Dynamic call tracking numbers on a plumbing site: per-visitor swaps or one number?

Say you run a plumbing business with a website, a Google Business Profile and a few Search campaigns, and you want to know which channel actually books jobs. That means some kind of call tracking, and there are two main routes: a pool of numbers that swaps per visitor source, or one number everywhere with tracking handled through the ad platform.

I lean towards the pool for anything that isn't a click-to-call ad. One number can't tell you whether a call came from a Search ad, the organic results or a directory listing. The catch is that the swap only works if the number loads from the page and the ad clicks still register properly, so test it on a real phone before you trust the data.

Some people prefer a single number because the office never has to ask callers how they found them. For a team of two or three techs, though, does the extra tracking really change where the budget goes, or is it mostly a nicer report?

Agreed, the pool only pays off if the calls get counted properly, and for plumbing the thing to watch is call length. A 20-second "do you cover my postcode?" call can register as a conversion unless you set a minimum duration on the call conversion action, and that throws off the channel comparison.

Agreed on the duration threshold. On Microsoft Ads the minimum call length is set per conversion action, so keep it as its own goal rather than one catch-all. A "calls over 60 seconds" goal for channel reporting and a "booked job" goal imported from the CRM will disagree, and that gap is useful. If the 20-second postcode calls sit in the goal bidding optimizes toward, the account learns from them.

Check volume before you split by channel, too. Bing usually brings a much smaller share of calls than Google, so a few techs might only get a handful of Microsoft calls a month, and numbers that small swing a lot from one month to the next.

Theo's volume point decides it for me, and it's really about how many calls the whole business gets, not just the Bing share. Split 40 calls a month across a pool of sources and you're left with a handful per channel, so one bad week changes the story. At that size I'd skip the pool and track booked jobs by source from the CRM. You get fewer numbers, but ones you can act on.

The pool also only sees calls that start on your website. Calls from the Google Business Profile listing go through the GBP number and show up in Business Profile insights, not in your site's tracking. For a local trade those can be the biggest share of calls, so a website pool can make Search look better or worse than it really is unless you account for them separately.

Agreed, and the GBP gap is bigger than it looks. For Search you don't need a website pool at all, because call reporting in Google Ads already counts calls from the ad itself, so you can separate Search from the listing without swapping any numbers. I'd also tag CRM bookings by job type. An emergency callout and a new boiler quote can come from the same channel and be worth very different amounts, so a source that brings in lots of cheap calls can look busy while most of the revenue comes from somewhere else.

On the Microsoft side the CRM link is weaker than people expect. Bing's offline conversion import matches a booking back to a click through the msclkid, and that only gets captured when the booking comes through a web form or a tracked page. A phone booking carries no click ID, so unless the office logs how the caller found you, that job lands in "unknown" and Bing looks worse than it is. For a trade where most bookings are calls, that's a big share to lose.

I'd start with a booking form that passes the msclkid into the CRM. After that, make "how did you find us" a required field on phone intake. It's boring, but it's the only way phone-booked jobs get attributed at all.

I'm not sure "the only way" holds. Calls that come through a Microsoft call extension or a Microsoft tracking number are already tied to the click inside Microsoft Ads, so Bing gets that credit without any intake question. The "how did you find us" field mostly catches people who searched and then dialled your main number directly.

Mira, I don't think the call credit solves this. A Microsoft tracking number tells you a call from a Bing ad happened and lasted long enough to count. It doesn't tell you the call turned into a booked job, and that's what the budget decision actually depends on.

Theo's problem was matching CRM bookings back to the channel, and call credit inside Microsoft Ads doesn't do that. The job still has to be recorded in the CRM, and if it's booked by phone with no click ID attached, it lands in "unknown" no matter what Microsoft reported for the call. Tracking numbers also only cover calls that come through that ad or extension, so someone who clicks, doesn't call, then rings the main line later still falls through.

The intake question is cheap, so keep it alongside the call tracking instead of treating call tracking as a replacement.

Agreed, and nobody's brought up the office side yet. If the intake question is a free-text box, you'll end up with "google" and "Google search" and "saw your van" all in one column. Give the field a short fixed list (Google ad, Bing ad, Google listing, website, referral, repeat customer) so the answers can actually be counted.

Agreed on the fixed list. The one thing I'd add is that it only works if "Bing ad" and "Google ad" are kept separate from "Google listing." A caller who saw your GBP profile and then clicked a Search ad will often just say "Google" and mean either one. Give the office a line like "if they found you on a Google result, was it the map pin or an ad?" so the answer lands in the right bucket. Otherwise your Search number quietly absorbs listing calls, and the channel split looks cleaner than the data really is.