1 of 8 Locksmith Sites Marks Up Name and Address

A locksmith sells one thing: a stranger arriving at your door and being let in. So we expected the websites of Singapore locksmiths to be loud about who they are. On 11 October 2026 we ran four locksmith searches through our own metasearch, picked the locksmiths' own sites out of the results, and read the code behind their homepages.

Of the 8 homepages whose HTML we could read, 1 declared a business entity carrying both of the properties Google lists as required for a local business. Three declared a business entity at all. Four declared none.

That is not a small technical nicety for this trade. It is the part of the page that tells a machine which business this is, where it operates and what number to ring.

What we counted, in one table

Every number below comes from the four searches and the page fetches described in the method section, all on 11 October 2026. Counts always carry their base, because the bases differ: we could audit 11 sites, read 9 homepages and parse the raw code of 8.

#	What we measured	The count
1	Homepages declaring a LocalBusiness-family entity	3 of 8
2	Homepages using schema.org's Locksmith type	2 of 8
3	Homepages declaring no business entity of any kind	4 of 8
4	Homepages with no structured data at all	1 of 8
5	Declared business entities carrying both name and address	1 of 3
6	Homepages showing at least one Singapore phone number	8 of 8
7	Homepages showing a street address with a postal code	2 of 8
8	Homepages showing any price figure in readable text	0 of 8
9	Homepages with a heading phrased as a buyer's question	2 of 8
10	Sites that refused our requests entirely (HTTP 403)	2 of 11
11	robots.txt files naming any of five AI crawlers	0 of 10
12	Homepages whose server answered in over 1,500ms	2 of 9

Row 8 is the one most owners will argue with, so read row 10 beside it.

Which searches, and which sites we measured

We used four phrases that Google's own autocomplete returns in Singapore: "locksmith singapore", "locksmith singapore price list", "locksmith singapore 24 hours" and "cheapest locksmith singapore". All four came back as live suggestions in the same run, so these are phrases people type, not phrases we invented.

The four searches returned 54 results. Four were about something else entirely, including two encyclopedia entries and two auto-generated blog pages, leaving 50 on-topic slots. Of those, 28 were a locksmith's own website and 22 belonged to somebody else: directories, review blogs, best-of listicles and a marketplace category page.

The split moved sharply by search. On "locksmith singapore 24 hours", 10 of 15 slots were a locksmith's own site. On "locksmith singapore price list", only 6 of 15 were, and 9 belonged to publishers. We put numbers on that pattern across other trades in best-of lists hold 28 of 60 slots.

From the 18 distinct locksmith domains we found, we attempted a live audit on 13. Eleven returned data. Two of those served us a 403 page instead of a homepage, so nine homepages were readable, and we could pull and parse the raw HTML of eight.

How many homepages declare a business at all

Three of the eight. One declared a generic Organization with no local-business type. Four declared nothing that identifies a business, and one of those served no structured data whatsoever.

The four that declared nothing were not empty pages. They carried WebPage, WebSite, BreadcrumbList and SearchAction markup, the set a popular content plugin adds by default. In other words, the code describes the website. Nothing in it describes the company.

Two of the eight used schema.org's Locksmith type, which exists for exactly this trade and sits under LocalBusiness in both of its inheritance paths. We did not expect that number to be as high as two, and in our view those two sites did the single most useful markup job in the sample.

Only one carried both properties Google calls required

Of the three declared business entities, one had both a name and an address. Google's local business structured data documentation, fetched the same day, lists exactly two required properties: address, "The physical location of the business", and name, "The name of the business".

The other two declared no address property at all. On one of them, the name field held a telephone prompt rather than a company name, so the only thing the markup could tell a machine about the business was a phone number formatted as a title.

Before anyone reads that as pure sloppiness, there is a real complication here. A locksmith who drives to you has no shopfront a customer visits, and marking up a home address invites a different set of problems, which we wrote about in why Google keeps suspending HDB home businesses. The one site that did supply an address handled it the sensible way: a PostalAddress giving the locality and country, plus an areaServed of Singapore, plus a telephone number. No street, no pretence of a shop.

So the fix is not "publish your home address". The fix is to declare a business with a name, a contactable number and a service area, in a type that says what the business does.

Two sites would not let us in

Two of the 11 audited sites answered with HTTP 403 instead of a homepage, and both still refused when we retried with an ordinary desktop browser user agent. We could not measure a single thing about their content.

Both of them are worth noticing for a second reason: one advertised "From S$80" in its search result title and the other, which blocked our direct fetch while allowing our audit crawler, advertised "from $39". The two sites in the sample that put a price where a buyer could see it before clicking were also the two we could not read.

We will not guess why a server refuses a request. A firewall rule, a hosting default and a bot filter all look identical from outside. What we will say is that if a page cannot be fetched, the systems that summarise the web cannot read it either, which is the argument in check the bot gets in before you buy AI visibility.

What every page did get right

All eight readable homepages showed at least one Singapore phone number, usually in the header and often as the first thing on the page. All nine audited homepages served their text in the HTML rather than building it with JavaScript, which means a crawler that does not run scripts still sees the words.

Every one of the 11 sites was on HTTPS. Of the 10 robots.txt files we could fetch, none named GPTBot, OAI-SearchBot, ChatGPT-User, PerplexityBot or ClaudeBot, and none carried a blanket disallow. That matches the wider pattern in our robots.txt census of 207 Singapore sites: the door is almost never locked on purpose.

