in either scheme, and their fees are not capped. That is the point. A private centre charging well above $650 is still being compared against $650, because the parent has read that figure somewhere before arriving. Staying silent does not remove the comparison. It only removes your side of it.

In our view this is the strongest argument for publishing a number that exists in any Singapore industry. A centre outside the schemes has a real answer available: here is our fee, here is what the extra buys, here is who it suits. A page that says nothing leaves the parent to assume the difference is margin.

Does the first paragraph say what the centre does?

Three of 11. We read the first paragraph of 15 words or more, outside the header, navigation and footer, and asked whether it named the service.

What the other eight opened with is more instructive than the count:

• One described a child as a unique spark of magic, then moved to curriculum philosophy.
• One opened by recruiting staff, inviting the reader to become an educator and grow their career. On the homepage, before anything addressed to a parent.
• One opened with a promotional strip: term enrolment enquiries are open, book a free tour.
• One opened with a stack of fragments rather than a sentence: registered status, languages, class ratios, minutes from an MRT station.
• One led with admission details, centre locations and fees as a list of links, without saying what the centre is.

Five of 11 named Singapore or a town in that first paragraph. Childcare is a walking-distance purchase, and the town is how a parent draws up a shortlist in the first place. In our view it belongs in the opening line, not only in the address block at the bottom.

Our study of 109 local pages in September found 65 of 109 opening with a direct answer. Three of 11 here is markedly worse, on a smaller sample.

Question headings, schema and getting in touch

Nine of 11 had a question heading or an FAQ block, the best result in the set. Six of 11 carried business-entity structured data: the types we saw included EducationalOrganization, LocalBusiness and ChildCare. There is no preschool-specific guide on this site yet, and the closest structural match is our guide for tuition centres, which covers the education-sector basics of headings, schema and enquiry routes.

Every readable site offered a clear way to enquire, 11 of 11, and 8 of 11 offered WhatsApp. Contactability is not this sector's problem. Several had a dedicated tour-booking page, which is the right conversion step for a purchase a parent wants to see with their own eyes.

Nothing blocked an AI crawler. Across the 10 robots.txt files we could read, none disallowed GPTBot, OAI-SearchBot, ChatGPT-User, PerplexityBot, ClaudeBot, CCBot, Google-Extended, Bytespider or Meta-ExternalAgent. One site served no robots.txt at all, which is not an error in itself. The door is open. What is behind it is thin.

What five speed audits showed

We ran our full audit crawler on 5 of the 11. All five scored under 70 on mobile PageSpeed: 34, 57, 60, 65 and 68. Lab LCP ran from 6.2 seconds to 38.1 seconds. Two had real-user field LCP recorded, at 4.1 and 3.7 seconds, both outside the 2.5-second threshold Google describes as good.

Lab LCP figures that high usually mean a homepage carrying a large number of unoptimised images. One sampled homepage had 70 images; another had 51, with 23 missing alt text. Centres photograph everything, understandably, and then publish it all at full size.

One site in the sample had 7 dead external links, and 5 of those pointed at Singapore government pages: three ECDA URLs, one MOE URL and one MSF URL, all returning 404. Citing the regulator is good practice and we recommend it. Government URLs move. If you link to them, re-check them once a year, because a 404 on the page where you explain subsidies is the worst place to lose a parent's trust.

What I would do this week

Four things, in order, none needing a developer.

1. Put a number on the homepage. If you are in a capped scheme, publish the cap and say the scheme's name. If you are not, publish your own monthly fee or a from-range per programme, with GST stated and the registration or deposit named separately. Where subsidies apply, link ECDA's subsidy scheme rather than restating the rules yourself. One honest line beats a contact form.
2. Rewrite the first paragraph. Name the programme, the ages and the town in the first sentence. Put the philosophy second and the recruitment advertisement on a careers page.
3. Build one page for the fee search. Title it the way parents type it, with the programme and the word fees. That page is currently held by publishers who have never seen your centre.
4. Re-check every government link you cite, then compress the homepage photographs.

