Google defines invalid clicks as "clicks on your ads that don't represent a genuine interest in your business" and lists three kinds: accidental clicks "that occur because of poor ad placement," automated clicks "from automated tools, bots, or other deceptive software," and manual clicks "that increase your ad costs or inflate publisher profits." Its invalid traffic page adds: "This includes competitors who manually click to increase your spend." See Google's invalid clicks definition and About invalid traffic.
Those definitions settle what counts as fake. They do not answer the operational question: is this account paying for traffic that will never become revenue? This guide covers the fraud types, what Google charges for, the warning signs, and a check, respond, prevent and recover plan with a detailed companion guide for each step.
What types of fake clicks and Google Ads fraud are there?
There are six common types, and only the first three are clicks in the narrow sense. The other three corrupt what happens after the click, which is often more expensive because Smart Bidding learns from it.
| Fraud type | What it is | Where it shows up | What to watch |
|---|---|---|---|
| Manual or repeated clicks | People clicking with no buying intent, including competitors | Search, Shopping, high-CPC campaigns | Repeated sources, spend spikes, no qualified leads |
| Bot traffic | Automated tools or scripts clicking ads or loading pages | Search, Display, PMax, Demand Gen | Short sessions, no engagement, data-center or proxy networks |
| Accidental clicks | Taps caused by poor ad placement | Display, PMax, Demand Gen, video, apps | Placement clusters with spend and weak engagement |
| Fake leads | Junk or automated form fills after a paid click | Landing-page forms, lead form assets, calls | Invalid contact data, duplicates, low sales acceptance |
| Conversion fraud | Conversions that are real in the tag but worthless in reality | Any campaign using automated bidding | Conversions that fail downstream CRM review |
| Attribution abuse | Credit taken by traffic that did not create the demand | Affiliate, partner, remarketing flows | Assisted conversions with no new pipeline |
Manual and repeated clicks
This is the classic version of click fraud: paid clicks with no buying intent, from a competitor, a click worker, or someone with a grudge. The practical signal is repeated cost without meaningful engagement or qualified leads. Do not judge it by one click. Judge it by pattern.
Bot traffic
Bot traffic is automated behavior that interacts with ads or landing pages. Some bots are obvious. Others mimic normal browsing well enough to pass basic checks. Useful signals include short sessions, repeated device patterns, proxy or data-center indicators, odd locations, and zero page interaction after expensive clicks. Our Google Ads bot traffic guide covers detection in detail.
Accidental clicks and low-quality placements
Display, Performance Max, Demand Gen, and video campaigns reach broad inventory. Some of it produces accidental clicks, weak attention, or traffic that never matches the offer. The fix is not to block every broad campaign. It is to review placements, apps, channel segments, and conversion quality before expanding spend.
Fake leads and conversion pollution
For lead-generation teams, fake leads can be worse than fake clicks. If raw form fills count as primary conversions, Smart Bidding learns from bad data and chases more of it. Import qualified lead stages, validate contact details, and separate raw inquiries from sales-accepted leads. Monitoring tools that work on your site, ClickFortify included, can only see the landing-page path. Lead form asset submissions, call-asset calls and call-only calls never reach your page, so those need CRM-side checks.
Does Google charge for fake clicks?
Not for the fake clicks Google catches, and Google catches them at three points. Its About invalid traffic page says: "You aren’t charged for invalid traffic detected before billing."
- Before billing. When Google detects invalid traffic, "Google filters it out automatically before you receive your invoice." These clicks show up in the invalid clicks column, not on your bill.
- After billing. "If Google finds invalid traffic after you receive your invoice, you'll receive a credit where appropriate and possible." The credit arrives as a billing adjustment, and Google says it "automatically issues an invalid activity credit in the next billing cycle, when appropriate and possible."
- On request. If you suspect invalid traffic that was not filtered, "you can submit an invalid traffic investigation request for activity within the last 60 days." Google states the limit plainly: "Google limits investigations to the past 60 days."
The limitation is in the wording. Google decides what is invalid, credits only "where appropriate and possible," and does not show which clicks it removed or why. A human who clicks with no intent to buy, and is not flagged, is billed like a customer. The 60-day window also means evidence older than two months cannot be used for an investigation request, so collect it as you go.
How do you check, respond to, prevent and recover from fake clicks?
Work in four steps, in this order, plus two related checks. Each one has a companion guide.
