Why Your Google Ads Leads Are Poor Quality and What to Fix First

When Google Ads produces enquiries your team cannot use, start with the reason each enquiry went wrong. A person asking for a service you do not sell, a fake form submission and a suitable customer nobody called back need different fixes. Putting all three under “bad leads” makes the campaign harder to improve.

Consider two hypothetical calls to a commercial cleaning company. One caller wants a house cleaned before moving out. The other manages an office inside the company’s service area but reaches voicemail and never receives a callback. The first points towards a service mismatch. The second leaves demand unresolved and exposes a response problem. Excluding more keywords might help the first situation while doing nothing for the second.

The useful question is where the mismatch begins: the search, the ad’s promise, the page, the form, the measurement or the follow-up. Review actual enquiries alongside those parts of the journey, then make the smallest change that addresses a demonstrated problem. Watch qualified enquiry volume as well as the percentage that qualifies.

What a short lead review can reveal

Imagine a fictional office-cleaning business reviewing two campaigns after both groups of enquiries have had the same opportunity for follow-up. The business offers recurring office cleaning within a defined service area. It does not offer domestic cleaning or supply cleaning staff for employment agencies.

Campaign A produced 20 unique new enquiries. Eight meet the company’s service requirements, seven request household cleaning, two need service outside the coverage area and three remain unresolved because contact has not been established.

Campaign B produced 15 unique new enquiries. Nine meet the requirements, two request household cleaning, one needs service outside the area and three remain unresolved.

Campaign A has more enquiries, but Campaign B has more confirmed qualified enquiries in this hypothetical review: nine compared with eight. The qualified share is 8 ÷ 20, or 40%, for A and 9 ÷ 15, or 60%, for B. Those percentages describe these small invented groups; they are not performance targets or proof that B will scale better.

The strongest clue is the household-cleaning mismatch in A. Before moving budget, the reviewer inspects its available search terms, ad wording and destination. Suppose the ad says “Reliable Cleaning Services” and the page prominently describes both home and office cleaning, although the business stopped residential work months ago.

That evidence supports a specific correction: make the advertising and page accurately describe the service the company now offers. It does not establish that every broad query is poor, that the whole campaign should stop or that a different bidding strategy will solve the problem.

The six unresolved enquiries also need attention. They should stay visible while the team checks delivery, call handling and follow-up. Their suitability cannot be inferred from silence. Some may eventually qualify; others may not.

This modest review is enough to choose a useful next investigation. It is not enough to claim a statistically reliable campaign winner. The distinction lets a small business act on an obvious offer error without pretending it has completed a controlled advertising experiment.

Agree on what makes an enquiry suitable

A qualified enquiry is a genuine request that fits the service the business can provide and the relevant requirements it has agreed to use. Qualification establishes a sensible next step. It does not promise a sale.

For the office-cleaning company, the decision might depend on the type of premises, required location, recurring-service need and practical scope. For another business, a different set of facts will matter. Write the requirements in terms a customer and an intake team can understand.

A useful starting sentence is: “We can help when the customer needs this service, at this location, under these relevant conditions.” Then ask which facts must be confirmed before offering a quotation, consultation or appointment.

Keep the criteria close to what the business actually delivers. An unexplained score based on how enthusiastic a caller sounds is difficult to apply fairly and difficult to diagnose. A clear service mismatch is much more informative.

Avoid rejecting ordinary buying behaviour

Someone comparing prices can still be a suitable customer. Someone using the wrong technical term may have exactly the problem you solve. A person who prefers email may be able to proceed without a telephone conversation. Do not turn those ordinary differences into rejection rules unless they genuinely prevent the service from being delivered.

Consider a hypothetical caller asking an electrician to “fix the fuse box”. The company may use a more precise service name. The intake task is to understand the problem, not to reject the enquiry because the customer did not use the keyword in the campaign.

Likewise, a customer who chooses a competitor after receiving a quotation may have been fully qualified. Record the lost sale separately. If you label every unsold enquiry unqualified, you remove the distinction between attracting suitable demand and winning the work.

Capacity deserves its own note. A customer who needs an offered service in the right area may be suitable even when your team is booked. If the ad promised immediate availability, that creates an expectation problem to address. Redefining the customer as unsuitable hides the commercial choice about capacity and advertising.

