SingRankFreight Sites: 0 of 8 Say How Fast They Will Quote
Contents12
- What this study measured, and on which sites
- Key takeaways
- Does any freight website tell a buyer what it costs?
- Not one site said how fast it would answer
- Which sites named a country a buyer might ship to?
- What the technical crawl found
- Two of ten sites would not let us in
- What we would do with a freight website this week
- What it costs
- Method and limits — read these before quoting
- Want these checks run on your own site?
- FAQ
A Singapore importer with a container of machine parts waiting in Shanghai wants two things from your website: a number and a date. On 3 October 2026 we measured ten Singapore freight forwarder and declaring agent websites. Eight could be read in full.
Not one of the eight showed a rate, a rate basis, or how long it takes to come back with a quote. All eight said what they do and where to reach them. Then every one of them asked the buyer to wait an unspecified time for an unspecified figure.
Published freight rates are a different argument, and we are not going to make it. Ocean and air rates move week to week, and a forwarder who prints a number is printing a number that will be wrong. The gap that matters is smaller and cheaper to close: say what you charge per, and say when you will reply. Both are sentences, not rebuilds.
What this study measured, and on which sites
We ran four searches a Singapore buyer would type through our own metasearch on 3 October 2026: "freight forwarder singapore", "customs clearance singapore", "sea freight companies in singapore" and "best freight forwarder singapore". Google Autocomplete returns all four as real queries.
That returned 75 result slots, which we tagged by hand:
| What held the slot | Slots |
|---|---|
| A Singapore freight or customs company's own website | 22 of 75 |
| Directories, listicles and "best of" pages | 23 of 75 |
| A foreign company's own website | 14 of 75 |
| Off-topic entirely (news, forums, academic papers) | 16 of 75 |
From those 22 own-site slots we drew ten sites to measure. One belonged to a business SingRank works for and was replaced with the next result. One more we dropped because our audit tool timed out on it three times, which is our problem rather than theirs.
Each site got two passes: a technical crawl of the home page and up to two more pages, and a reading of the home page against seven checks a B2B buyer cares about. Nine sites returned the technical pass. Eight returned both. Two served HTTP 403 to a fetcher that was not a browser, each to a different one of our two tools.
Key takeaways
Every figure below comes from the passes described above, taken on 3 October 2026. Counts always carry their base, because the bases differ.
| # | Finding | The number |
|---|---|---|
| 1 | Home pages showing any price, rate or rate basis | 0 of 8 |
| 2 | Home pages stating how fast they reply to a quote request | 0 of 8 |
| 3 | Home pages whose opening names the service and Singapore | 8 of 8 |
| 4 | Home pages with a visible phone, email, form or messaging route | 8 of 8 |
| 5 | Home pages naming a specific country or port, not just "worldwide" | 5 of 8 |
| 6 | Home pages with a question heading or an FAQ block | 5 of 8 |
| 7 | Sites stating a licence, registration or accreditation | 3 of 8 |
| 8 | Sites carrying business-entity structured data | 6 of 9 |
| 9 | Sites serving their content in the HTML, no JavaScript needed | 9 of 9 |
| 10 | Sites that returned 403 to one of our two fetchers | 2 of 10 |
Rows 1 and 2 are the ones we would act on. Row 9 is the good news nobody mentions.
Does any freight website tell a buyer what it costs?
No. Across the eight home pages we could read in full, zero showed a price, a rate, or a basis for one. No "per kg", no "per CBM", no chargeable-weight explanation, no minimum. Three used a phrase in the family of "competitive rates" or "affordable prices" without a figure or a unit attached.
We have measured price transparency in consumer trades repeatedly, and our census of 211 Singapore service websites found 99 of them showing a price. Freight does not work that way, and we would not tell a forwarder to publish a rate card.
But "we cannot publish a rate" is not the same as "we cannot explain how charging works". A shipper comparing three forwarders by email is trying to work out whether the quotes are even comparable. Whoever explains chargeable weight, what is excluded, and which charges land at destination has answered the question the buyer brought, and has done it before the quote arrives.
