Why it matters: Know whether to fix the reporting, the way calls are handled, or the quality of the enquiries.
What does the ad report count?
Google’s website phone-number click tracking records a tap on the number, not an actual call. A person can open their dialler and decide not to ring. Other call-tracking methods count different events, so this does not explain every Google call report.
Ask which method produced the number. Then compare the same dates and phone numbers, and check any call-duration rule. A difference between reports is a reason to investigate, not proof that calls or revenue went missing.
A simple example
The same spend can tell two very different stories.
Imagine A$5,000 in advertising spend, 100 clicks on the phone number and 50 recorded calls. That is A$50 per click, or A$100 per recorded call.
That does not prove half your calls were lost. It shows why you need to know what you bought before deciding whether it worked. The next step is to see which recorded calls were suitable and became paying work.
- Per phone-number click
- $50
- Per recorded call
- $100
Illustrative example. Clicks and calls measure different things. Amounts are AUD, excluding GST.
What happened to the calls that reached you?
Once you have the actual call records, check what happened. Were calls answered? Did unanswered callers get an appropriate callback? Were they asking for work you offer?
The fix depends on the answer. A reporting mismatch needs clearer measurement. Genuine unanswered enquiries need a decision about how they are handled. Calls for the wrong service give your marketing team something else to investigate.
- Ask what the reported call total counts.
- Match the dates and numbers before comparing totals.
- Follow genuine enquiries through to their outcome.
How Gibson checks the difference.
Where suitable, we put a tracking number on the website and connect its call records with the source that brought the visitor. We check what your ad report is measuring alongside what reached the phone system.
Where your systems support it, we link those enquiries to the customer record, quote or job. A caller who also sends a form should stay one opportunity, with both contacts visible. Missing links are flagged rather than guessed.
You get an explanation of the gap and an agreed action plan. Your ads team gets useful information about the calls that became suitable enquiries and work. They keep control of the campaigns.
What you can do next
Start with one useful question.
Ask: “Can you show me which calls reached us and what happened next?” Gibson can help connect that picture, identify the gaps and agree what to fix with your team. Your marketing team keeps control of the ads.
See the calculations, sources and checks
The evidence
What the published sources say.
Google states that phone-number click tracking measures clicks, not actual phone calls.
A click count and a phone-system call count can legitimately differ. The difference needs investigation before it becomes a lost-call claim.
Google Ads: phone-number clicks ↗Google distinguishes calls from call conversions. A conversion can depend on a chosen duration, and its reporting date can differ from the call date.
Compare matching numbers, dates and definitions. A long call still does not establish that a job was sold.
Google Ads: call reporting ↗The measures that make this useful.
- Cost per recorded call
- Relevant ad spend ÷ recorded calls linked to that spend
Shows the advertising cost of a call rather than the cost of a tap.
- Answer rate
- Answered calls included in the review ÷ recorded calls included in the review
Compare busy hours and out-of-hours separately before changing staffing.
- Suitable-enquiry rate
- Suitable new enquiries ÷ recorded calls reviewed
Separates useful demand from repeat contacts, wrong numbers and work you do not offer.
- Cost per paying job
- Relevant ad spend ÷ paid jobs confirmed in the records linked to those enquiries
Connects advertising cost to work that has actually paid. Keep unknown outcomes visible.
Worked value example
The same spend can tell two very different stories.
Illustrative calculation. All inputs assumed. AUD; amounts exclude GST. Not a customer result, forecast or Gibson price.
Assume one matched reporting period: A$5,000 in ad spend, 100 reported activities, 50 recorded calls, 40 answered calls, 20 suitable enquiries and five paid jobs. These figures are invented. The 50-record difference is not assumed to be lost calls.
- Cost per phone-number click
- $50
- Cost per recorded call
- $100
- Recorded-call answer rate
- 80%
- Cost per suitable enquiry
- $250
- Cost per verified paid job
- $1,000
- Modelled additional paid jobs
- 1
- Modelled additional paid revenue
- $2,400
- Modelled gross contribution before extra costs
- $960
10 eligible missed enquiries × 50% reached × 80% suitable × 25% becoming additional paid jobs × A$2,400 = A$2,400 modelled additional paid revenue.
The recorded-call cost is A$100, not A$50. That is a correction to the denominator, not a saving. Separately, the callback assumptions produce one additional paid job and A$960 gross contribution at a 40% margin before follow-up costs and overheads.
Show every assumed input
- Relevant ad spend
- $5,000
- Phone-number clicks in this example
- 100
- Recorded calls
- 50
- Answered calls
- 40
- Suitable enquiries
- 20
- Verified paid jobs
- 5
- Genuine missed enquiries eligible for contact
- 10
- Assumed contact rate
- 50%
- Assumed suitability among contacts
- 80%
- Assumed paid-job rate among suitable contacts
- 25%
- Assumed paid value per extra job
- $2,400
- Assumed gross margin
- 40%
What would change the result? The callback rates and additional job are assumptions. If the work would have arrived anyway, it is not incremental recovery. Zero extra paid jobs means zero extra revenue. Confirm eligibility and contact preferences before agreeing any contact.
From our Blind Marketing build
One customer can leave several records.
When we built the Blind Marketing walkthrough, we separated contact records from opportunities. A person can try calling, submit a form and call again about the same job. Counting those contacts as three potential jobs would inflate the opportunity total. Counting only the final answered call would hide the earlier attempts. We kept both views: contacts to explain what happened, and linked opportunities to follow the work.
What to take from it: Before calculating a cost, write down what you are counting: taps, calls, people or jobs. Then use that same definition throughout the comparison.
This is a design decision in our mock-data walkthrough, not a measured customer result.
When do unanswered calls happen?
Look at calls by day and hour. How many arrived, how many were answered and how long did people wait for a callback? This helps you see whether calls are being missed at particular times.
- Compare similar weeks. A percentage based on only a few calls can be misleading.
- Keep time to callback separate from time to paid work. A quick callback is not the same as a sale.
- After a change, check what happened to those enquiries and what the extra work cost. A later sale does not by itself prove the change caused it.