Which ad brought the phone call? A practical lead-source tracking setup

By Noe Guerrero III

An ad gets a click. Someone visits your website, taps the phone number, and asks for a quote. A week later, that job is on the calendar. Can your team connect those steps?

To track which ads produce leads, you need three things working together: campaign information on the visit, a reliable record of the inquiry, and a way to update what happened afterward. A dashboard full of clicks only covers the beginning.

For a service business in McAllen or elsewhere in the Rio Grande Valley, phone calls may be part of the sales process alongside online forms. I’d build the measurement plan around both, then check whether those inquiries become work the business actually wants.

Decide what counts as a lead before installing anything

Write down the stages your team will report. Keep the definitions simple enough that two people would classify the same inquiry the same way.

• Contact attempt: Someone taps a call button or begins a form.

• Inquiry: A submitted request reaches your team, or a caller asks about your services.

• Qualified lead: The inquiry fits your service, service area, and agreed qualification criteria.

• Booked job or sale: The customer completes the specific commitment your business uses to mark a win.

A booking and a payment can be separate stages. Choose names that match your process instead of squeezing everything into “conversions.”

For a hypothetical repair company, an existing customer asking about an invoice would be a real call but not a new sales lead. A request outside the company’s service area would be an inquiry but not a qualified opportunity. Those distinctions make the advertising report more useful.

Assign one person to maintain these definitions. Otherwise, “qualified” can quietly change whenever somebody wants the monthly numbers to look better.

Give each campaign a consistent source label

UTM parameters are labels added to a destination URL. Google Analytics can use them to identify the campaign that referred a visit. Use a consistent source, medium, and campaign name; add a content label when you need to distinguish creative versions. Values are case-sensitive, so a shared lowercase convention prevents avoidable splits. Google’s campaign URL guide explains the fields.

For example, a hypothetical paid Instagram campaign promoting estimates could use:

• Source: instagram

• Medium: paid_social

• Campaign: fall_estimates

• Content: crew_video

Keep a small campaign register with those labels, the destination page, and the person responsible. Someone building the next ad should copy the agreed naming pattern instead of inventing another abbreviation.

Google Ads has its own important setup detail: Google recommends auto-tagging for the most detailed Ads data in Analytics. Don’t switch it off just to make everything use manual UTMs. Have the person managing tracking check the account connection and destination URLs. Google’s auto-tagging documentation covers that distinction.

A campaign label identifies a source you can observe. It doesn’t prove that one ad deserves all the credit for a customer’s decision.

Carry the source into the inquiry record

Ask your developer to show what happens between the landing page and the saved form submission. Seeing a campaign name in Analytics doesn’t establish that your inbox or customer relationship management system received it.

For a form-based inquiry, I’d want the internal record to contain the submission time, requested service, landing page, available campaign labels, and a stable inquiry ID. Customer contact details belong in the appropriate protected business system. Staff should be able to find the record and update its status without reading a long tracking URL.

Choose a source rule explicitly. You might store the campaign present when the inquiry was submitted, while retaining an earlier known source in a separate field. Label those fields clearly. Don’t silently overwrite one with the other and later report them as though they meant the same thing.

If no source is available, save “unknown.” It’s better than automatically crediting your current advertising campaign.

Keep customer names, email addresses, phone numbers, and form-message text out of campaign labels and Analytics events. Google’s policy prohibits sending personally identifiable information to Analytics, including through URLs and campaign fields. Its PII guidance is a useful implementation check.

Treat a phone-button click as its own event

A visitor can tap your phone number, open the dialer, and decide not to call. Label that event “phone click,” so the report says what was measured.

Google Ads distinguishes mobile website number clicks from tracked calls. Its number-click tracking measures the click, not the call itself. Website call conversion tracking can use a Google forwarding number and a minimum call duration to count qualifying calls after an ad click. Availability and setup requirements apply. Google’s phone conversion guide describes the options.

Even a duration-based call conversion still needs business context. A long call could involve an existing order, a wrong service, or someone selling to you. Call length alone doesn’t tell your team whether the caller became a qualified lead.