Not one site said how fast it would answer
Zero of eight stated a turnaround for a quote request. The closest any came was a promise to reply "as soon as possible", and two sites offered round-the-clock contact, which is about availability rather than a commitment to a timeframe.
In our view this is the cheapest unclaimed advantage in the sample. A buyer sending the same enquiry to four forwarders on a Tuesday afternoon is running a race nobody told them the rules of. "Rates for most lanes within one working day; unusual cargo may take two" is one sentence, needs no developer, and is the only claim on the page a buyer can check within 24 hours.
It also makes a promise you then have to keep, which is presumably why nobody writes it. That is the point. The cost of the sentence is that your operations team has to honour it.
Which sites named a country a buyer might ship to?
Five of eight named at least one specific country or port. The other three described their reach only as worldwide, global, or a network of a given number of countries.
This matters more in freight than in most trades, because a freight buyer does not have a generic need. They have an origin and a destination. Our companion piece on how Singapore freight buyers actually search found 12 of 46 Autocomplete suggestions naming a country other than Singapore, and none asking for a forwarder nearby.
A page that says "over 200 countries" matches no search and answers no question. A page that names the lanes you are genuinely strong on will miss some buyers and convince the rest. The second trade is better.
What the technical crawl found
Nine sites returned the technical pass. Six carried business-entity structured data, by way of Organization, Corporation or ProfessionalService types. One declared a LocalBusiness. Two carried FAQ markup. Three declared no business entity at all, only page-level and site-level types, which tells a machine that a website exists without telling it whose it is. Our guide to schema for AI search covers the types.
The ordinary on-page housekeeping was weaker than the schema:
| Check (base: 9 sites) | Result |
|---|---|
| Exactly one H1 on the home page | 5 of 9 |
| Home-page title within 65 characters | 6 of 9 |
| A sitemap our crawler could find | 7 of 9 |
| Served over HTTPS | 9 of 9 |
| robots.txt readable | 9 of 9 |
Three home pages carried more than one H1, one of them 13, and one carried none. The longest home-page title ran to 159 characters, which no search result will ever show in full. Across the nine home pages, 256 of 578 images had no alt text.
We could only score mobile PageSpeed on four sites before our quota ran out, so treat this as four data points and not a pattern: mobile scores of 14, 30, 58 and 60, with largest-contentful-paint times of 7.6, 9.2, 10.4 and 19.3 seconds. All four are slow. Our Core Web Vitals guide explains which of those numbers a buyer actually feels.
The genuinely reassuring finding: all nine served their content in the HTML, with no JavaScript required to see the text. AI crawlers do not run JavaScript, so this sample's pages are readable by the systems that increasingly summarise them.
Two of ten sites would not let us in
One site returned HTTP 403 to our audit crawler and served no readable robots.txt to it. A second returned 403 to our page-reading fetcher while the audit crawler got through normally.
We are careful about what this means. A 403 to one non-browser fetcher is not proof that a site blocks AI crawlers, and it is not evidence of intent. It usually comes from a firewall or a CDN rule written for a different reason, and it may not apply to Googlebot at all. We only know what happened to our two requests on 3 October 2026.
It is still worth an owner's hour. The site that refused our crawler also holds the top slot for "best freight forwarder singapore", with a list it published itself, so whatever blocked us plainly is not blocking everything. Our note on checking crawler access before buying AI visibility shows how to test your own site in a few minutes.
What we would do with a freight website this week
Four moves, in the order we would take them.
- Add one sentence about turnaround, above the fold. Name the window for a standard lane and the exception. Nobody in our sample of eight has done it.
- Explain the charging basis, not the price. One short section on chargeable weight, what is excluded and which charges land at destination. It answers the comparison the buyer is already running.
- Name your real lanes. Replace "worldwide" with the countries and ports you file for every week. Three of eight home pages still describe reach only in the abstract.
- Put your registration on the page. Three of eight sites state a licence, registration or accreditation. Singapore Customs is explicit that importers must obtain the relevant customs permit before the goods are imported into Singapore, and a declaring agent account takes about three working days to process for straightforward applications. A buyer who needs clearance next Tuesday cares whether you already hold that account.
