Only 4 of 13 Aircon Sites Put a Price in the Title

A homeowner in Bedok types "aircon servicing price" at nine at night, reads a list of ten results, and picks two or three to open. What she reads in that list is not your website. It is your title tag and your meta description: roughly 60 and 155 characters that you chose. In a trade that competes on the price of one fan coil, we wanted to know how many aircon companies put that price where she can see it before she clicks.

Four of thirteen did. On 28 September 2026 we ran three Singapore searches through our own metasearch, kept the aircon companies' own websites, and fetched 15 of them. Thirteen answered. Four carried a price figure in the title tag, five carried one in the meta description, and six carried one in either. Seven showed no price at all in the two fields a buyer reads first.

The second number surprised us more. The slowest of the thirteen servers took 8,172 milliseconds to return its homepage. The fastest took 116. Same trade, same island, a 70-fold difference before a single pixel appears.

Key figures from 13 Singapore aircon websites

Every figure below comes from our own crawl on 28 September 2026. Counts are always shown with their base. Percentages are rounded.

#	What we measured	The number
1	Sites with a price figure in the title tag	4 of 13
2	Sites with a price figure in the meta description	5 of 13
3	Sites with a price figure in either field	6 of 13
4	Sites declaring the HVACBusiness schema type	5 of 13
5	Sites declaring any LocalBusiness-family entity	9 of 13
6	Sites declaring no business entity at all	2 of 13
7	Homepages returned in 0.8 seconds or less	5 of 13
8	Homepages slower than 1.8 seconds	5 of 13
9	Homepages that took 4 seconds or longer	4 of 13
10	Median server response time	955 ms
11	Sites whose homepage needed JavaScript to read	0 of 13
12	Sites where robots.txt returned 200	13 of 13
13	Sites of the 15 fetched that refused our crawler with HTTP 403	2 of 15

Row 3 is the one we would put in front of an aircon boss. Row 8 is the one we would put in front of his web developer.

Why a price belongs in an aircon title tag, and not only on the page

Because the decision happens in the results list. Aircon servicing is a small, repeat, price-led purchase. Nobody shortlists five companies and builds a comparison table for a $60 job. They read the ten lines Google gives them and open two.

That makes the title tag and meta description the only part of your site most searchers ever see. Google decides what appears there and can replace it: its documentation lists the <title> element first among the sources it uses to determine a title link, and says that "if we've detected an issue on the page, we may try to generate an improved title link from anchors, on-page text, or other sources" (Google Search Central, title links). So the title you write is an input, not a guarantee. It is still the only input you control.

We have measured the other half of this before. Our study of 211 Singapore service websites counted prices printed on the page. This one counts prices printed in the search result. They are different jobs, and a company can do the first and skip the second, which is exactly what we found.

How many showed a price before the click?

Six of thirteen showed a price figure in the title tag, the meta description, or both. Four had it in the title, five in the description, and three had it in both places.

The four titles that carried a price used it as the whole proposition: a per-unit figure, sometimes with "same-day" beside it. The nine that did not used the space differently. Three spent it on a superlative — "No.1", "#1 Rated", "Top-notch" — and the rest on a service list. Counting all 13 titles, four leaned on a superlative, and one of those four also carried a price.

In our view that is a bad trade. A superlative is a claim every competitor in the list is also making, so it carries no information. A number is the one thing on the line that a buyer can act on, and the only thing that separates two identical-looking results.

One detail is worth copying. Two of the four price titles named the unit as well as the figure, so the buyer knows whether the number is per fan coil or for the whole flat. A bare "from $20" invites a phone call that starts with a correction, and those calls do not convert well.

Which schema type did these sites declare?

Nine of thirteen declared a business entity in the LocalBusiness family. Five of those nine used the specific type schema.org reserves for this trade: HVACBusiness, which schema.org defines as "A business that provides Heating, Ventilation and Air Conditioning services". Four declared the generic LocalBusiness instead.

Two sites declared only an Organization, which says a company exists but not that it serves a place. Two declared no business entity at all: one had no structured data anywhere on its homepage, and one had only page-level markup.

