In brief
To isolate which network is sending bad traffic, you need to break the account down by traffic source, compare each network against real business outcomes, and look for the point where click quality starts to collapse.
The mistake is looking only at top-level campaign results. If Search, Search Partners, Display, Performance Max, YouTube, Discover, Gmail, and remarketing-like traffic are all mixed together, the account may show one average performance number that hides the real source of the problem.
Bad traffic usually leaves clues. It may show up as short sessions, fake forms, irrelevant locations, low-quality calls, repeated clicks, high spend with no pipeline, or conversions that never become qualified leads. The goal is to trace those clues back to the network or campaign type that produced them.
You are not trying to prove every bad click is fraud. You are trying to find which traffic source is creating activity without business value.
A cleaner way to trace the source of junk traffic
The first step is to stop looking at the account as one blended traffic pool. A Google Ads account can include several very different traffic sources. Core Search behaves differently from Search Partners. Display behaves differently from Search. Performance Max may blend multiple surfaces. Remarketing traffic behaves differently from cold prospecting. If all of those are judged together, the diagnosis becomes muddy.
Start by identifying the symptom. Are you seeing fake lead forms? Out-of-area clicks? Repeated visitors? Spam calls? A sudden drop in lead quality? A rise in conversions that sales teams reject? The symptom tells you where to look first.
If the problem is fake form submissions, compare form quality by campaign and network. If the issue is foreign clicks, compare location data by campaign type. If the issue is repeated clicks, review click patterns by area, device, and source. If the issue is low-quality conversions, compare platform conversions with CRM-qualified outcomes.
Timing is also important. Bad traffic often appears after a change. Maybe Display partners were enabled. Maybe Search Partners were turned on. Maybe PMax was launched. Maybe a campaign moved to broad match. Maybe a new conversion action was added. Maybe the budget increased, and automation expanded into cheaper traffic. The change does not always prove the cause, but it gives you a starting point.
Next, compare traffic quality across networks using more than media metrics. CPC, CTR, and cost per conversion are not enough. A network can produce cheap conversions and still damage the account if those conversions are fake, unreachable, or irrelevant.
Use business-quality indicators instead. For lead generation, compare valid phone numbers, reachable leads, booked calls, qualified prospects, sales opportunities, and closed deals. For eCommerce, compare add-to-cart quality, checkout behavior, purchase rate, refund patterns, and revenue per session. For larger B2B advertisers, compare demo quality, company fit, job title, lead source, and sales acceptance.
Then review post-click behavior. Bad network traffic often has weak engagement. Visitors may bounce quickly, avoid important pages, submit forms too fast, repeat the same pattern, or trigger soft events without meaningful intent. One weak session means little. A network that repeatedly sends shallow sessions deserves attention.
Location data can help as well. If one network sends a high share of clicks from outside the target market, that is a signal. The cause may be targeting settings, user interest, VPNs, proxy traffic, partner traffic, or suspicious behavior. The point is not to assume. The point is to isolate.
If the account allows testing, run controlled comparisons. Disable a suspect network in one campaign or for a defined test period. Watch what changes. Does raw lead volume drop but qualified lead rate improve? Do spam forms fall? Does the CRM become cleaner? Does cost per valid opportunity improve? That tells you more than top-level conversion volume.
You should also avoid feeding bad traffic into automated bidding. If a poor network is generating raw form submissions, and those forms are primary conversions, the campaign may learn to chase more low-quality users. That makes isolation urgent. Bad traffic is not only wasting money. It can train the account in the wrong direction.
The cleanest diagnosis usually comes from combining Google Ads, analytics, CRM data, and click-level protection. The ad platform shows where the spend and clicks came from. Analytics shows behavior. The CRM shows quality. Click protection helps identify suspicious patterns that standard reports may not expose clearly. This connects directly to the broader process of diagnosing bot traffic and fake leads in Google Ads campaigns.
Real-life example
A national healthcare provider runs Search, PMax, and Display campaigns for private clinic appointments. The dashboard shows rising conversions, but the call center reports more fake numbers, irrelevant locations, and users who never answer.
Instead of pausing everything, the team separates the data by campaign type and network. Core Search produces fewer leads but a higher booking rate. Display produces many cheap forms, but most are invalid. PMax sits in the middle: some real appointments, but also a growing number of weak submissions.
The team then checks timing. The spike began after Display budgets increased and PMax started using a broader conversion action. Display is excluded from the worst placements, and PMax is shifted away from raw form fills toward qualified appointment data.
Lead volume drops, but booked appointments improve. The issue was not “Google Ads traffic” as a whole. It was specific networks and signals producing noise. For advertisers that need better visibility across paid sources, paid marketing protection can help identify the traffic sources that create spend without real outcomes.
Bottom line
To isolate which network is sending bad traffic, do not rely on account-level averages. Break performance down by campaign type, network, location, behavior, and CRM quality.
The winning source is not the one with the cheapest clicks. It is the one that creates valid leads, real prospects, purchases, appointments, or pipeline.
Once you identify the source of junk traffic, you can restrict it, exclude it, protect against it, or stop feeding it into automated bidding.