Call quality scoring means grading every phone lead from Google Ads against a short rubric: did the caller state a real need, are they inside your service area and offer, can they decide, and did the call end in a booking or quote. A four-grade scale (A to D) plus a spam tag is enough. Review a sample each week, link grades to campaigns and keywords, and cut spending on terms that only produce D calls. Duration alone is a weak proxy, so add human or CRM tags.
Why a call count is not a quality measure
Open the call report of almost any service account and you will find the same pattern: a few calls that turn into jobs, a larger group of questions that never will, and a layer of noise such as wrong numbers, job seekers, suppliers and sales agents. Google Ads counts all of them unless you tell it otherwise. If your bidding strategy learns from that mix, it learns to find more of everything.
Scoring fixes this at the point where the data is cheapest to improve: the phone call itself. It gives you a common language between the person who answers the phone and the person who manages the ads, and it replaces arguments like "the calls are bad" with a count of how many were A, B, C or D. If you have not yet set up the basics, start with our guide to conversion tracking for calls and WhatsApp and come back here.
What Google gives you and what it does not
According to Google's help page on measuring calls from ads, calls can be tracked from call and location assets, from calls on your website, from calls on your Business Profile that are linked to an ad interaction, and from conversions imported by upload. You can count a call as a conversion when it lasts longer than a minimum duration that you set. Call reporting must be switched on, and a Google forwarding number or verified number has to be available.
The same page describes AI analysis of call recordings to judge lead quality, but says that recording applies only to calls where both parties are in the United States or Canada. For a business in Dubai, Abu Dhabi or Riyadh that means you should not count on automatic call analysis at the time of writing. Check the current help page before you design your process, because features and country lists change.
That leaves you with three signals you control: call duration, the details Google already logs about each call, and what your own team records after the call. A rubric combines all three.
The four-grade rubric
Keep it short enough that a busy coordinator will actually use it. The table defines each grade by what happened on the call, not by how the caller sounded.
| Grade | What it means | Typical signs | Action |
|---|---|---|---|
| A | Qualified and ready | States a service, an area you cover and a timing; accepts the next step (visit, quote, booking) | Count as a full conversion; protect the keyword that produced it |
| B | Qualified but not ready | Real need and right area, but needs a quote, a comparison or approval | Count as a lead; follow up on WhatsApp |
| C | Unclear or partial fit | Asks for something you do only in part, or the wrong district, or price shopping only | Review the ad and the keyword for mismatched promises |
| D | Not a lead | Wrong service, job seeker, supplier, existing customer support, misdial | Look for negative keywords or a page change |
| S | Spam or sales | Cold sales calls, robocalls, abusive callers | Block if possible; exclude from reporting |
Add one more column to your sheet for the outcome: booked, quote sent, won, lost, no answer. Grade is a judgement about quality at the first call; outcome is what happened later. Comparing the two over time shows whether your grading is realistic.
What goes into a score: five quick questions
- Need: did the caller name a service you actually sell?
- Place: are they inside the area you serve and want to serve?
- Timing: did they give or accept a date or a deadline?
- Authority: are they the person who decides, or a messenger?
- Next step: did the call end with a booking, a visit, a quote request or a promise to send details?
Four or five yes answers is an A; three is usually a B; one or two is a C; none is a D. If you prefer numbers, give one point per question and keep the same cut-offs. The lead qualification questions guide offers wording for the person on the phone so the answers appear naturally during the call.
Add a handling score: separate lead quality from answer quality
Many accounts blame the ads for outcomes that start with the phone. Give each call a second, very short rating for how it was handled: answered within a few rings, caller greeted by business name, key details collected, clear next step given. When a call is graded A for lead quality but poor for handling, the fix is training or staffing, not a new keyword. When the handling is good and the leads are still D, the fix is in the campaign.
Missed calls deserve their own line. Count calls outside working hours and calls that rang out, then compare them with your ad schedule. If a large share of clicks happen when nobody answers, read our guide to ad scheduling and budget pacing by hour.
A weekly routine that takes under an hour
- Export the call details for the past week: time, duration, campaign, ad group and status, as far as your setup reports them.
- Sample calls: grade every call if volume is small, or a fixed sample of 20 to 30 if it is large. Choose the sample randomly, not only the long calls.
