← Back to blog

Click Fraud Detection for Dealerships: A Practical Guide

November 5, 2026
Click Fraud Detection for Dealerships: A Practical Guide

Click fraud detection comes down to three checks. Compare Google's click data with your own lead and DMS data. Look for patterns that no single metric proves. Escalate to Google with logs and exact dates when the pattern holds. For a dealership, the question is rarely "is there fraud?" It is "how much of my paid search spend produces real shoppers, and who can prove it?" This guide shows what invalid clicks look like, what Google already filters, how to investigate, and how to tell fraud from ordinary underperformance.

Autoroiq
See Where Marketing Spend Works
Get an independent view of dealership marketing performance, with vendor-agnostic analysis that helps clarify where budgets are being used effectively.
Explore independent analysis

TL;DR

  • Click fraud is only one part of invalid traffic. Google's own definition also covers accidental clicks, ad stacking and competitor clicks, and it filters much of it before you are billed.
  • No single signal proves fraud. High CTR, short sessions or shared IP addresses can all have legitimate causes, so look for several signals together.
  • Compare Google click data with CRM, call and DMS outcomes. Spend that produces clicks but no real leads is the clearest sign something is wrong.
  • Tighten exposure (location, networks, placements, keywords, schedules) and protect lead forms before you spend time chasing individual clicks.
  • Google accepts investigation requests for recent activity. Prepare your evidence before you file.

Table of Contents

What click fraud is and what it looks like in a dealership account

Click fraud happens when a person or bot pretends to be a legitimate visitor and clicks an ad to make a platform record a genuine interaction. Cloudflare's overview lists the common forms: bots and botnets, click farms, competitor clicks, publisher-driven clicks, and clicks meant to inflate click-through rates. Huntress adds that human click farms are harder to catch because they resemble real users.

Google uses a broader term, invalid traffic. Google Ads Help defines it as traffic that does not represent genuine interest in your business. Its examples include bots and automated software, accidental clicks, clickjacking, ad stacking, competitor clicks and publisher-incentivized clicks. The distinction matters. Not every wasted click is fraud, and not every wasted click is credited.

For a franchise dealership, exposure shows up in a few predictable places:

  • Branded and model-specific search campaigns. Competitor and curiosity clicks gather here because the keywords are valuable and the CPCs are visible.
  • Display, partner and app inventory. Placements you never chose can produce high CTR with no downstream activity.
  • Lead forms. Credit application, trade-in valuation and "check availability" forms attract bot submissions that look like leads in the platform and fail the first phone call.
  • Broad location settings. Users outside your market can click, which looks like fraud in the report but is a targeting configuration problem.

What Google already filters, and what it does not

Start from what is handled for you. Google says it uses automated filters, machine learning and manual reviews to detect invalid traffic. Activity caught before billing is removed from campaign metrics and billing. Activity found after billing may be credited where appropriate and possible, according to Google Ads Help.

Google Ad Traffic Quality states that it uses more than 200 filters to stop most invalid traffic in real time or soon after it occurs. Its filters can draw on signals such as denylisted user agents or IP addresses, unusually high click-through rates, repeated traffic from a single user, and patterns that suggest intentional, accidental or non-human interactions.

That gives you a useful baseline for your own review. If a pattern is obvious enough to trip a filter, Google probably removed it. The remaining risk sits in traffic that looks plausible at the click level but never becomes a shopper. Understanding how filters, monitoring and manual review work together is also useful for anyone running automotive campaigns, and this walkthrough of Google Ads setup and tracking covers the account-level basics.

What Google already filters, and what it does not

Read the Invalid Activity Credit Report first

Before investigating anything, open the credit report. The Invalid Activity Credit Report shows credited clicks, credited interactions and the credited amount. It also shows adjusted cost, clicks, CTR, average CPC, conversion rate and cost per conversion.

This changes how you read performance trends. Google notes that some activity is removed automatically and later adjustments can change reported results. A cost-per-lead swing may be partly a reporting adjustment, not a new problem. Check the report, billing and adjusted metrics before you conclude that something unusual is happening.

Signals worth investigating

Research on invalid traffic points to a consistent set of warning signs. Anura lists sudden CTR increases, high traffic without conversions, high bounce rates, short sessions, unexpected geographies, repeated interactions from related IP ranges, unusual click frequency or timing, and unexplained performance changes.

The same source is clear that these are not proof. Multiple legitimate users can share an IP address, and a high bounce rate can have an ordinary explanation. A practical detection guide describes IP and network analysis, geography, device and browser characteristics, behavior, click velocity and conversion outcomes as complementary signals. No single one, including a low conversion rate or short sessions, proves fraud.

The working rule: one anomaly is a question, three independent anomalies are a case.