Compare a few real decisions before changing settings

Ask the owner, intake lead and campaign manager to review a varied set of redacted enquiries. Include apparently good requests, obvious mismatches and uncertain cases. Discuss disagreements about why a request did or did not qualify.

This discussion often reveals a practical problem: the campaign manager is advertising work the delivery team no longer accepts, or the intake team is applying an unwritten minimum project size. Resolve those differences before interpreting a performance report.

Keep the final definition short enough that people will use it. If a requirement changes, record when it changed. Otherwise, a later improvement may reflect a different definition rather than better enquiries.

When people ask for the wrong service

Start with what the people actually requested. Group recurring mismatches in ordinary language: job applications, household work instead of commercial work, product purchases instead of repairs, or support for a service the business does not provide.

Then trace the likely expectation backwards. Read the available search-term themes, the ad’s wording, the destination and the choices in the form. Look for a place where a reasonable person could conclude that the wrong service is available.

Google’s search terms report shows actual searches for reported traffic, but it omits some low-activity queries for privacy. Treat the visible terms as useful evidence with incomplete coverage. Do not assign a particular search to an individual enquiry unless your permitted measurement genuinely connects them.

A keyword in a tracking field is also not necessarily the person’s search term. Confusing those two can make the diagnosis sound more certain than it is.

Choose between a query change and an offer change

Suppose the cleaning company sees employment-related searches in the report and receives job applications. Those observations justify examining the matching path and considering relevant exclusions. The business should still check whether the page itself contains recruitment material or an ambiguous “join us” action near the enquiry form.

Now suppose the reported searches are about office cleaning, yet people ask for one-off household cleaning. The destination may be creating the wrong expectation. Excluding more terms would be an indirect response to a page problem.

The best intervention depends on the evidence. A clearly irrelevant search theme may call for a negative keyword. A relevant search sent to an unsuitable page may call for a destination change. A broad but legitimate search may need clearer service wording so people understand the offer before enquiring.

For the detailed query decision, use the guide to high-intent search queries. The enquiry review supplies the business context that a search-term report alone cannot provide.

Be careful with words that serve more than one intent

“Jobs” often signals employment, but a contractor’s customer might describe a small repair as a job. “Cost” and “price” can be normal parts of choosing a provider. A single word does not always tell you the complete buying purpose.

Read the phrase and the pattern before applying a broad exclusion. Ask which useful searches the proposed change could also remove. Where the mismatch is narrow, a narrow control may be more appropriate than an account-wide rule.

A customer asking whether you offer a related service can also reveal unclear navigation. If the site has several business lines, separate their enquiry routes sufficiently that the receiving team can tell what the person wants. Advertising cannot compensate for a form that sends every service to an unlabelled inbox.

When enquiries come from outside your service area

Use the location where the work must happen. A caller’s area code, current location or billing address may be different. An owner living in Calgary could reasonably request cleaning for an office in Toronto.

Compare the actual service location with your coverage, the areas described in the ad and page, and the campaign’s location settings. Look for repeated misunderstandings rather than assuming every remote caller is wasted traffic.

Google distinguishes people likely to be in or regularly in a location from people who have shown interest in it. The broader Presence or Interest option can include both. Presence focuses on the former. Google also states that location accuracy is not guaranteed. See its advanced location options.

For a strictly local service receiving repeated requests for work elsewhere, a Presence test may be worth considering. For a business serving relocating customers or remote property owners, interest-based reach may have a valid purpose. The service model should decide which question you test.

Make the boundary understandable

“Serving the GTA” can mean different things to different readers. If the company has practical coverage limits, state them clearly enough for people to judge whether an enquiry is worthwhile. You may not need a long municipality list above every form, but you should not promise a region the delivery team routinely refuses.

A hypothetical moving company advertises “Ontario Moves” while accepting only moves that start within a particular local area. People planning moves elsewhere in the province are responding to a plausible interpretation of the ad. A more specific description of the service can reduce confusion without assuming those people ignored the page.

A short location question may help route the enquiry: “Where is the property that needs service?” Ask for the level of detail needed at that stage. A city or postal-code area may be enough to establish coverage; a full address may be unnecessary before the person decides to proceed.

After changing geography or wording, watch what happens to both out-of-area requests and qualified volume. A higher qualified percentage can look encouraging even while the business receives fewer useful opportunities. That trade-off belongs in the decision.