Two findings sat on the weaker side. Two of nine homepages took over 1,500ms for the server to respond at all, one of them 3,045ms, before a browser could begin drawing anything. And five of nine had at least one image with no alt text.

Nobody published a price where we could read it

Not one of the eight readable homepages contained a price figure. No dollar amount, no "from" rate, no range. We checked for dollar signs, the letters SGD and the word dollar across the full visible text of each page, and the only dollar sign we found was a decorative bullet in a list of safe brands.

Three of the eight linked to a price somewhere else: a dedicated price page, a cost-guide article and a cost calculator. One put the nearest thing to a commitment on the homepage itself, in words rather than numbers: charges confirmed before dispatch.

Compare that with the cleaning trade, where nine of ten sites showed a price and five showed their minimum job, as we reported in 9 of 10 cleaning sites show a price. Here it was none of eight, in a trade where "price list" and "cheapest" are among the phrases buyers actually type.

Which headings read like a buyer's question

Five of the eight homepages had at least one heading ending in a question mark. Only two had a heading that reads like something a buyer would type.

The distinction matters more than the raw count. "Need a locksmith?" and "What we can provide?" are question marks attached to a sales prompt. "How much to change a lock in Singapore?" and "Locked out of your home?" are the buyer's own words, and the second sentence underneath them can be the answer.

The two sites that did this were also the two using a local-business type correctly, which is probably not a coincidence. Someone on those sites had thought about the reader.

What we would fix first if this were our site

Four things, in this order, and the first three are an afternoon's work.

• Declare the business. One JSON-LD block with @type of Locksmith, a real company name, a telephone number and an areaServed. If you have a workshop, add the address. If you do not, leave the street out rather than inventing one.
• Put a number on the page. A callout fee, a from-price for a standard door, or a stated range with what moves it. Every search we ran contained money words. The buyer is pricing you with or without your help.
• Rewrite two headings as questions you are actually asked. Put the answer in the first two sentences underneath.
• Check whether your own server answers. Open your homepage from mobile data with the browser cache cleared. If it takes a beat, that is the beat a locked-out buyer spends going back to the results page.

Our SEO guide for locksmiths in Singapore covers the Maps, reviews and trust side that this study did not measure. The buyer-side half of today's work is in our note on how Singapore lock buyers actually phrase it.

Method and limits, read these before quoting

• Metasearch is not Google. Results came from SingRank's self-hosted SearXNG on 11 October 2026. Most of its engines errored that day and the ordering is blended, not Google's ranking and not its Maps pack.
• Four searches, one morning. A snapshot, not a trend.
• Three different bases. 11 sites audited, 9 readable homepages, 8 homepages whose raw code we parsed ourselves. We quote the base with every count and never blend them.
• One site excluded from the schema counts. Our audit tool flattens compound @type arrays, which undercounted types on two sites when we checked by hand. So the schema figures use only the 8 homepages we parsed from raw HTML, not the audit's summary.
• Homepages only. A site may declare a business entity on a contact page, or publish a price list we did not open. We measured the page a search result sends you to.
• No mobile PageSpeed finding. The scoring step timed out on most sites, so we have a usable mobile score for one. We report server response time instead, which came back for every site.
• No business is named. This page describes practices in an anonymised sample and is not a rating of any company.

Our method page sets out how we separate findings from assumptions.

Want these checks run on your page?

Send us the one page that should win "locksmith singapore" or whatever its equivalent is in your trade. We will read its code, tell you which business properties are missing and what the first fix is. It starts as a free visibility check through our contact page, and our SEO, AEO and GEO programme starts from S$500 a month if you want us to do the work.

One thing we will not promise: that correct markup gets you cited by an AI assistant. Those platforms choose their own answers. Markup makes you legible, not chosen.

Findings describe pages fetched on 11 October 2026. This page names no business.

FAQ

Does a locksmith need LocalBusiness schema to rank?

No. Google's documentation treats structured data as a way to make business details eligible for certain features, not as a ranking guarantee. In our sample it was simply the cheapest unclaimed fix: four of eight homepages declared no business at all, which costs an afternoon to correct.

Should a mobile locksmith publish a street address?

Not if there is no premises a customer visits. In our sample the one entity that handled this well declared a locality, a country, a service area of Singapore and a telephone number, with no street. That tells a machine where you operate without inviting a listing suspension.

Is schema.org's Locksmith type real?

Yes. We fetched the page on 11 October 2026. It defines the type as "A locksmith" and places it under HomeAndConstructionBusiness, which sits under LocalBusiness in both inheritance paths that schema.org lists. Two of the eight homepages we parsed used it.

Why did two sites return 403?

We do not know, and we will not guess. A firewall rule, a hosting default or a bot filter all look the same from outside. What matters commercially is that both refused a normal browser user agent too, so this was not a crawler-only block.

Why is no price the most interesting finding?

Because every search we ran carried money words, including "price list" and "cheapest". None of the eight readable homepages answered that with a figure. When a page stays silent, the publisher listicles holding nine of fifteen slots on the price search answer for it.

Can I cite these figures?

Yes. Please cite "SingRank, Singapore locksmith homepage markup study, 11 October 2026" with a link to this page. Quote every count with its base, such as 1 of 8, and keep the metasearch caveat. Corrections are welcome through our contact page.