- Check: compare invalid-click columns with post-click behavior and lead quality, looking for patterns rather than single clicks. Start with the 12 warning signs of click fraud in Google Ads, then learn how to check invalid clicks in Google Ads.
- Respond: contain the source with exclusions and setting changes, and document each action with a date. Follow how to stop click fraud in Google Ads, step by step.
- Prevent: tighten targeting, placements and conversion signals so repeat sources stop reaching your ads. See how to prevent click fraud in Google Ads, done without blocking real customers.
- Recover: read what Google already credited, then file for what it missed inside the 60-day window. Read about Google Ads invalid-click credits and how to dispute invalid clicks.
- Find the waste: some loss is not fraud at all. Compare why your Google Ads budget is being wasted, CPA inflation from bad traffic, and a weekly traffic quality review.
- Plan for peak season: a dated Black Friday and Cyber Monday checklist.
Which campaign types carry the most risk?
Every campaign type can waste budget, but the risk pattern changes with inventory and intent.
| Campaign type | Primary fraud or quality risk | Best control |
|---|---|---|
| Search | Repeated paid clicks on expensive terms or weak query matching | Search term review, invalid-click monitoring, lead-quality validation |
| Shopping | Product clicks from non-buyers or repeated comparison traffic | Product segmentation, query review, location checks |
| Display | Low-quality placements, accidental clicks, and weak engagement | Placement exclusions and engagement review |
| Performance Max | Broad inventory and limited source visibility | Qualified conversion imports, value rules, placement review where available |
| Demand Gen | Low-attention traffic and broad audience expansion | Audience quality checks and post-click behavior review |
| Video | View or click activity that does not support pipeline | Placement, audience, and assisted-conversion analysis |
| Lead form assets | Fake or low-quality submissions | Phone, email, duplicate, and sales-acceptance validation |
This map helps avoid broad conclusions. If Search is clean but broad inventory is weak, the answer is not "pause Google Ads." It is to isolate the weak source and protect the campaign types that still produce qualified demand.
Which warning signs should PPC teams monitor?
The strongest fraud signal is a mismatch between paid activity and real commercial outcomes, not any single metric.
- Clicks rise while qualified leads, sales calls, or pipeline stay flat
- High-CPC keywords spend heavily but produce no valid sales conversations
- The same location, device type, network, or source repeats without conversion quality
- Sessions from paid traffic are unusually short or show no interaction
- Google Ads reports conversions that the CRM marks as invalid, duplicate, or spam
- Display, PMax, or Demand Gen placements spend but produce weak engagement
- Invalid clicks or invalid-click rate rise at the same time lead quality falls
- Budget is exhausted before your historically strongest conversion windows
If the pattern is broad and random, fix campaign structure first. If it repeats around specific sources, devices, networks, locations, placements, or leads, investigate fraud.
Rule out setup mistakes before calling it fraud
Some accounts look badly hit because the setup is loose. Clean these up before making strong claims:
- All lead form submissions count as primary conversions even when sales rejects them
- Broad match runs without enough negative keyword review
- Brand, generic, research, and competitor-intent traffic are mixed together
- Locations include people interested in a place rather than only people in the market area
- Display or PMax spend is judged by raw conversions instead of qualified outcomes
- Tracking issues create click and session mismatches that look suspicious
- Landing pages do not match the ad promise, causing legitimate users to bounce
Fixing these does not replace fraud protection. It gives you a cleaner baseline so suspicious patterns stand out.
Run the detection checklist
Use this workflow before adding exclusions or changing bidding.
1. Review Google Ads columns
Add invalid clicks, invalid click rate, clicks, cost, conversions and cost per conversion, then segment by search term, location, device and hour of day. Google's invalid-click data shows what Google has already identified. Compare it with your own post-click and lead-quality data.
2. Check analytics behavior
Look for paid traffic with:
- very short sessions
- no scroll or click behavior
- repeated landing-page visits
- high bounce from expensive terms
- location or device patterns that do not match buyers
Clicks and sessions differ for normal reasons, including tracking setup and user behavior, and Google notes that Analytics may report data differently from Google Ads. Use the mismatch as a clue, not as proof.
3. Validate lead quality
Compare suspicious campaigns against CRM outcomes: valid phone rate, valid email rate, duplicate lead rate, booked appointment rate, sales accepted lead rate, and qualified opportunity rate. A campaign that looks efficient in Google Ads can still be bad if the CRM rejects the leads.