When the offer attracts the wrong expectations

Read the ad, page and form as one conversation. What is the person invited to do, what will they receive and what happens after they respond? If those answers change along the way, an apparently unsuitable lead may be a predictable response to the message.

Here is a hypothetical before-and-after example for a business that sells paid design consultations.

Before: “Free Design Help. Book Your Consultation.” The destination’s form asks for contact details and says the team will be in touch.

After, assuming it accurately describes the offer: “Discuss Your Renovation Project. Start with a Free Fit Call.” The page explains that the initial call establishes scope, while the subsequent design consultation is a paid service with its fee disclosed clearly.

The revision gives the first step a precise purpose. Someone seeking a full free design session can make a better decision before submitting. Someone who is suitable for paid work understands what the introductory call does.

This is an editorial hypothesis to test, not a claim that adding price information always improves performance. The right wording depends on the real service. The important point is that marketing and intake should describe the same offer.

Put important conditions where they can inform the decision

A minimum project size hidden in a long FAQ cannot reliably set expectations at a form that says “All Projects Welcome”. A service area mentioned only in the footer may be missed by someone arriving directly on a mobile landing page.

Ask whether an important condition is visible at the moment it affects the decision. You do not need to overload the page with disclaimers. You need enough relevant detail to prevent a predictable misunderstanding.

Use plain descriptions of fit. “Recurring office cleaning for occupied workplaces” is more informative than “Premium solutions for serious clients”. The latter sounds selective but leaves the actual service vague.

If the business cannot describe a minimum budget honestly before assessing the work, explain the quotation process rather than inventing a screening threshold. Ask questions that help determine scope and the next step. Avoid requesting private financial detail merely to make a lead feel more qualified.

Inspect the page people actually reach

Do not assume every click goes to the URL you remember setting months ago. Check the relevant destination reports and campaign configuration.

In Performance Max, Final URL expansion can use other relevant pages on the domain; a page feed alone does not restrict traffic to those pages while expansion is enabled. Google documents Performance Max URL expansion. Search campaigns using AI Max can also have Final URL expansion, so check whether that feature applies to the campaign you are reviewing.

A recruitment page, an outdated service page or a general article may create a different expectation from the intended enquiry page. If the evidence points to an unsuitable destination, examine the available URL controls. Do not assume an ordinary Search campaign uses every automation feature simply because it exists in the platform.

When forms contain spam or obviously false information

First establish that the submissions are non-genuine. Repeated nonsense, unrelated sales pitches or a clearly automated pattern are different from a prospect with unfamiliar spelling, a personal email account or limited English.

Keep examples and timestamps in the appropriate internal system, then ask the website owner to inspect how the submissions entered. Did they use the visible form? Did required-field validation run? Did a recent form change remove a safeguard? Is one route disproportionately affected?

Google explains that its invalid-traffic filtering does not by itself stop fake website leads. A bad form submission is therefore not automatic proof of a chargeable invalid click or entitlement to a refund. Its invalid-lead guidance separates advertising protections from safeguards on the advertiser’s own forms.

A developer may recommend server-side validation, rate controls or an appropriate bot challenge. Choose a measure that addresses the observed pattern and test the legitimate customer journey afterwards. A safeguard that prevents genuine customers from submitting creates a different problem.

Do not use friction as a substitute for understanding

Adding more fields can make a form harder to complete without proving that the person is genuine. A determined bot may still submit it; a real mobile visitor may abandon it.

Likewise, requiring a business email address may be reasonable for a particular business process, but it can exclude legitimate sole proprietors or small organizations using personal addresses. Decide whether the rule serves the actual service requirement.

If verification is needed, consider when it belongs in the journey. A confirmation step can help validate contact details, but it may delay an urgent request. The business should understand the operational trade-off before choosing it.

Review spam separately from service fit. A genuine household-cleaning enquiry sent to a commercial cleaner is unsuitable for that offer, but it is not spam. Keeping those reasons distinct makes the next fix easier to choose.

When good-looking enquiries are unreachable

Trace the delivery before blaming the audience. A form may save successfully while its email notification goes to junk. A CRM record may exist without an assigned owner. A phone call may reach an automated menu without reaching someone who can help.

Take one unresolved request and follow what happened: received, delivered, assigned, first useful response and next attempt. Look for the first missing step. This is often more revealing than comparing another campaign-level conversion rate.

If the team never reached the person, their fit may remain unknown. Keep any facts already established. A form that clearly requests an unavailable service is unsuitable even if the person does not answer later. A known suitable request does not become unsuitable merely because the conversation stops.

