Why Does a Hong Kong Business Website Get Traffic but No Enquiries?

01 / THE FIRST ILLUSION

You measured arrivals, not demand

On Monday morning the dashboard shows 4,800 visitors last month. The number is up. The room feels optimistic—until someone asks how many enquiries arrived.

The answer may be three. It may be zero.

A student researching a topic, a competitor checking prices, a supplier confirming an address and an automated crawler can all create visits. The traffic report is not designed to tell you who has budget, authority and urgency.

Owners rarely need another arrival count. They need to know how many visitors decide that speaking with the company is worth the risk. Those are very different populations. A Hong Kong professional-services spike may come from a careers page, an address page, a downloadable document or an existing client login; one report column buries the valuable minority.

A visitor is not a stable unit either. Several people from one account may appear from the office, home and mobile. When headcount stands in for an account or a buying situation, management sees a large market while sales receives a handful of vague questions.

02 / THE HIDDEN DENOMINATOR

The wrong people may be crushing your conversion rate

Businesses often divide enquiries by all visitors. That denominator is too blunt.

If an educational article is shared overseas, traffic may double without adding a single Hong Kong prospect. The displayed conversion rate falls by half even though the commercial audience did not shrink.

Separate people who could realistically buy from people who merely passed through. Only then discuss conversion.

A better denominator is Hong Kong visitors who reached relevant service, pricing, evidence or contact pages. Better still: accounts that match the company size, sector and need you can serve.

Use a denominator closer to the business. Imagine 4,800 visits, 220 reaching a service, case or pricing page, 24 starting phone or WhatsApp contact, and three becoming qualified opportunities. The site-wide rate is 0.06%; the commercial-visitor rate is 1.4%; the contact-path rate is 12.5%. All three are true, but they lead to three different decisions.

This is not an argument for flattering the numbers. Keep both denominators: total-site health and qualified-demand density. The first warns that audience mix has changed; the second shows which pages actually move a buyer closer.

03 / INVISIBLE ENQUIRIES

The site may influence business your dashboard cannot see

A Hong Kong buyer may copy the phone number, open WhatsApp and write the next day, share the URL in a company group, then have a colleague search the brand again on another device.

Google Analytics counts an important action as a key result only after it is configured. Cross-device behaviour, offline deals and untagged WhatsApp conversations do not automatically assemble themselves into a complete journey.

Create an “unknown source” field instead of forcing staff to choose Google, Facebook or referral. Each week, sample ten enquiries and compare call notes, the first WhatsApp sentence and the CRM record. Some “direct” contacts will have seen the site days earlier. The sample may not look statistically elegant, but it exposes false channel credit.

Also separate contact temperature: an address question, a price question, a fit question and a proposal request are not the same event. If every one is called an enquiry, the site is optimised for cheap questions rather than fewer, more decision-ready conversations.

04 / THE TRUST TAX

People often leave for reasons no button colour can fix

Professional services, education, fintech and enterprise software share one condition: before a buyer submits details, they must reduce personal and organisational risk.

Who is accountable? What exactly has been done before? What might it cost? Will data be handled safely? Will one enquiry trigger aggressive follow-up? Every missing answer charges the buyer a trust tax.

Nielsen Norman Group’s B2B research shows that business sites remain subject to usability and trust principles. Its contact-page guidance recommends limiting forms to the 3–5 fields genuinely needed to respond. Asking for more does not look more professional; it transfers risk to a stranger.

Hong Kong buyers also use a site as a document they can forward internally. Procurement sends the service page to the owner, finance asks about retention, and IT asks whether systems can connect. If the page only makes sense to one person, the second decision-maker pauses the conversation. Conversion is not a button press; it is whether one page can survive inside a company group chat.

That is why proof should not be a logo row alone. State who did the work, how it was done, where it is not suitable and what the client had to provide. Clear limits reduce risk more effectively than another adjective. The sharper the boundary, the easier it is for a suitable buyer to trust the supplier.

05 / THE 48-HOUR TEST

Pause acquisition and run a no-media test

Find five people who resemble your ideal buyer but do not know the company. Give them 90 seconds on the site, then ask:

1. Who exactly is this company for?A broad answer means the positioning did not land.
2. What costly problem does it solve?Remembering a service label is not the same as understanding value.
3. What evidence makes you believe it?A logo wall is decoration; process, context and limits are evidence.
4. What happens after contact?Uncertainty about response and required information delays action.

If five people give five different answers, the site lacks consensus, not visitors. More promotion will amplify the ambiguity faster.

Add a Hong Kong context test: ask a colleague to enter from a Google result on a phone and complete contact as if travelling on the MTR or eating lunch. Do not explain the site. Note the steps to send a message, the fields that require thought and what happens after the tap. A smooth desktop path does not prove a smooth WhatsApp journey.

Classify findings as “unclear,” “unbelievable,” “inconvenient” or “unattended.” Pick one smallest fix in each class and retest after seven days. Change ten things at once and you will only know that the page changed, not which change made the decision easier.

06 / THE LAST NUMBER

The most useful metric may not be conversion rate

Record where every qualified enquiry began, what it viewed, how it made contact and whether it became revenue. Then add the number owners often miss: time from first contact to a real company response.

A website can build trust in two minutes and an internal process can destroy it with a two-day delay. When costs rise and cash flow is tight, the cheapest growth may be to stop wasting people already at the door.

After two weeks you should be able to answer four questions more useful than traffic: which pages attract qualified visitors, which evidence keeps them moving, which contact path has the least friction, and how long the company takes to take over. Connected, these answers form a website funnel that can be managed.