What this costs you

Publishing a fee costs you the enquiry from the parent who was only ever going to choose on price, and you will feel that loss before you feel the gain. It also commits you to updating the page when fees change, and to explaining an increase in writing. Some operators cannot publish a single figure honestly, because the programme mix genuinely varies by centre and age band; publish the range and the three things that move it.

And a fee page will not fill a centre that is in the wrong place. Childcare demand is a catchment question first. If there are not enough young children within walking distance, a better website changes the arithmetic by very little. There are times when search is the wrong spend, and a half-empty centre two MRT stops from its nearest catchment is one of them.

Method and limits

• Metasearch is not Google. Results came from our own SearXNG instance on 1 October 2026. Six of its engines returned errors that day, so the ordering is a blend of what remained and is not Google's ranking or its Maps pack.
• Twelve sites, 11 measured. Small numbers. One site more or less moves every percentage by about nine points, which is why we give counts, not percentages.
• Homepages only, raw HTML only. A centre may well publish fees on an inner page. We did not crawl inner pages for fees, and we did not run JavaScript, so anything assembled in the browser was not measured rather than absent.
• One sample member is not a preschool. A school for ages 5 to 12 appeared in the preschool fee results through a blog post about preschool fees, and we measured its homepage as it was returned. We have left it in and disclosed it.
• The speed pillar covers 5 of 11, not the whole sample.
• No business is named, and this page is not a rating of any company. No centre in the sample is a SingRank client.
• This is marketing guidance, not legal or regulatory advice. Check any statement about schemes, caps or licensing against ECDA's own pages before you publish it.

One next step

Send us your centre's homepage. We will run these same checks on it and tell you, in plain words, what a parent and a crawler each see first. That starts as a free visibility check through our contact page.

If you want the pages built and measured for you, our SEO, AEO and GEO programme starts from S$500/month on the live page today, with a three-month minimum. We will not promise you a ranking, an AI citation or an enrolment. Google and the AI platforms choose their own answers, and a parent chooses a centre by walking into it.

These findings describe 12 websites sampled on 1 October 2026 through metasearch. Counts will change. This page names no business.

FAQ

Should a Singapore preschool publish its fees online?

In our view yes, with conditions stated. Parents arrive already holding ECDA's published fee caps as a reference, so the comparison happens whether or not your page joins in. Publish the monthly figure or a from-range per programme, say whether GST applies, and name deposits separately.

What are the 2026 preschool fee caps in Singapore?

ECDA's Anchor Operator page states a monthly cap of $610 for full-day childcare, $1,235 for full-day infant care and $150 for kindergarten, excluding GST. Its Partner Operator page states $650 and $1,290, excluding GST, for full-day childcare and infant care for Singapore Citizen children.

Do fee caps apply to every preschool in Singapore?

No. Caps apply to centres in the Anchor Operator and Partner Operator schemes. Centres outside both set their own fees. Most sites in this sample were uncapped, which is why we argue an uncapped centre has more to gain from publishing a number, not less.

How many of these sites blocked AI crawlers?

None. Across the 10 robots.txt files we could read, not one disallowed any of the nine AI crawlers we checked, including GPTBot, PerplexityBot and ClaudeBot. One site served no robots.txt at all. Access was not the limiting factor in this sample; what the pages say is.

Why were only 11 of 12 sites measured?

One site answered our request with a 202 status and 169 bytes, a bot challenge rather than a page. We counted it as attempted and excluded it from every content measurement, rather than record absent signals it may well have.

Can I cite these figures?

Yes. Please cite "SingRank, Singapore preschool website fee study, 1 October 2026" with a link to this page. Quote each figure with its base, such as 2 of 11, and carry the metasearch caveat. If you think a number is wrong, tell us through our contact page and we will recheck it.