Read call metrics as clues

Google’s call reporting can provide information such as start time, duration and whether a call connected when the relevant feature is supported. Those observations help investigate handling. They do not independently establish that a conversation was commercially suitable.

A long call might include hold time. A short call might efficiently confirm a relevant appointment. Use the actual conversation outcome, when lawfully available and appropriately reviewed, to understand suitability. This guide does not require recording calls or changing recording settings.

Suppose a hypothetical company reports that evening enquiries rarely qualify. Its advertising owner plans to stop evening delivery. Before doing so, the team discovers that evening calls go to a voicemail box nobody checks until several days later.

The proposed schedule change would remove enquiries from that period, but it would not establish that the demand was poor. The immediate operational question is whether the business can provide an appropriate callback or another clear response route. If the advertised service requires immediate assistance that the company cannot provide at night, the promise and schedule may genuinely need to change.

For a deeper diagnosis of phone outcomes, use the call-tracking guide. For assigning and responding to enquiries, see the speed-to-lead response guide. Here, the important decision is whether the observed problem belongs to demand or handling.

Make the next step obvious to the customer

An automated acknowledgement can reassure someone that a request arrived, but it is not the same as a useful response. Explain who will respond, how and within a timeframe the team can actually support.

Honour the person’s stated contact preferences and use permitted channels. Repeated calls at inconvenient times can make a viable enquiry harder to develop. A planned project and an urgent repair need different response expectations.

The intake team should also know what to do after an unsuccessful attempt. A request without a next action can remain apparently active for weeks while nobody owns it. A simple assignment and follow-up process is usually more useful than adding another “lead score”.

Check whether the report counts the thing you care about

Google Ads conversions, form submissions, telephone calls, CRM contacts and buying requests are different units. A person may submit a form, call later and create more than one record about the same project. A broken tag may count one accepted submission twice.

Before judging quality, compare a small set of source records across the systems. Confirm what caused the conversion, whether a genuine enquiry reached the business and whether repeated contact belongs to the same request.

Google’s conversion-counting options include One and Every. One operates for a conversion action after an ad interaction; it does not deduplicate every person or buying request across a CRM. A configuration choice cannot replace understanding the unit you want to count.

Separate obvious spam, repeat contact about the same request and existing-customer support from the new-demand review. Keep those records for their own purposes. Existing customers may be commercially valuable, but counting their support calls as new-customer acquisition gives the campaign a different meaning.

Inspect the goals the campaign is using

Write down the important conversion actions and what actually triggers each one. Is it a submitted enquiry, a click on a phone link, a booked appointment or a later confirmed outcome? A campaign receiving only initial form signals cannot infer your private sales classification unless an appropriate feedback process supplies it.

In standard goals, primary conversion actions are used for bidding when the relevant goal is selected; secondary actions are generally observational. Custom goals are an important exception because their included actions can be used for bidding regardless of that label. Google explains the distinction in its primary and secondary conversion guidance.

Do not move settings around until you understand the campaign’s actual goals and the reliability of the replacement signal. Changing a label while leaving the same weak action inside a custom goal may not accomplish what the team intended.

If qualified outcomes should feed back into Google Ads, treat that as a separate implementation project. The offline-conversion guide covers mapping and validation. A lead-quality review can define what should count; it is not itself a conversion-import file.

Leave unattributed enquiries unattributed

Use reliable permitted source evidence where available. A prospect saying “I found you on Google” might mean an ad, organic search or Maps. Preserve that self-report without turning it into a precise campaign attribution.

Missing source data may limit how far you can take a conclusion. You can still observe that the business receives many wrong-service requests and correct a misleading page. You may not be able to say which campaign caused each request.

Avoid filling gaps by matching nearby timestamps or choosing the most plausible keyword. That creates confident-looking data without a reliable connection. Keep an unknown-source group visible and improve measurement separately where appropriate.

When a qualifying question could prevent a predictable mismatch

A form can help establish fit when it asks for information that changes the next step. Start with the decision the receiving team needs to make. Then ask the simplest question that supplies it.

For the fictional cleaning company, “What type of premises need cleaning?” could distinguish an office, another commercial property, a home or an uncertain request. That answer can guide routing and make the offer clearer. A generic “Tell us about yourself” box asks the customer to guess which details matter.