SignalWhere to lookCommon innocent explanation
CTR jumps with no change in ads or bidsGoogle Ads campaign and keyword reportsA new promotion, seasonal demand or a competitor pausing
Clicks with near-zero engagementAnalytics sessions, VDP views, time on siteSlow mobile page, mismatched landing page
Clicks from outside your marketGeographic report, server logsLocation option that includes people showing interest in an area
Bursts at odd hoursHour-of-day reportAd scheduling left open overnight
Many form fills, few reachable peopleCRM, call tracking, BDC notesTypos, shoppers using throwaway emails
High CTR on a few placements, no conversionsPlacement and domain reportsPoorly chosen inventory rather than fraud

The last column matters as much as the first. Targeting configuration is not fraud, and Search Engine Journal specifically notes that broad location eligibility can admit users outside the intended service area. Fixing the setting costs nothing and solves the problem immediately.

A step-by-step investigation for a dealership

The order below keeps the work proportionate. Most accounts get answers by step 4.

  1. Define the symptom. Write one sentence: which campaign, which metric, which dates. "Used-vehicle search leads rose 40% in a week while appointments stayed flat" is a symptom. "Our numbers feel off" is not.
  2. Check the credit report and billing. Confirm what Google already removed or credited, as described above.
  3. Audit placements and domains. Search Engine Journal recommends reviewing placement and domain reports, checking whether spend goes to intentional targets, and excluding unfamiliar or clearly low-quality sites and apps when campaign inventory is responsible.
  4. Compare Google's clicks with your own data. Search Engine Land recommends comparing internal lead data with Google click data and excluding low-quality apps or sites with high CTR and no conversions. Pull server logs with IP addresses, devices and GCLIDs, and match them to CRM leads.
  5. Grade lead quality, not lead count. Have the BDC mark each lead in the suspect window as real, unreachable, duplicate or junk. Keep the GCLID with each assessment.
  6. Follow the lead to the DMS. A lead that never becomes an appointment or a deal is a cost, whatever caused it. Our paid search audit approach ties campaign spend to DMS sales before any bid changes.
  7. Decide: fix, filter or escalate. If the cause is targeting or inventory, fix the settings. If clicks look automated or coordinated and Google did not catch them, escalate.

Where Performance Max and automated campaigns complicate this

Automated campaign types spread spend across inventory, and reporting is thinner than in manual search. That makes placement review and lead-level validation more important, not less. Conversion actions that count low-value events, such as a page view or a click-to-call tap, can make weak traffic look productive. Our Performance Max audit framework covers how to verify what a conversion actually means before you judge traffic quality.

Escalating to Google: what to submit

If you suspect unfiltered invalid traffic, Google Ads Help says to submit an invalid-traffic investigation request for activity within the last 60 days. That window is the reason to review monthly instead of quarterly. Search Engine Land lists what to include:

  • Your Google Ads customer ID
  • Exact dates of the suspicious activity
  • Campaign and ad group names
  • Suspicious keywords
  • Server logs with IP addresses, platforms or devices, and GCLIDs
  • For invalid leads, GCLIDs and your lead assessments
  • A short explanation of the trend you see

The quality of the request depends on the quality of your records. A dealership that keeps GCLIDs attached to CRM leads can file a request in an afternoon. One that only has a platform lead count has little to send. Outcome is never guaranteed, and credits apply where Google finds it appropriate and possible. Record the date, the request and the response so the result can be reconciled against your spend report later.

Reducing exposure before fraud reaches you

Prevention is mostly exposure management. Google recommends using its automated protections, setting location options and negative keywords, refining content suitability, focusing on lead goals, avoiding untrusted traffic partners and never clicking your own ads. Its guidance for preventing invalid clicks also recommends understanding your visitors through segmented traffic reports and Google Analytics.

Applied to a dealership account, that translates to:

  • Location. Use presence-based targeting where it fits your market, and review the geographic report for out-of-market clicks.
  • Networks. Turn off search partners or display expansion unless you can show they produce appointments.
  • Placements. Exclude low-quality apps and sites with high CTR and no conversions, as Search Engine Land advises.
  • Keywords. Add negatives for job seekers, parts, DIY repair and other intents that do not buy vehicles.
  • Schedule. Compare hour-of-day performance against when your BDC and showroom are staffed.
  • Conversion goals. Optimize toward qualified leads, such as appointments set or calls over a minimum length, instead of raw clicks.

Protect the lead form

Google's guidance mentions securing lead forms with reCAPTCHA, server-side validation or double opt-in. Search Engine Journal's lead-generation guide recommends layers: automated detection as an initial signal, form validation such as honeypot fields and conditional conversion triggers, and manual review of patterns and lead quality.