Build an evidence package
If you need to escalate internally, review with an agency, or file an investigation request with Google, the evidence should be specific enough for someone else to see the pattern.
- Campaign, ad group, keyword, placement, or asset group affected
- Date range and time-of-day pattern, inside Google's 60-day window
- Cost, clicks, invalid clicks, conversions, and qualified outcomes
- Locations, devices, networks, or sources that repeat
- Landing-page behavior such as session duration or engagement
- CRM outcomes such as invalid, duplicate, spam, rejected, or sales accepted
- Screenshots or exports from Google Ads, analytics, CRM, or monitoring tools
- The action you took and the date you took it
If you have Google Ads API, Google Ads scripts or BigQuery access, add the click-level rows too. Our guide to the Google Ads click view report shows how to match each GCLID to your server logs, GA4 export and CRM.
A good evidence package does not need to prove intent. It needs to show that a pattern is wasting budget or polluting conversion quality. Keep it after the case closes, because low-quality sources often return when budgets, match types, or campaign types change.
Follow a 30-day prevention plan
Use this plan when an account has suspected fraud but no mature review process.
Week 1: clean the signal
- confirm auto-tagging and conversion tracking
- separate raw leads from qualified outcomes
- add invalid-click columns
- export search terms, placements, locations, devices, and hours
Week 2: remove obvious waste
- add negatives for clear non-buyer intent
- exclude placements with spend and no qualified outcomes
- review location settings and unsupported markets
- document all exclusions and reasons
Week 3: validate patterns
- compare suspicious traffic against CRM outcomes
- look for repeated devices, networks, locations, or placements
- check whether campaign changes explain traffic spikes
Week 4: decide what needs automation
- keep manual review for low-risk patterns
- automate monitoring where spend, CPC, or fake leads are material
- import better conversion stages if bidding depends on lead quality
The plan is staged on purpose. If you block first and investigate later, you may remove real customers. If your busiest weeks fall in Q4, the Black Friday checklist above puts these steps on a dated calendar from October setup to the January credit deadline.
Each layer of the plan maps to one control:
| Layer | What to do | Why it helps |
|---|---|---|
| Tracking | Use auto-tagging, clean conversion goals, and CRM-qualified lead imports | Prevents bad traffic from training bidding systems |
| Campaign hygiene | Review search terms, match types, locations, devices, and schedules | Removes ordinary waste before calling it fraud |
| Inventory controls | Review placements, apps, partner traffic, and PMax source quality | Limits low-quality reach in broad campaign types |
| Exclusions | Block IPs, locations, placements, or sources only when evidence supports it | Reduces waste without blocking real prospects |
| Lead validation | Separate raw leads from qualified leads and spam outcomes | Protects Smart Bidding and sales-team time |
| Monitoring | Use click-level monitoring when manual review cannot keep up | Finds repeat suspicious traffic faster |
When should you use click fraud protection software?
Use it when manual review cannot keep pace with the spend at risk. Manual review is enough for some small, low-risk accounts. It is not enough when:
- CPC is high
- daily budget can be wasted quickly
- lead quality matters more than raw form volume
- campaigns run across broad inventory
- suspicious sources rotate across devices or networks
- an agency needs consistent reporting across many accounts
Google's credits arrive after the fact and without click-level detail. ClickFortify scores every paid click in real time on 100+ device, network and behavior data points, adds the IPs it flags to your Google Ads IP exclusions automatically, and logs the reasons for each flag. Start with click fraud protection software for Google Ads. For Search campaigns specifically, see Google Ads click fraud protection.
What should you report to leadership?
Executives need to know whether spend is reaching qualified demand, not every click detail. Report spend at risk, qualified and rejected lead rates, the top waste sources, invalid clicks and credits from Google, the actions taken, and the cost per qualified lead after cleanup. That moves the conversation away from "are there bots?" and toward "is the account buying real opportunities?"
What should you do next?
Fake clicks on Google Ads are best handled as traffic-quality control. Google's filtering and credits matter, but they only cover what Google classifies as invalid, and investigations stop at 60 days. Your account still needs its own review across clicks, placements, lead quality, and conversion signals: clean tracking, strong campaign hygiene, lead validation, evidence-backed exclusions, and monitoring that catches repeated waste before it becomes normal performance.