For any call-tracking setup, ask the provider or developer to demonstrate the actual route: the number a visitor sees, the destination that rings, the report entry, and the way staff attach an outcome. Test both tapping the number and manually dialing it.

Also test after-hours handling. A campaign can generate interest while your business is closed; your process still needs to distinguish missed calls, callbacks, and completed conversations. Counting all three as separate new leads would inflate the result.

Close the loop with a short staff routine

The tracking system needs an answer from the people handling the work. Give each inquiry an owner and a few useful statuses, such as new, contacted, qualified, booked, and closed without a sale.

Keep the loss reasons practical: outside service area, service not offered, duplicate, spam, price, or no response. Avoid a giant dropdown that nobody finishes. Set a regular review time and decide who resolves records that remain unclassified.

If someone submits a form and then calls about the same project, connect both interactions to the same opportunity. Keep the interaction history, but report one distinct opportunity in the business totals. A repeat customer asking for a separate new job may need a new opportunity; document that rule too.

An optional “How did you hear about us?” answer can add context. Store it separately from the observed campaign source. A customer remembering a recommendation and a system recording a later ad click can both be useful facts.

Start with a process staff can maintain, then automate repetitive steps. The person answering the phone should know exactly where to record the outcome and who takes the next step. A technically accurate report won’t help much if the underlying requests sit unanswered.

Compare campaigns using the same stages

Here’s an invented example, not a client result or a performance benchmark. Assume both campaigns ran over the same period, their inquiries have had enough time for follow-up, and duplicate opportunities have been removed.

Campaign A looks cheaper when the report stops at inquiries. Campaign B produced more qualified leads and booked jobs in this example. That’s the reason to keep the stages separate.

The calculation is straightforward: divide campaign spend by the number of qualified leads attributed to that campaign. If there are none, show zero qualified leads and no calculated cost per qualified lead; don’t display a misleading zero-dollar cost.

This table still doesn’t establish profit. Job values, delivery costs, cancellations, and the time spent answering inquiries matter. Nor would I redirect a whole budget based on a handful of bookings. First check the classifications, unfinished follow-up, and whether both campaigns had a comparable chance to produce outcomes.

Keep business totals separate from platform conversion totals

Audit the actions feeding your advertising reports. One successful form might be measured by a website tag and an Analytics import. A phone click and the resulting conversation might also appear as separate actions. Adding every row together won’t necessarily give you a count of distinct prospects.

In Google Ads, primary actions can be used for bidding when their standard goal is used by the campaign. Secondary actions usually serve observation, with an exception for actions in custom goals. Review both the action settings and the campaign’s selected goals. Google’s primary and secondary conversion documentation explains the distinction.

My recommendation is to write down what each action means and why it’s included. Keep preliminary signals available for diagnosis without presenting them as completed sales. Don’t make a blanket change to bidding goals until you understand the existing configuration.

Use the CRM’s distinct opportunities for business totals, and retain the platform reports to understand what each platform measured. Investigate differences instead of forcing the numbers to agree.

Test one complete customer journey before trusting the report

Pick one campaign destination and run a controlled test with your team. Label test records so they don’t become sales results.

1. Open the final destination with the intended campaign labels and confirm that the right page loads, including through any redirects.

2. Move to the contact form, complete it, and verify that the request was actually received. A button click or a validation error shouldn’t be reported as a successful inquiry.

3. Check the saved record’s source and inquiry ID. Reload the confirmation page and confirm it doesn’t create another lead.

4. Test the phone route separately. Confirm which event records a tap, which records a call, and how the team logs a qualified inquiry.

5. Update a test opportunity through qualification and booking. Check that the reporting view reflects the intended stages once, with understandable timestamps.

For a bilingual South Texas campaign, repeat the journey in each language you advertise. Check the destination, form, confirmation, and staff handoff. Our bilingual form guide covers that review in more detail.

Keep a short list of known gaps, including missing sources, incomplete staff follow-up, and reporting delays. Don’t relabel uncertain inquiries to make a channel look complete.

Start with one campaign, one form, and one phone route. Once you can follow an inquiry through to its outcome, expand the setup. RGV Web Design’s marketing and connected software services can help bring the website, campaign, and follow-up process together around that practical question: which inquiries are becoming useful business?