It also describes a dealership-related example where a much longer form, with specific questions and a vehicle-photo upload, cut signup volume while materially improving lead quality. The same guide warns that longer forms can reduce legitimate completion rates. For a trade-in or appraisal form, a photo upload step is a reasonable friction point because a real owner has the car. For a simple "confirm availability" form, keep the form short and rely on validation behind it.

Fraud, or just underperformance?

Many accounts that feel defrauded are simply misattributed. Call tracking may credit the wrong source. Form leads may be duplicated across vendors. A third-party site vendor and an ad agency may each claim the same appointment. When two vendor reports contradict each other, fraud is the least likely explanation and attribution is the most likely.

Use this quick test:

  • Clicks look wrong and leads look wrong: investigate for invalid traffic.
  • Clicks look right but leads do not convert: examine landing page, inventory availability, pricing and BDC response time.
  • Platform leads look strong but DMS sales do not move: examine conversion definitions and attribution. Our guide to lead source attribution walks through how to reconcile sources.
  • Website metrics shift without any campaign change: review site behavior and tracking first. See our dealer website metrics fix for executives.

This is also the reason a detection tool alone rarely settles the matter. A tool can flag suspicious clicks, but it cannot tell you whether a legitimate click turned into a vehicle sale.

Our perspective

We review dealership marketing without selling advertising, managing campaigns or taking vendor commissions. That position shapes how we look at click fraud.

The usual conversation about invalid traffic starts at the click. Dealership leaders are better served starting at the outcome. When a vendor's report shows strong lead volume and the DMS shows flat sales, the first job is to establish which numbers are defensible. Sometimes the answer involves invalid traffic. More often it involves conversion definitions, duplicate leads, loose targeting or attribution that favors whichever vendor built the report.

Our marketing intelligence work is designed to give executives one independent view across channels. We compare spend to lead quality and downstream sales, so that a suspicious pattern can be traced to its cause and a budget decision can be defended in front of an owner or a factory representative.

A dealership that can show Google exact dates, campaign names and lead assessments is in a far stronger position than one that can only say it feels overcharged. Independent records are what make that possible.

AutoROIQ

Working with an independent reviewer

If your vendors' reports disagree and you cannot tell whether traffic is the problem, our Marketing Intelligence Review is built for that situation. We evaluate vendor performance and marketing channels without any advertising to sell, and deliver executive-level recommendations and vendor accountability scorecards. Dealerships that want continuing oversight can add ongoing strategic advisory. Details are on our Marketing Intelligence Review page, and you can see the broader approach at AutoROIQ.

Autoroiq

You can also browse our marketing intelligence insights for channel-by-channel guidance, including Facebook versus Google for dealerships.

A monthly routine that keeps you inside the review window

Because investigation requests cover only recent activity, a short monthly routine does more than a large annual cleanup.

  1. Open the Invalid Activity Credit Report and note credited amounts and adjusted metrics.
  2. Scan the geographic, placement and hour-of-day reports for outliers.
  3. Export server logs or analytics data with GCLIDs for any campaign that moved sharply.
  4. Have the BDC grade a sample of leads from those campaigns.
  5. Match graded leads to appointments and DMS sales.
  6. Log changes to settings, exclusions and conversion goals so later results can be explained.
  7. Escalate within the window if the evidence supports it.

Twenty to thirty minutes per campaign group is usually enough when the data is already connected. If the data is not connected, that gap is the first thing to fix, because every other check depends on it.

FAQ

Does Google refund money lost to click fraud?

Google removes invalid activity it finds before billing. For activity found after billing, it may issue credits where appropriate and possible, and the Invalid Activity Credit Report shows what was credited. If you suspect activity that was not filtered, you can submit an investigation request for the last 60 days.

Are high bounce rates and repeat IP addresses proof of click fraud?

No. Multiple legitimate users can share an IP address, and high bounce rates have many ordinary causes, such as slow pages or mismatched landing pages. Treat them as prompts to look at more signals, such as geography, click timing, device patterns and lead outcomes.

What do I need to send Google to report suspected invalid clicks?

Prepare your customer ID, exact dates, campaign and ad group names, suspicious keywords and server logs with IP addresses, devices and GCLIDs. For invalid leads, add GCLIDs and your assessment of each lead, plus a brief explanation of the trend.

Can competitors really click my dealership's ads?

Google lists competitor clicks among examples of invalid traffic, and it uses filters to catch repeated traffic from a single user. Not every pattern of repeat clicks is deliberate, though, so check geography, timing and lead outcomes before concluding that a competitor is responsible.

What is the fastest way to reduce wasted paid search spend?

Tighten exposure first: location options, networks, placements, negative keywords and ad schedules. Then optimize toward qualified leads instead of raw clicks, and add validation such as reCAPTCHA or double opt-in to your forms.

Sources