If clicks rise for an HVAC or cleaning campaign while calls and requests do not, it can feel as if a competitor is clicking the ads. But repeat visits or high click volume alone do not identify the visitor or prove a competitor caused it. Separate three records first: Google Ads clicks it filtered as invalid, website sessions in analytics, and qualified requests the service team can verify.
Google says it filters most invalid traffic before billing, adjusts campaign metrics, and may issue credits when applicable. Advertisers can add the Invalid clicks column to review filtered activity. Google does not disclose details about individual clicks that could help people bypass its detection systems. If specific concerns remain, Google provides an investigation form for other activity it did not filter; its current guidance limits the lookback to the last 60 days. Google Ads Help: Invalid traffic
Find where the numbers diverge
Compare the same dates and time zone, and write down each metric’s definition and source. Google Ads reports ad interactions under Google’s rules; analytics records website sessions; a phone system or CRM records inquiries that actually arrived. The counts can differ because a click may not load the page fully, tagging and consent differ, attribution windows vary, or a call starts without a website session.
For a defined period, collect campaign clicks, cost, Invalid clicks and any displayed adjustments; analytics sessions and landing-page events; call attempts and connected calls; and CRM inquiries classified by service area, qualification, and outcome. Note the exact reporting window, currency, and any tracking outage or site change.
Check service-area and lead routing before suspecting fraud
Review location targeting and presence settings, excluded areas, search terms, ad schedule, and landing-page promises. A real person outside your service area may click if settings or campaign behavior permit it; a qualified customer may also be uncontactable or call later. Check that location names and service claims match areas the team actually covers.
Test the mobile page, phone link, form, call tracking, and CRM handoff. Confirm the phone was answered, the request reached the correct team, and the person asked for a service the business provides. A click on a phone button is not a connected call, and a connected call is not automatically a qualified lead.
Request an investigation with a clean evidence pack
If the discrepancy still suggests unfiltered invalid activity, preserve the date range and time zone, campaign and account identifiers, click and cost reports, invalid-click column, website-session definitions, service-area settings, call records without unnecessary personal data, and known changes. Submit the official Google investigation form for eligible recent activity. Describe the discrepancy and evidence precisely; do not accuse a named competitor without evidence. Google may not disclose individual click details, and a request does not guarantee a credit or finding.
Short-video script: Clicks do not identify who clicked
Hook: “HVAC ads got clicks but few calls. Can you conclude a competitor clicked?”
Beat 1: “Compare Google Ads invalid-click data with site sessions and verified CRM or call outcomes.”
Beat 2: “Check service-area settings, page load, call routing, and tracking before assigning a cause.”
Beat 3: “If evidence remains, use Google’s official investigation process for eligible recent activity.”
Close: “A click count alone does not identify a person or prove fraud.”
Educational carousel: Diagnose strange Google Ads clicks
- Cover: “More clicks, fewer calls? Separate the evidence.”
- Google Ads interactions are not the same as website sessions.
- Invalid clicks are filtered activity; use the report column.
- Check location targeting against the actual service area.
- Test page load, call tracking, and CRM handoff.
- Classify requests: eligible, out of area, unanswered, unavailable, or incomplete.
- Preserve a recent date range and use the official investigation form when warranted.
- Do not name a competitor or expect a credit without evidence.
Field experiment thread: Audit the click-to-call path
Proposed diagnostic plan, not a fraud finding.
1/ Select a recent period and record its time zone and reporting lag.
2/ Compare Ads clicks and Invalid clicks with analytics sessions using each system’s definition.
3/ Review service locations, search terms, exclusions, and ad schedule for the same period.
4/ Test a mobile landing page, phone link, call tracking, form, and CRM routing.
5/ Classify redacted inquiries by service-area fit, contactability, and actual outcome.
6/ Record known changes or outages; separate unexplained discrepancy from proven invalid activity.
7/ If evidence remains within Google’s current eligibility window, submit the official investigation request and log its outcome without presuming a credit.
Tool-neutral prompt pack: Reconcile click and lead evidence
Metric reconciliation: “Compare these redacted records [data] for the same dates and time zone. Keep Google Ads interactions, invalid clicks, analytics sessions, calls, and CRM requests separate. Explain which definition or gap may account for each difference; do not identify a clicker.”
Service-area review: “Review these campaign location settings, search terms, landing page, and service map [data]. Flag mismatch between targeting and actual coverage. Do not infer competitor activity.”
Lead-quality classification: “Classify redacted requests as qualified, out of area, not reached, service unavailable, or incomplete using these rules [paste]. Remove unnecessary personal data and do not label a case fraud without evidence.”
Investigation pack: “Organize the supplied recent discrepancy into a factual summary for Google’s official process: dates/time zone, account/campaign, reported counts, invalid-click column, session definition, targeting, call/CRM evidence, and known changes. State limits and avoid unsupported accusations.”
Stop-doing swipe file: One number does not prove a competitor clicked
Stop: Accuse a competitor because clicks rose on one day. Use: Compare the same period across invalid-click reporting, sessions, and CRM outcomes.
Stop: Block an IP address from an unverified third-party report. Use: Verify the report source and your own settings; avoid blocking real customers behind shared or changing addresses.
Stop: Demand the identity of each click from Google. Use: Use available reports and the investigation form, recognizing that sensitive click details are not disclosed.
Stop: Count a phone-button click as a qualified lead. Use: Verify call connection and classify its result in the CRM.
Stop: Treat every out-of-area request as fraud. Use: Review targeting, search terms, and service coverage first.
Hook–Value–Template: Document the discrepancy before an investigation
Hook: “Before calling clicks fake, did you compare the right sources?”
Value: Reconcile filtered Ads data, site sessions, and verified requests, then review targeting and tracking.
Copy-ready note: “For [dates/time zone], the campaign recorded [clicks, cost, Invalid clicks]. Analytics recorded [sessions and definition]. CRM/calls recorded [requests and qualifications]. Targeting covers [areas] under [setting]. The discrepancy began [date] alongside [known changes]. We cannot identify the people behind interactions. We are asking Google to review [eligible period] through its official process.”
Illustrative example: If analytics has fewer sessions than Ads clicks while Invalid clicks are also reported, record the difference and test page load and tagging before claiming fraud.
This is a diagnostic method; it does not establish who caused activity or guarantee a credit.