Simulated example · Simulated educational case study

Case study: a home renovation company measures its calls for eight weeks — simulated example

Content updated:

Short answer

In this fictional case, the company answers 76.7% of 180 calls, but only 41.2% between 13:30 and 15:30. Google Ads counts 54 conversions against 24 qualified calls because the action had no useful duration threshold, and listing calls cannot be separated because the listing shares the main number. A qualified Ads call costs €100, an accepted quote €480, and ROAS at a 30% margin is 5.19 against a break-even of 3.33.

Key verification: Compare Google Ads call conversions with your qualified calls for the same period and review the configured minimum call duration.

Sources and limitations

A home renovation company that measures its own calls for eight weeks may find that its problem is not the ads, but lunchtime, a badly defined conversion and an untagged number. This case follows every call through to the accepted quote and calculates cost and return with margin. All data is fictional: it does not come from a CallsIQ customer or from a provider test.

The scenario and the recording rules

The company in the example renovates kitchens and bathrooms in Valencia and its metropolitan area. It has four people: the manager, who visits and quotes; an administrator, who answers the phone; and two tradespeople. Calls arrive through three routes: Google Ads with a call asset, its Google Business Profile listing and the organic website, which displays its own tracking number.

The rules are set before looking at results. A call is qualified if it asks for a renovation within the service area and accepts or suggests a visit. Amounts are accepted quotes, excluding VAT. The funnel only follows calls answered at the first attempt; callbacks are recorded separately. Quotes are observed until four weeks after the period: of 34 sent, 13 were accepted, 12 were rejected and 9 are still pending.

The eight-week journey by source

A renovation company measures 180 calls: table 1
SourceCallsAnsweredQualifiedVisitsAcceptedAccepted amount
Google Ads725424165€41,500
Organic website383115114€30,200
Untagged main line (includes the listing)705321134€28,800
Total180138604013€100,500

The overall answer rate is 138 / 180 = 76.7%. By source: Google Ads 75.0%, website 81.6% and main line 75.7%. Of the answered calls, 43.5% were qualified (60 / 138). Two in three qualified calls led to a visit (40 / 60) and 32.5% of visits to an accepted quote (13 / 40). The average accepted quote is 100,500 / 13 = €7,730.77.

First problem: nobody covers the phone at lunchtime

When the log is crossed with the time of day, the 42 unanswered calls are not evenly spread.

A renovation company measures 180 calls: table 2
Time slotCallsAnsweredUnansweredAnswer rate
13:30-15:3034142041.2%
Rest of opening hours1461242284.9%
Total1801384276.7%

The cause is organisational: the administrator takes lunch at the same time as the manager is out on visits. A rough estimate, not a proof, applies to those 20 calls the overall ratio of accepted quotes per answered call (13 / 138): that gives 1.88 quotes, about €14,565 in value and €4,370 in contribution at a 30% margin. Some of those callers may have rung again and already appear among the answered calls, so the figure is an order of magnitude, not an accounting loss.

Second problem: Google Ads counts conversions that are not customers

Google Ads showed 54 "Calls from ads" conversions in the period, while the log held 24 qualified calls. The action had been created without a useful duration threshold, so it also counted wrong numbers, suppliers and out-of-area enquiries. The Google Ads help on call conversions states that the advertiser defines the minimum duration and that shorter calls are not counted; the pages consulted on 22 October 2026 do not publish a recommended value, so the threshold is a business decision. The page on tracking calls from ads requires a call or location asset and call tracking switched on, and the list of Google forwarding numbers includes Spain.

In the log, 32 of the 54 answered calls lasted 60 seconds or more, and that group contains 23 of the 24 qualified calls. Even 60 seconds does not separate customers from the merely curious: 9 long unqualified calls remain. That is why the threshold is paired with an import of qualified calls, as described in the guide to importing call conversions into Google Ads.

Third problem: the Google Business Profile listing does not appear in the data

The listing shows the main line, the same one painted on the vans, printed on invoices and passed on by word of mouth. That is why there are 70 calls nobody can assign. The performance panel records 61 "calls" over the eight weeks, but Google's help on profile performance defines that metric as clicks on the call button, not completed calls. It cannot be concluded that 61 of the 70 came from the listing.

The fix is to give the listing its own local tracking number. The Business Profile guidelines require the business to control the number directly and recommend a local one; the guide to call tracking on Google Business Profile explains how to do it without losing profile consistency.

Cost per result and margin-based ROAS for Google Ads

Spend was €2,400 excluding VAT, €300 a week. A contribution margin of 30% is assumed after materials, labour and job travel, plus €40 per quote visit (time and travel), which the margin does not include.

A renovation company measures 180 calls: table 3
IndicatorCalculationResult
Cost per call2,400 / 72€33.33
Cost per conversion according to Ads2,400 / 54€44.44
Cost per qualified call2,400 / 24€100.00
Cost per accepted quote2,400 / 5€480.00
Cost per accepted quote including visits(2,400 + 16 × 40) / 5€608.00
ROAS on value41,500 / 2,40017.29
Margin-based ROAS41,500 × 0.30 / 2,4005.19
Break-even ROAS1 / 0.303.33
Contribution after ads and visits12,450 − 2,400 − 640€9,410

The result looks comfortable, but it rests on five jobs: €14,200, €9,800, €7,600, €5,900 and €4,000. Without the largest, margin-based ROAS falls to 27,300 × 0.30 / 2,400 = 3.41, almost at the threshold. The guide to ROAS and ROI for services explains why the remaining contribution must still cover fixed costs.

What changes, at what cost and how it is checked

A renovation company measures 180 calls: table 4
ChangeProblemAssumed costCheck
Stagger lunch breaks and divert to a second phone41.2% answer rate at lunchtime€0; reorganise shiftsAnswer rate in the slot at or above 80% for 4 weeks
Return missed calls within 30 minutesCalls with no answer and no follow-up10 minutes a dayPercentage of missed calls returned in time
60-second threshold and import of qualified calls54 conversions against 24 qualified calls2 hours of set-upGap between Ads conversions and the log
Local tracking number on the listing70 untagged calls€8 a monthListing calls against call-button clicks

The costs are assumptions of the example: ask the provider for the real price of an additional number. For callbacks, follow the method for returning missed calls without duplicating contacts. The Google Ads budget stays unchanged until the measurement is repeated for another eight weeks with the corrected conversion, as explained in the guide to measuring Google Ads calls in Spain.

How to repeat the exercise in your business

  • Log every call with source, time, answered or not, duration and whether it is qualified.
  • Compare Google Ads conversions with your qualified calls for the same period.
  • Check whether your listing shows a number used only by the listing.
  • Calculate cost per accepted quote and ROAS with your real margin, and repeat it without the largest job.
  • Let the decision period for pending quotes run its course before drawing conclusions.

If you change several things at once, a later improvement will not show which one caused it. Record the date of each change. This case is simulated and its amounts must not be quoted as results from a real business.

Sources and limitations

Documentary review: . Content type: Simulated educational case study.

Linked sources describe the terms and capabilities each provider publishes. Examples that use fictional data are labelled in the text.

How to report a correction

How this guide was prepared

Official sources, explained calculations and clearly labelled examples. Read about our methodology and use of AI in writing.