- Record grades in the same sheet you use for leads. The lead tracking spreadsheet guide shows a layout that works with WhatsApp enquiries too.
- Summarise by campaign: share of A and B, share of D and S, average duration of each grade.
- Decide one change: a negative keyword, a different ad promise, a page edit, a schedule change or a training point.
If you record calls for review, tell callers that calls may be recorded and check the privacy rules that apply to your business in each country. This is general guidance, not legal advice.
A hypothetical example: reading a week of graded calls
The numbers below are invented for illustration and are not a benchmark or a client result. Imagine a home service company that received 60 calls from ads in one week across three campaigns.
| Campaign | Calls | A | B | C | D or S | Share of A and B |
|---|---|---|---|---|---|---|
| Emergency repair | 24 | 9 | 5 | 4 | 6 | 58% |
| Annual maintenance contracts | 20 | 3 | 6 | 7 | 4 | 45% |
| Brand and general terms | 16 | 1 | 2 | 4 | 9 | 19% |
The overall picture looks acceptable at 60 calls. The campaign view tells a different story: the brand and general terms campaign sends mostly D calls, so its search terms report is the first place to look, and our negative keyword guide is the next. The maintenance campaign has many C calls, which suggests the ad promises something the call handler cannot confirm. The emergency campaign deserves protection. Shifting a small part of the budget from the third campaign to the first, then re-grading the next two weeks, is a controlled test, not a leap of faith.
Feeding quality back to Google Ads
Once you have a few weeks of graded calls, you can pass that information into the account in more than one way. Use the simplest that fits your volume.
- Duration threshold: set the call conversion minimum length near the point where grade A and B calls usually begin, based on your data rather than a generic number.
- Separate conversion actions: create one for all calls (secondary) and one for qualified calls (primary). Bidding then learns from the stronger signal while you still see the whole picture.
- Imported conversions: where your setup and volume allow, upload graded or booked calls as offline conversions so the account knows which clicks led to real jobs.
- Exclusions: use D and S patterns to build negative keywords, location exclusions and schedule changes.
Do not switch bidding to the strict signal until you have enough of it. Too few qualified conversions give automated bidding little to learn from; our overview of bidding strategies for service businesses explains when to move. Because Google is retiring call-only ads, check the call-only ads guide if any of your phone traffic still depends on them.
How we work at ZED
Our tracking and conversion service links your website and ads to calls, WhatsApp and forms so you can see what your campaigns bring. Our Google Ads service includes keyword research, negative keywords and call and WhatsApp tracking. A call scorecard sits on top of that setup: we can help you design the rubric and read the results with you. We do not publish promises about outcomes, and every figure in this article is a hypothetical example.
Frequently asked questions
What is call quality scoring in Google Ads?
It is a manual or semi-automatic way to grade phone leads from your ads, for example from A to D, based on whether the caller had a real need, was in your area, could decide and took a next step. It complements Google's own call duration setting.
Is call duration enough to judge a lead?
No. Duration helps remove very short calls, but a long call can be an argument about an invoice and a short one can be a perfect booking. Use duration as a filter and a scorecard as the judge.
Can Google analyse my call recordings in the UAE or Saudi Arabia?
Google's help page states that recording analysis applies only to calls between parties in the United States or Canada, so do not rely on it for Gulf calls. Check the help page for current country lists.
How many calls should I grade each week?
If you receive fewer than 30 calls, grade all of them. Above that, grade a random sample of 20 to 30, and keep the same sampling method every week so trends are comparable.
Who should grade the calls?
Ideally someone who understands the service and was not the person who took the call, such as the manager. If the call handler grades their own calls, add a second reviewer for a small share to keep the scores honest.
How do I include WhatsApp enquiries?
Use the same rubric on the first conversation: need, place, timing, authority and next step. That puts calls and WhatsApp in one comparable scale.
Does scoring work for all sectors?
Yes, the rubric is generic. Change the questions to match your offer, for example property type for a renovation firm or lesson type for a driving school. For regulated sectors, keep recordings and notes within the rules that apply to you.
How can ZED help with call scoring?
Message us on WhatsApp or use the contact page with a description of your tracking and phone process. We will explain what we would review first.
Turn your call log into a decision tool
Send us how your calls are tracked and answered today. We will suggest a scorecard layout and what to check first, with no promise of results.
Share my call setup on WhatsApp Call 01097079170