Skip to main content
NewNMC’s advertising guidelines for doctors are in force from 6 Oct 2026. See what to change, and get a free compliance auditNew NMC advertising rules for doctors: what to change
Analytics

Tracking Phone Calls and WhatsApp Clicks as Conversions: A Clean Event Design

A tap on a call or WhatsApp button is intent, not an enquiry. Here is an event design that separates click, conversation started and conversation qualified, and keeps each counted once.

Tracking Phone Calls and WhatsApp Clicks as Conversions: A Clean Event Design

For most diagnostic labs and clinics, the phone and WhatsApp are where bookings happen. A visitor reads a test page, taps the green icon, and the real conversion takes place in a chat that the website never sees. Many sites count that tap as a conversion and report a very healthy-looking number. A tap is not an enquiry. Some people tap and close the app. Some tap three times. Some start a chat that never reaches a person who can book.

This article is for marketing leads and analysts at diagnostic centres, labs and clinics who track calls and WhatsApp in GA4 and Google Ads. It separates the stages, shows the events, and explains how to count each stage once. The examples use a hypothetical diagnostics chain with home sample collection in several Indian cities.

Three stages, three events

Treat every phone or WhatsApp contact as moving through up to three stages.

StageMeaningWhere it is observedEvent name
ClickThe visitor tapped a call or WhatsApp buttonThe browsercall_click, whatsapp_click
StartedA call connected or an inbound message arrivedCall tracking or WhatsApp systemcall_connected, chat_started
QualifiedA staff member confirmed a genuine booking enquiryCRM or staff actionqualify_lead (with lead_channel)

Only the first stage can be measured by a website tag. The second and third need your telephony or messaging system and, for qualified, a person.

Which stage should Google Ads optimise on? Not the click. If bidding optimises on clicks to the WhatsApp icon, it will find people who tap icons. Use "started" at minimum, and "qualified" where you have enough volume and the import is reliable. Report all three, with their ratios, so you can see where the drop-offs are.

Stage 1: click events

GA4's enhanced measurement has an automatic outbound `click` event for links that lead away from your domain, with parameters including link_url, link_domain and an outbound flag. A wa.me link will trigger it. That is useful as a check but it is not a clean design, because the event name is generic and every outbound link produces it.

Build your own events instead:

  • whatsapp_click with click_location (header, testpage, stickybar, bookingconfirmation), `pagetype and language`.
  • call_click for tel: links with the same parameters.

Do not send the phone number, the pre-filled message text or the page text as parameters. A WhatsApp link can include pre-filled text; if that text can describe a test or condition, keep it out of analytics and out of the event.

Do not mark the automatic click as a key event. If you mark whatsapp_click as a key event, use it as a secondary signal and not as the primary one.

Stage 2: conversation started

This is the first stage with real information, and it happens outside the browser.

Calls. A call tracking number reports whether a call was answered and its duration. Google's website call conversion documentation describes forwarding numbers swapped onto the site by a snippet, with a call length minimum you set so that very short calls are not counted as conversions. Choose a minimum that fits your call pattern; a lab that takes quick bookings needs a lower one than a clinic with longer consultations. Our guide to call tracking for healthcare covers choosing a provider.

WhatsApp. An inbound message to your business number is the equivalent. If you use the WhatsApp Business Platform or a tool built on it, you can record the first inbound message time. Our own tool WaCRS (wacrs.com) helps plan WhatsApp outreach campaigns and manage business conversations, and any comparable system can supply the same record. The requirement is only that someone can export "first inbound message from this chat, at this time".

Joining a click to a conversation

A GA4 event cannot know that a chat started unless something carries an identifier from the browser to the message. A practical method:

  1. On the click, the page generates a short random reference, for example four characters.
  2. The WhatsApp link's pre-filled text includes it: "Hi, I'd like to book a test. Ref K7QX". Do not put personal data in the text.
  3. The browser sends whatsapp_click with that reference as enquiry_ref, and the reference is saved on your side together with the GA4 client_id and session_id.
  4. When the first inbound message arrives, your system finds the reference, marks the chat as started and sends a server-side event to GA4.

The Measurement Protocol exists for this: Google describes it as a way to augment automatic collection with server-to-server and offline interactions, joined to online interactions by client_id. To tie the event to a session's geographic and device data, include the session_id and send the event within 24 hours of the session start. After that, the latest tagging data for the client is used. Validate payloads with the validation server Google provides.

The reference may not always survive. Some people delete the pre-filled text; some open WhatsApp from a saved contact. Expect only a share of chats to match, and report matched and unmatched separately. Do not guess at the unmatched ones.

For calls, most call tracking tools give you a session identifier or click ID when the call is tied to a visit; send the connected-call event with the same method.

Want this applied to your practice? Get a free analytics audit.