Our guide for logistics companies covers the longer programme.
What it costs
The turnaround sentence costs you an operational promise, and a sales team that has to be reachable on a Friday evening to keep it. The charging explainer costs an afternoon with whoever quotes, and it tells your competitors exactly how you price. Naming your lanes costs you the enquiries you were never going to win anyway, which feels like a loss on the enquiry count before it shows up as a gain in conversion.
The 256 missing alt attributes and the four slow mobile scores are a developer's week, not an afternoon, and they are the lowest-value items on this list for a business whose buyers arrive by email. Fix the words first.
And the honest version of our own position: none of the four moves above needs an agency, and none of them is content we could sell you by the article.
Method and limits — read these before quoting
- Metasearch is not Google. The four searches ran through SingRank's self-hosted SearXNG on 3 October 2026, which blends other engines and is not Google's ranking or its Maps pack. Several of its engines returned errors that day, and 16 of 75 slots were off-topic.
- Small sample, different bases. Ten sites attempted, nine measured technically, eight read in full. Mobile PageSpeed covers four. Every count above carries its base for that reason.
- One day, one page each. Mostly home pages, plus up to two more where the crawl reached them. A snapshot, not a trend.
- Our checks are literal. "Shows a price" means a figure or a unit on the page we read. A rate inside a logged-in portal or a PDF would not count.
- We dropped one measurement entirely. Our crawler's meta-description extractor returned page markup rather than the description on one site, so we have left meta-description findings out of this study rather than publish a number we do not trust.
- Two sites we could not fully read. Both returned 403 to one fetcher. We report that as a fact about our requests, not as a finding about their intent.
- One exclusion. A site belonging to a business SingRank works for was replaced with the next result. No business is named anywhere on this page.
Our method page sets out how we separate findings from assumptions.
Want these checks run on your own site?
Send us the page you would want a shipper to land on. We will run the same seven buyer checks and the same technical pass, and tell you which sentence to write first. It starts as a free visibility check, and our SEO, AEO and GEO programme starts at S$500 a month if you want the work done for you.
What we will not do is promise a ranking or an AI citation. AI platforms choose their own answers.
These findings describe ten websites sampled on 3 October 2026 through metasearch and our own crawler. This page names no business and is not a rating of any company. It is marketing guidance, not advice on customs compliance.
FAQ
Should a Singapore freight forwarder publish rates on its website?
Usually no. Ocean and air rates move week to week, so a published figure dates quickly and invites disputes. Publishing the charging basis is the useful half: what you charge per, what is excluded, and which charges fall due at destination. None of the eight sites we read did either.
What is the single cheapest fix on a freight website?
A stated quote turnaround. Zero of the eight home pages we measured said how fast they reply to an enquiry, so one sentence naming a window for standard lanes puts you ahead of every site in this sample. The cost is operational: you then have to meet it.
Does structured data help a freight forwarder get cited by AI tools?
It helps a machine identify who owns the page, which is a precondition, not a guarantee. Six of nine sites carried a business entity and one declared LocalBusiness. Accuracy matters more than presence: schema that contradicts the visible address gives a machine two answers.
Why did two sites return 403 to your crawlers?
We do not know. A 403 to a non-browser fetcher usually comes from a firewall or CDN rule rather than a deliberate block, and it may not apply to Googlebot at all. We report only what happened to our two requests on 3 October 2026, and we recommend owners test their own sites directly.
Is page speed worth fixing for a B2B freight site?
It is real but it is not first. We scored four sites, with mobile results of 14, 30, 58 and 60 and load times from 7.6 to 19.3 seconds. For buyers who arrive by email and compare quotes in a spreadsheet, the words on the page move more money than the milliseconds.
Can I cite these figures?
Yes. Please cite "SingRank, Singapore freight website study, 3 October 2026" with a link to this page, and quote each count with its base, such as 0 of 8 or 6 of 9. Corrections and challenges are welcome through our contact page.