Three numbers people confuse
These are not interchangeable, and reporting one as another is how agencies produce flattering nonsense.
A button click
Someone tapped the call button or pressed submit. It proves intent and nothing more. A tap does not mean the call connected, and a submit press does not mean the message was delivered.
A received inquiry
A message actually landed in your inbox, or a call actually rang through. This is verifiable outside analytics — the email exists or it does not.
A qualified lead
A real person, in your service area, wanting something you actually sell, who you would be willing to quote. Spam, recruiters, and out-of-area requests are inquiries, not leads. Only you can classify this, and it usually takes thirty seconds per message.
A minimum viable setup
- Analytics installed once, initialized before any page-view event, with a single measurement ID
- One page-view method chosen deliberately on a single-page site — either correctly configured automatic tracking or manual route tracking — never both, so navigation is not double-counted
- An event when a phone link is tapped
- An event when the primary call-to-action is clicked
- Separate events for a successful and a failed form submission
- No names, emails, phone numbers, message text, or other personal data sent to analytics — ever
- A simple record — a spreadsheet is fine — where each received inquiry is marked qualified or not
Testing without polluting your data
Your own visits, your web person's visits, and your test submissions all look like demand if you do not separate them.
- Label test submissions unmistakably in the message field so they can be excluded later
- Keep a note of the date and time of every test so those events can be filtered
- Test the failure path too — disconnect and submit — so you know the error state is visible to visitors
- Prefer a staging or preview environment for repeated testing when one exists
- Do not judge a change on the first week; small local sites have noisy weekly numbers
Reading the numbers honestly
Hypothetical example, for illustration only. In one month a site records 500 sessions (sessions, not unique people), 40 call-button taps, 15 form submissions, 14 delivered emails, and 9 of those delivered emails are messages you would happily quote.
The useful reading is not '500 visitors'. The 55 contact actions (40 taps plus 15 submissions) are actions, not 55 different people — one person can tap and also submit. One submission never arrived and needs investigating. And roughly 64% (9 of 14) of the delivered form inquiries were worth quoting — that ratio describes those emails only, not calls, which are not measured here. Each observation points at a different fix.
These figures are invented to demonstrate the reasoning. Your own ratios will differ by trade, season, and market.
Common measurement traps
- Counting form submit clicks as leads when delivery was never verified
- Two analytics snippets on one page, doubling every number
- Judging performance by impressions or keyword positions rather than inquiries
- Comparing a busy season with a quiet one and calling it a trend
- Sending customer details into analytics, which is both a privacy problem and unnecessary
- Never checking the spam folder, then concluding the form is broken
Questions about this page
- Do I need call tracking numbers?
- Not to start, and they carry a consistency cost: multiple phone numbers published for one business can contradict your other listings. Begin with tap events and a simple call log.
- How long before the data means anything?
- For most local sites, think in months rather than weeks. Low weekly volumes swing wildly for reasons that have nothing to do with your website.
- What if the numbers say the site is fine but no one calls?
- Then check delivery end to end first: submit a real test, confirm the email arrives, and check spam. Silent delivery failure is easy to miss and worth ruling out before assuming a traffic problem. The conversion checklist covers the rest of the path.
Want help setting this up?
Measurement is part of every engagement — see ongoing optimization or get your free website plan.
Get My Website PlanPrime Local Media does not guarantee rankings, traffic, leads, revenue, or inclusion in any AI answer. Timelines and outcomes depend on your market, your website, and how quickly information and approvals come back.