Do not read too much into the subtype. Going from LocalBusiness to HVACBusiness is a one-word edit and will not move a ranking by itself. The reason we measured it is that it reveals who thought about the markup and who let a plugin decide. The gap that matters is the second one: four of thirteen homepages gave a machine nothing to attach a business to. Our guide to schema markup for AI search covers what to declare and in what order.

A 70-fold spread in how fast these servers answered

Eight of thirteen homepages came back in under 1,500 ms, the threshold our audit tool uses. Four took four seconds or more, and the slowest took 8,172 ms. The median was 955 ms.

Google's own guidance on Time to First Byte puts "good" at "0.8 seconds or less" and "poor" at "greater than 1.8 seconds", measured at the 75th percentile of real visits (web.dev, Time to First Byte). Our figure is one lab request per site, not a field percentile, so the two are not the same measurement. Read loosely against those bands anyway: 5 of the 13 came back inside 0.8 seconds, 3 sat in the middle, and 5 were slower than 1.8 seconds.

This is one sample per site, taken once on one morning, so treat an individual figure as a snapshot rather than a verdict. The spread is still too wide to be noise. Four seconds of waiting before any content renders is not a tuning problem; it is usually a hosting or platform problem.

Aircon is an impatient purchase. The searches that convert are the ones typed by someone whose bedroom is 30 degrees, and that person has nine other results on the page. Slow server, back button, next company.

Two of the four slowest sites also listed 900 or more URLs in their sitemaps, against a median of 55 across the twelve sites that had one. A large site on a slow host is the combination we would fix first, and it is a fix your developer does, not your marketer. Our Core Web Vitals guide explains what the field data measures once the server is out of the way.

What the sitemaps showed, including one page nobody at the business wrote

Twelve of thirteen sites had a reachable sitemap.xml. One did not, which means its pages get discovered through links alone.

Two sitemaps listed 1,000 URLs or more, and a third listed 943. Against a median of 55, that is a different kind of website: one with a page per job, per town or per unit model. Scale is not the problem. Unattended scale is.

On one of the large sites, the second URL our crawler picked from the sitemap returned a page whose own title announced it had been created by someone outside the business. It carried no service content. We are naming nobody, and we have no idea how long it had been there, but the lesson generalises: if your site publishes 1,000 URLs and nobody reads the sitemap, you do not know what is on your website.

Open yours today. It takes two minutes and it is the cheapest check on this page.

The three fixes we would make this week

In this order, for an aircon company with a working site:

1. Put your entry price and the unit in the title tag of the page that should win "aircon servicing price". Something in the shape of "Aircon Servicing Singapore from $X per unit". Rewrite the meta description to repeat the figure and add one differentiator you can prove, such as same-day slots or a warranty length.
2. Check your server's response time on a mobile connection. If the HTML takes more than about a second to start arriving, talk to your host before you buy anything else. No amount of content fixes a four-second wait.
3. Open /sitemap.xml and read it. Then declare an HVACBusiness entity on the homepage with a name, address and phone that match the footer exactly.

Nothing here needs a rebuild. Items one and three are text edits.

What it costs, honestly

Publishing a price costs you flexibility. Once "from $40 per unit" is in your title tag, you cannot quietly quote $65 to a customer in a condo with awkward access, and you will lose some enquiries from people who would have paid more had they called first. Some aircon firms make that trade deliberately and keep the phone for complicated jobs. That is a legitimate strategy, and if it is yours, then rows 1 to 3 of our table are not your problem.

The other cost is honesty about what these seven checks buy you. They are hygiene. They make your result readable and your page quotable. They do not create demand, they do not beat a competitor with 2,000 reviews, and they will not put you above a listicle that has been collecting links for six years. If your real problem is that nobody knows your name, fix the name problem first, and our guide on when SEO is the wrong thing to buy says when to walk away from search spend entirely.

Method and limits — read these before quoting

We would rather you know exactly what this sample carries.

