Straight answers for local service businesses
The only reliable way to track where leads come from is to ask every caller and log the answer, then back that up with a separate phone number or tracking link for each channel you pay for. Analytics tools show you clicks and sessions, not booked jobs, so they will always undercount word of mouth and over-credit whatever a customer touched last. Give it 90 days of honest logging and you will see which channels bring real work and which ones just bring noise.
Most contractors I talk to have a gut feeling about where their work comes from. "Mostly referrals." "Google, I think." Then they look at a real log for 90 days and find out half their best jobs came from one source they almost cancelled, and the thing they pay the most for brings tire-kickers.
Here is how to find out for sure, without buying software you will never open.
Because analytics measures screens, not jobs. It can tell you someone came from a Google search and spent 40 seconds on your water heater page. It cannot tell you that person was your customer's sister-in-law, that she already had your truck number saved, or that she called and booked a $4,200 job.
Three specific blind spots show up over and over:
None of that means analytics is useless. It means it is one input, not the answer.
Ask, and write it down. That is it. A shared spreadsheet or a notebook by the phone beats every tool you have not configured.
Log these columns for every single inbound lead:
Thirty seconds per lead. After a month you will have something no dashboard can give you: a list of actual jobs with sources attached.
One rule that matters more than the rest: log the nos too. If a channel sends you 20 calls and you book two, that is a channel problem, not a sales problem. You will never see that if you only track the wins.
Put it at the end, after you have the job details and the address. By then the caller is relaxed and you have already been helpful.
Phrasing that gets real answers:
That second one is the useful one. It splits referral from search right away. If they say online, one follow-up: "Do you remember if you clicked an ad or just the regular listing?" Half will not remember. That is fine. "Online, not sure" is still better data than nothing, and the numbers you do get will be enough to see the pattern.
If you have an office person or an answering service, make this a required field in your intake script. Not optional. Not "if it comes up."
Yes, for anything you pay for. A separate phone number per paid channel means calls sort themselves and nobody has to remember anything.
A practical setup for a small shop:
All of them forward to the same phone. You answer normally. The difference is your call log now tells you which bucket each call came from.
Two warnings. First, keep your main number consistent on your Google Business Profile and major listings. Mismatched numbers across the web muddies your local visibility. Second, do not go overboard with dynamic number swapping on your main site unless someone is managing it. For most trades, two or three fixed tracking numbers covers everything worth knowing.
Make the form tell you where it came from before the customer ever sees it. Two easy moves:
Then do the thing almost nobody does: forward every lead email into one folder. At the end of the month you have a stack you can actually count.
Track two answers instead of one. Where they first heard of you, and what they touched last. They are different questions and both matter.
A landscaper might find that yard signs and referrals create almost all first contact, while Google handles the last step. Cut the signs because "Google gets the credit" and the calls dry up a few months later with no obvious cause. Logging first-touch protects you from that mistake.
Fuzzy attribution is normal. You are not trying to build a perfect model. You are trying to answer one question: if I turned this channel off tomorrow, would my calendar notice?
Sort by booked jobs and dollars, not by leads. Then run three quick numbers per channel:
Then act. Cut anything with a bad cost per booked job and no strategic reason to exist. Feed the one with the best close rate. And if a channel brings plenty of leads but few bookings, look at your intake and your follow-up speed before you blame the channel.
If you only do one thing from this post, start the log. A notebook by the phone, filled in for 90 days, will tell you more about your marketing than any report you have ever been sent.
Only if you are spending real money on ads or paid lead services. Most phone providers can give you extra forwarding numbers cheaply, and for two or three channels that is plenty. Skip the enterprise tools.
Not if you keep your main number on your Google Business Profile and your core business listings. Use tracking numbers for ads and paid lead sources only, and leave your primary number consistent everywhere else.
That is normal and it is still useful. Log it as "unsure" and move on. Enough answers will be specific that the pattern shows up anyway after a couple of months.
Give it 90 days. A single month is too small for most trades, especially if your work is seasonal. Ninety days of consistent logging is enough to make real spending decisions.
Yes, because "all referrals" usually turns out to be partly referrals and partly people searching your name after hearing it. Knowing the split tells you whether your website and Google profile are doing their job in closing those referrals.
We build it, launch it, and handle everything. $499 one time, live in minutes.
See how it works