Book my free audit

Stage 3: qualified

"Qualified" is a human judgement: did this person want a booking for a service we provide, in an area we serve? Define it in one sentence and put it in the CRM as a status. When a staff member sets it, send qualify_lead, which is one of GA4's recommended events described as a lead that meets the criteria to become a qualified lead. Include the lead_channel (call or whatsapp), the reference and, if you use it, the click_id for Google Ads import. Our CRM tracking article shows one way to structure this.

Do not send the patient's name, number or the test requested.

Deduplication rules

Every one of these three stages can double-count. Set rules.

  • One click event per user per interaction. Debounce the listener so a double tap fires once. Add a short lock, such as a few seconds, before the next whatsapp_click from the same session.
  • Counting method. GA4 lets you choose how a key event is counted. Google's key events article refers to a separate setting for the counting method of key events. For click events, count once per session; for qualified leads, count each event.
  • One started event per chat. Key the server event on the chat or call ID. If the same ID arrives twice, ignore the second.
  • One qualified event per enquiry. A status that flips back and forth in the CRM should send only the first transition to qualified.
  • No double-counting across channels. A visitor who taps WhatsApp, then calls, then books is one person. Decide whether you count one qualified enquiry or two, and use enquiry_ref or a hashed internal ID to merge them. For a diagnostics chain, counting per booking is usually more useful than counting per touch.
  • Do not also count the form. If the same visitor also submits the booking form, you have a generate_lead too. Use one reporting definition across them.

A hypothetical worked example

Take a diagnostics chain running a month of search ads for home sample collection in one city. The numbers below are invented to show how the stages read, not benchmarks.

  • 900 taps on the WhatsApp button (whatsapp_click), after the debounce rule removes repeat taps.
  • 310 chats started that can be matched to a click through the reference code.
  • 190 qualified enquiries, as marked by the booking team.
  • 120 booked collections, from the CRM.

Reading it stage by stage: roughly one in three taps becomes a chat, which is a plausible drop for a button that opens an external app. Of the chats, about three in five are real enquiries. The rest are questions about reports, price comparisons or wrong numbers. Cost per qualified enquiry is calculated from the 190, not from the 900 taps. If the team had optimised on taps, it would have reported an attractive cost per conversion and bought more of the traffic least likely to book.

The same table, split by landing page, would show which test pages produce taps without chats. A page with many taps and few chats may have a button that opens correctly on some phones and fails on others, which is a development fix rather than a media one.

Reading the numbers

Report these four ratios each month by channel and by location:

  • Clicks to started: how many taps turn into an actual contact.
  • Started to qualified: how many contacts are genuine enquiries.
  • Qualified to booked: from the CRM.
  • Cost per qualified enquiry and per booking, using ad spend.

A low clicks-to-started ratio on mobile usually points to the button, not the ad: the WhatsApp link may fail on certain browsers, or the number may be wrong. A low started-to-qualified ratio usually points to targeting or to ad copy that attracts the wrong question. Pair this with the guidance in healthcare website conversion optimisation and landing page design for healthcare ads.

Where to start

  1. Define click, started and qualified in writing.
  2. Replace generic outbound-click reporting with whatsapp_click and call_click.
  3. Add a short reference in the pre-filled message and store it with the session.
  4. Send chat-started and call-connected events through the Measurement Protocol or your call tracking integration.
  5. Define qualified and send qualify_lead from the CRM.
  6. Apply the deduplication rules and test with ten real chats.

If you want help with the build, see our conversion tracking service.

Filed undertrack whatsapp click conversionwhatsapp click tracking GA4call click conversion trackingGA4 phone call trackingwhatsapp conversion event designdeduplicate conversions diagnostics

About the author

Co-Founder & CTO · Gurugram, India

Arush builds the systems the rest of the company runs on — the patient-acquisition pipelines, the AI chatbots, the analytics that tie a single click to a booked appointment. If a campaign reports a number, it's because something his team built is tracking it.

Free guide · 15 pages

The Patient Acquisition Blueprint

A 90-day plan from first search to booked appointment.

The Pioneers Brief — one essay, three benchmarks, one teardown, every Thursday. Subscribe free →

Free analytics audit

Your next chapter starts with a conversation.

Three quick questions. We come prepared with a plan for your specialty and city.

  • A 30-minute call with a healthcare specialist
  • Your top three growth opportunities, in writing
  • Honest advice — even if that means not hiring us
  • No obligation and no hard sell
You’ll speak with a senior strategist
Consultations are hosted by Arush Thapar’s team.
Step 1 of 3Takes 30 seconds

What do you want to improve first?

Details stay privateNo obligationReply within 1 working day

Client portfolio · BTL & activations

Work relevant to tracking phone calls and whatsapp clicks as conversions a clean.

Explore client work →