• Three searches, one morning. "aircon servicing singapore", "aircon chemical wash singapore price" and "aircon repair singapore same day", run on 28 September 2026.
• Metasearch, not Google. Results came from SingRank's self-hosted SearXNG. Between seven and eight of its engines returned errors on each query, so the ordering is a blend of what remained and is not Google's ranking or its Maps pack.
• 60 results read. 41 slots were an aircon provider's own page. The other 19 were listicles, booking marketplaces, auto-generated blog spam, social platforms or businesses outside Singapore.
• 29 provider domains appeared. We audited 15. We chose the highest-placed firms that sell aircon servicing directly, and skipped booking marketplaces and electronics retailers.
• Two of the 15 refused us. Both returned HTTP 403 to our crawler, so all counts above have a base of 13, not 15. A 403 to us is not proof of a block on Google or on AI crawlers; it means we could not measure.
• One site redirected. A well-known aircon domain redirected to a parent home-services brand, so the page we measured was its aircon service page rather than a homepage.
• Up to two pages per site. Mostly the homepage plus one sitemap entry, with raw HTML only. No JavaScript rendering, no field speed data, no review counts, no Maps position.
• Server time is a single sample. One request each, from one network, on one morning. Google's TTFB bands describe the 75th percentile of real visits, so our figures are indicative only and cannot be read as a field measurement.
• No business is named. This page describes a sample of pages, not a rating of any company.

Our method page sets out how we separate findings from assumptions. Our earlier AI search readiness study found aircon servicing the weakest of the four trades it sampled, which is part of why we came back to it with a different set of checks.

Want these seven checks on your own page?

Send us the one page that should win "aircon servicing price" for your company. We will run the same seven checks on it, in writing, and tell you which one to fix first. Those seven: a price in the title tag, a price in the meta description, a business entity in your schema, server response time, whether the page reads without JavaScript, whether robots.txt answers, and whether your sitemap is reachable and clean. That starts as a free visibility check through our contact page.

If you would rather we did the work, our SEO, AEO and GEO programme starts at S$500 a month on the Premium Starter tier, with a three-month minimum. The trade-specific playbook is in our guide to SEO for aircon servicing companies.

What we will not do is promise a ranking. We can make your result readable and your page fast. Where it lands is Google's decision.

These findings describe 13 websites sampled on 28 September 2026 through metasearch. This page names no business and is not a rating of any company.

FAQ

Should an aircon company put its price in the title tag?

In our view yes, if you are willing to hold the figure. Aircon servicing is a small repeat purchase decided inside the results list, so a per-unit number is the only differentiator a buyer can act on before clicking. Four of the 13 sites we measured did this; seven showed no price in either field.

What is the HVACBusiness schema type?

HVACBusiness is the schema.org type for a heating, ventilation and air conditioning business. Schema.org defines it as "A business that provides Heating, Ventilation and Air Conditioning services", and places it under LocalBusiness. Five of the 13 sites we measured declared it; four used the generic LocalBusiness type instead.

How fast should an aircon website respond?

Google's guidance calls a Time to First Byte of 0.8 seconds or less good, and anything above 1.8 seconds poor, measured across real visits. In our single-request sample of 13 aircon homepages, 5 answered inside 0.8 seconds and 5 took longer than 1.8 seconds. Server response sits upstream of every other speed fix.

Why did two websites return HTTP 403 to your crawler?

We do not know. A 403 can come from a firewall rule, a content delivery network, or bot filtering that does not distinguish our crawler from anything else. It does not prove Google or an AI crawler is blocked. It does mean an automated check of those two sites was impossible, so we excluded them from every count.

Does this study say which aircon company is best?

No. We measured seven technical and presentational signals on up to two pages per site, and named no business anywhere. Nothing here speaks to workmanship, punctuality, pricing fairness or safety. A site that passes all seven checks can still send a poor technician to your flat.

Can I cite these figures?

Yes. Please cite "SingRank, Singapore aircon website study, 28 September 2026" with a link to this page. Quote each figure with its base, such as 4 of 13, and carry the metasearch caveat. If you think a number is wrong, tell us through our contact page and we will recheck it.