Most of the local clients I work with (clinics, repair services, small shops) get almost all their leads from a WhatsApp button or a tap to call link, not from forms. In GA4 I usually set a click event on wa.me and tel: links and mark it as a key event, then import it to Google Ads. The problem is a click is not a real lead, people tap and never send the message. Some clients see 40 clicks and maybe 10 actual chats. How are you handling this? Do you just optimize for clicks and accept the noise, use call tracking numbers, or something else like asking the client to log real leads and uploading them as offline conversions? Curious what actually worked for small budgets.
For WhatsApp a prefilled message helps a lot. Put a short code in the wa.me text, different for ads, the map listing and the site, and then you can count actual chats per source on the client's phone instead of guessing from clicks. It's manual, but even a weekly count tells you more than 40 taps.
Clicks are a weak signal, so I'd add one cheap qualifier on top of them. A short thank-you or auto-reply message, or asking callers a single question like 'how did you find us?', lets the client tag each enquiry as real, wrong fit, or spam. Once a month, total the tagged leads against the click count to get a click-to-lead ratio per channel, and use that ratio to judge the ads rather than raw clicks. If a client is on a tiny budget, that simple tally is usually enough before paying for call tracking software.
Thanks both, the prefilled code idea is a good one. I would keep the codes short and boring, like AD1 or MAP1, because people sometimes edit the text before sending and a long code gets deleted. A monthly tally of tagged leads against clicks also gives you a ratio to set a target cost per real chat instead of per click.
Good thread — this is the exact measurement headache with local lead gen. A couple of things that work on small budgets: Treat the click as a micro-conversion, not the conversion. Keep the wa.me and tel: click events firing into GA4, but mark them as secondary. Optimising spend against clicks alone will just buy you more accidental taps. For WhatsApp, the prefilled code idea is solid — keep the codes short and source-specific (site, GBP, each ad group). One extra step that costs nothing: once a month, ask the client to count real chats per code from their phone and compare with the click events. That click-to-chat ratio per source is the number that tells you where to shift budget, and it usually varies a lot by source. For phone calls on a small budget, you don't need fancy software. Get one cheap tracking number per source (site vs GBP vs ads) and forward it to the client's real number. A single number per channel is enough to split the lead count, and the call log becomes your offline lead list. On the website you can swap the number per source with a tiny bit of script; on GBP just put the GBP-specific number. Then close the loop: once a week, take the real leads (chats that got a reply, calls over ~30 seconds) and import them as offline conversions into Google Ads. That is the only way the bidding learns to buy leads instead of taps. Enhanced conversions for leads accepts a simple spreadsheet upload, so even without a CRM you can do it manually in a few minutes. And a practical one: train the client to answer fast. A WhatsApp click that gets a reply in five minutes converts; one that waits three hours doesn't. Speed of response is honestly half the game with click-to-chat leads.
Agreed on keeping clicks as secondary. Importing only the tagged real leads back as offline conversions has worked better for me than optimising on raw clicks, even with low volume. Thanks for the detailed write up.