The form should explain what happens when a service is unavailable. If residential cleaning is not offered, make that clear before someone spends time completing the enquiry. Where appropriate, a brief explanation beside the service choices may be enough. Avoid a misleading form that accepts every choice and leaves the sales team to reject people later.

Compare two ways of asking about scope

Imagine that the company needs an approximate property size to decide which team should respond. One form requires an exact measurement before it can be submitted. Another asks for a practical size range and offers “I’m not sure”.

The exact field may look more rigorous, yet many legitimate buyers will not know the measurement. A range can provide enough information for routing while allowing the team to clarify later. The appropriate choice depends on the decision, not on a belief that longer forms always produce better leads.

Suppose the business truly cannot assess a request without a particular detail. Explain why it is needed and help the person find or describe it. If the detail is only required before preparing a final quotation, consider collecting it later instead.

The same reasoning applies to budgets. If there is a genuine minimum project commitment, state it accurately. If the price depends on scope, forcing a precise budget answer may produce guesses that the sales team later mistakes for verified buying capacity.

Watch what the new question excludes

After adding a question, review both the answers and the completion experience. Test it on a phone, check that labels are understandable, and ensure an error does not erase the person’s work. A form improvement should remain usable by the customers you want to help.

If the question produces many “Other” or “I’m not sure” answers, read a sample of the requests before adding more restrictive choices. Customers may describe the service differently from your internal departments. Their uncertainty can identify a wording problem rather than poor fit.

Ask intake whether the information changed what it did. If the new answer is never used for routing, qualification or the first response, the field may be adding friction without a corresponding benefit.

Also watch confirmed qualified volume. A lower submission count can be sensible if it mainly removes clearly unsuitable requests. It can be harmful if the business loses suitable customers who found the question confusing or intrusive. The enquiry rate alone cannot distinguish those possibilities.

For the detailed design of fields, labels and error recovery, use the lead-form optimization guide. The lead-quality decision comes first: identify the uncertainty you need to resolve, then design the question around it.

When several problems appear in the same review

You may find wrong-service requests, fake submissions and missed callbacks together. You do not have to choose one explanation for the entire account. Separate the immediate repairs from the uncertain campaign decisions.

A notification failure can be repaired while the marketing team investigates search intent. A misleading service claim can be corrected while a developer assesses spam controls. These independent actions address different failures, so waiting for a single grand diagnosis would delay useful work.

What you should avoid is treating the combined outcome as proof of one tactic. If the page, form validation and response process all change in the same week, the next review may look better for several reasons. Describe the improvements at the level the evidence supports.

Prioritize by consequence and clarity. A verified failure that prevents suitable requests from reaching the team deserves prompt attention. An ambiguous search phrase with one recent enquiry may need more investigation. This keeps the review focused on recoverable business problems while preserving room to learn about less obvious causes.

Keep sales feedback small enough to use

For the next review, you need a request reference, source evidence when known, the requested service, the fit decision, the reason and any response problem. A brief factual note can explain a borderline case. Detailed customer records should stay in the operational system that is authorized to hold them.

The reason should suggest what to investigate. “Bad” is too vague. “Needs residential cleaning; campaign promotes office cleaning” is useful. “No callback completed; service fit unknown” directs attention to a different owner.

Allow uncertainty. Intake staff should not have to invent a reason to close a required field. A person may need to clarify the location or scope before anyone can make a fair decision.

The existing lead-quality feedback kit provides a more detailed CSV, data dictionary and fictional records for teams that need a structured review. It is an internal analysis aid, not a Google Ads upload template. Start with the information that changes your decisions; use the additional fields when they answer a real operational question.

Keep access and retention proportionate. Names, contact details, precise addresses and sensitive conversations usually do not belong in a campaign summary. Pseudonymous references can still be linked to people, so handle them appropriately rather than treating them as automatically anonymous.

A CRM makes this easier when it gives each request an owner, current state and next step. Software cannot settle a disagreement about what the business considers suitable. Agree on the meaning first, then configure the system to preserve it.

Choose one change that follows from the evidence

Return to the recurring symptom. If wrong-service enquiries share misleading wording, correct the wording. If unsuitable destinations are involved, inspect destination controls. If real requests are lost after submission, repair the handoff. If confirmed fake submissions dominate one form, investigate that form’s safeguards.

