Measuring WhatsApp leads from Hong Kong British campaigns

To measure WhatsApp enquiries from a campaign designed for Hong Kong British audiences, connect each placement to a source code, then track whether conversations become qualified enquiries and completed outcomes. Report campaign source separately from customer identity: a Cantonese message or community-media referral does not prove someone is Hong Kong British.

The question: did the campaign generate useful enquiries?

A prospective customer sees your advert, opens WhatsApp and asks a question. Your advertising dashboard records a click, your customer-service team sees a conversation, and your CRM may record a booking days later. Without a shared process, those events remain disconnected.

This article addresses that specific measurement gap. It applies when audience research supports WhatsApp as an enquiry route—not because every Hong Kong British customer supposedly prefers it.

People differ by generation, faith, region, migration history and individual preference. Some want Cantonese support; others prefer English, written Traditional Chinese or another language. Keep alternative contact routes available, and measure them separately.

Your primary metric should be cost per qualified enquiry, not cost per WhatsApp click. Add completed bookings or purchases where the customer journey allows a reliable connection.

Define a qualified enquiry before buying media

Start with a definition that your marketing and customer-service teams can apply consistently.

For a UK service business, a qualified enquiry might mean a person who:

• Is asking about a service the business actually provides.
• Falls within the service area or delivery coverage.
• Has expressed a relevant need, rather than sending only a greeting.
• Has provided enough information to take the agreed next step.

The next step could be a consultation, quotation or appointment. Do not require every customer to disclose a budget or telephone number simply to improve your reporting.

Use a short status list: new conversation, qualified, not qualified, booked, completed. Record broad reasons for non-qualification, such as “outside service area” or “service unavailable”.

Agree whether your unit is a person, a conversation or an enquiry. For most acquisition reporting, count a distinct enquiry once within a defined reporting period. Several messages about the same booking should not become several leads.

Give each placement a traceable route

Build a source map before launch. Each publisher, creator, paid placement or printed QR code should have a distinct campaign identifier.

For example, an illustrative code such as HK26-P02-TC could represent campaign, placement and creative language. It describes the marketing activity—not the person responding.

For a standard WhatsApp click-to-chat link, include that code in an editable pre-filled message:

Hello, I’d like to ask about your service. Reference: HK26-P02-TC

Explain that the reference helps the team identify the offer or advert. Avoid putting personal information in links or codes.

Opening the chat does not send the message. The customer must send it and can remove or change the reference. A link click therefore cannot be treated as a received enquiry.

If an advert first opens your website, use consistent UTM parameters to measure that visit and the onward WhatsApp click where your analytics setup permits. Do not assume those parameters automatically arrive in the WhatsApp conversation or your CRM.

Where a business platform provides advertising referral metadata, test what is actually available in your account and integration. Do not build a reporting promise around an unverified feature.

Test the entire handover

Before spending, follow every route on a real phone:

1. Open the advert link or scan the QR code.
2. Check the correct business account and message appear.
3. Send a test enquiry.
4. Confirm the source reaches the reporting record.
5. Move that enquiry through qualification and booking.

Also test a message with the reference deleted. Your team needs an “unknown source” option, not a requirement to guess.

Record source without inventing ethnicity

A community publisher can distribute your advert beyond its expected readership. A customer may forward the link to a friend. An English-language response may come from someone who identifies as Hong Kong British; a Cantonese response may not.

Keep three concepts separate:

• Campaign source: the placement or link associated with the enquiry.
• Service language: the language the customer requests for communication.
• Identity: how someone describes themselves, if collecting this is genuinely necessary.

For routine campaign attribution, you usually need the first and may need the second. You generally do not need the third.

Do not label people by inferred ethnicity using surnames, profile pictures, telephone prefixes or language. UK data-protection law gives special protection to personal data revealing racial or ethnic origin. Intentionally inferring that information can create obligations that ordinary source tracking does not.

The ICO’s guidance on special category data is a useful starting point. Seek appropriate advice before adding identity fields or audience classifications to your CRM.

Build a report that exposes the gaps

Use one reporting view with clearly defined stages:

• Spend: attributable media spend, with other costs stated separately.
• WhatsApp link clicks: recorded attempts to open the contact route, not necessarily unique people.
• New enquiries: distinct relevant enquiry records received.
• Qualified enquiries: records meeting the agreed definition.
• Bookings: confirmed next steps, not merely proposed appointments.
• Completed outcomes: attended appointments, fulfilled orders or another agreed result.

Calculate cost per qualified enquiry = attributable spend ÷ qualified enquiries. If you include translation, production or handling costs, label the result accordingly and use the same basis across comparisons.

Also report source-code coverage: the proportion of enquiries with a recorded campaign source. A strong-looking result is harder to trust when many conversations remain unattributed.

Allow time for outcomes to mature. Compare enquiries received during equivalent periods and give each cohort the same follow-up window. Yesterday’s enquiries should not be judged against last month’s completed bookings.

Separate creative performance from service performance

A campaign can generate relevant interest and still lose customers after the first message.

Record response time, whether requested language support was available, and whether the advertised service or appointment slots were actually accessible. Otherwise, a slow handover can be misreported as weak audience demand.

When comparing English and Traditional Chinese creative, keep the offer, destination, qualification rules and response standards as consistent as possible. Traditional Chinese is a writing system, not proof that a respondent speaks Cantonese or holds a particular identity.

Where feasible, randomise creative within a comparable audience. If one version runs through a community publisher and another through broad paid social, describe the comparison as two campaign packages, not a clean language test.

Avoid declaring a winner from a handful of bookings. Report underlying counts alongside rates, and treat small differences as directional until there is enough evidence to support a decision.

Attribution is not proof of additional demand

A source code shows an observed route into the business. It does not establish that the campaign caused an enquiry that would otherwise never have happened.

Existing customers may respond to a new advert. Someone already considering your service may use the campaign link because it is convenient.

If the decision requires evidence of incremental demand, consider a suitably designed holdout or phased rollout. Comparable areas or periods can help, but differences in demand, availability and seasonality must be considered. Without a credible comparison, call the result attributed enquiries, not incremental growth.

Practical pre-launch checklist

• Define one primary outcome and a consistent qualification rule.
• Assign distinct source codes to placements and creative versions.
• Test links, QR codes, messages and CRM handovers.
• Provide an unknown-source category and a deduplication rule.
• Agree language support, response standards and follow-up windows.
• Check privacy information, lawful basis, access controls and retention.
• Review consent requirements for your actual analytics and messaging setup.
• Keep enquiry handling separate from permission for future promotions.

A service enquiry is not blanket permission to send promotional WhatsApp messages. Check the ICO’s direct marketing guidance before adding contacts to marketing lists. Keep reporting extracts lean; full chat transcripts rarely belong in a campaign dashboard.

Review cultural risk before launch

Check the advert, pre-filled message, automated reply and human response scripts together. Look for inappropriate translations, identity assumptions, misleading language-support promises and imagery that presents Hong Kong British life as a single experience.

Kulture adapts ad copy, images and video for UK communities and runs cultural risk reviews. The final check should establish both that the creative respects the intended audience and that the business can deliver the experience it advertises.