Some problems should be repaired immediately rather than left in place for an experiment. A broken notification route or an offer the business no longer supplies is an operational fault. Record the repair and its timing so later comparisons remain understandable.

For an uncertain improvement, write a plain prediction before changing anything: “Making recurring office cleaning explicit should reduce household enquiries while preserving suitable office requests.” Then decide which observations would support or challenge it.

Watch the underlying counts, not just a favourable rate. If suitable requests fall from eight to four while unsuitable requests fall further, the qualification percentage may rise while the business receives less useful demand. Whether that is acceptable depends on costs, capacity and the value of the opportunities.

Keep major unrelated changes out of the same comparison where possible. Changing the offer, geography, form and bidding together makes it difficult to learn which intervention mattered. A supported campaign experiment can help for eligible changes; a before-and-after review may be the practical option elsewhere. The Google Ads experiments guide covers that design in more depth.

Decide whether quality actually improved

Compare enquiry groups that have had a similar opportunity to develop. Yesterday’s unanswered form submissions should not be judged against last month’s fully reviewed requests. Choose a follow-up period that reflects how your business establishes fit, and keep unresolved enquiries visible.

Use the same qualification definition. If the sales team tightens its minimum project requirements halfway through the period, explain that change rather than attributing the resulting shift entirely to the campaign.

Check whether handling improved at the same time. Better response coverage can increase confirmed qualified volume by resolving unknowns. That is valuable, but it is different from proving that the advertising attracted a new type of customer.

With a small sample, a few enquiries can move percentages sharply. You can make a cautious operational decision without announcing a proven winner. Keep collecting relevant evidence when the decision remains uncertain, and avoid extending a test indefinitely without a clear question it can answer.

For the fictional cleaning company, the first review should ask whether household-cleaning confusion decreased after the offer was clarified, whether office-cleaning demand held up and whether unresolved requests received proper follow-up. That completes the reasoning started by the original complaint. It also tells the team what to investigate next if the result disappoints.

A useful lead-quality meeting ends with one explanation supported by the records, one action, an owner and a time to review the outcome. If you need help connecting those enquiries to your advertising, discuss your Google Ads management with Canada Create. Bring a redacted sample of suitable, unsuitable and unresolved requests so the conversation can start with what is actually happening.

Frequently asked questions

Why are my Google Ads leads poor quality?

Start with the reason each enquiry went wrong, since a request for a service you do not sell, a fake form and a good customer nobody called back need different fixes. Grouping all three as bad leads makes the campaign harder to improve. Find where the mismatch begins: the search, the ad's promise, the page, the form, the measurement or the follow-up.

How do I identify a quality lead from Google Ads?

A qualified enquiry is a genuine request that fits the service you can provide under the requirements you have agreed to use. Write those requirements in terms both customers and your intake team understand, such as type of premises, location and scope. Qualification sets up a sensible next step, which is different from a promised sale.

Does a high Quality Score mean better Google Ads leads?

Not necessarily. Quality Score is a keyword diagnostic based on expected click-through rate, ad relevance and landing page experience, and it does not judge whether enquiries suit your business. Measure lead quality separately by reviewing real enquiries against your qualification rules.

Is a high click-through rate a sign of good Google Ads leads?

Click-through rate shows how often people click after seeing the ad, not whether they need what you sell. An ad promising something vague can draw plenty of clicks from the wrong people. Track qualified enquiry volume alongside the share that qualifies.

How do I measure whether Google Ads lead quality improved?

Compare qualified enquiry volume as well as the qualified share, using groups that had the same time for follow-up. In the guide's invented review, one campaign had eight qualified out of 20 and another had nine out of 15. Small groups like that support a next investigation, not a firm winner.

Why do Google Ads enquiries come from outside my service area?

Location settings, the wording of the ad and the page can all invite people you cannot serve. Make the boundary clear in the ad and on the destination, and check the location option the campaign uses. Ask where the work is needed before rejecting a caller from elsewhere.

How do I reduce spam form submissions from Google Ads?

Protect the form against obvious junk, but do not rely on friction alone. Extra hurdles can put off real customers while determined spammers carry on. Review where spam clusters, such as specific campaigns or placements, and fix that source.

Should I add a qualifying question to my Google Ads form?

Add one when it prevents a predictable mismatch, such as people asking for a service you do not offer. Compare two ways of asking about scope and watch what the new question excludes, since it may also turn away suitable customers. Keep sales feedback short enough that the team will actually use it.