Put the pattern in context.
Start with what changed. Was there a budget increase, a new placement, a different audience, a website release, or a change to tracking? Compare similar time periods and look at the customer journey before drawing a conclusion.
A high bounce rate or poor lead quality does not prove fraud. Irrelevant targeting, an unclear offer, slow pages, and duplicated conversion events can produce disappointing results too.
Review several signals together.
Look for patterns that repeat across available data sources. One unusual metric is a reason to investigate, rather than a verdict.
- Changes in traffic by placement, referral, location, device, or time.
- Repeated submissions with similar content or inconsistent contact details.
- A gap between reported conversions and independently verified enquiries.
- Abrupt changes in engagement that coincide with a new traffic source.
Validate the basics before adding controls.
Test your conversion journey and inspect the event configuration. Confirm that a submission is counted at the right point and that the same action is not counted twice.
Review suspicious enquiries with the team receiving them. Preserve the evidence you need for an investigation while limiting access to personal information and following an appropriate retention policy.
Take a proportionate next step.
Use the controls your platforms make available, such as placement reviews, exclusions, frequency settings, and appropriate validation. Consider whether additional monitoring would answer a clear question.
Advertising platforms may already identify and filter certain invalid interactions. Review their reporting and support options before assuming every unusual click has been charged or that an external tool can eliminate all risk.
Make quality a regular conversation.
Keep a short record of the pattern, evidence, action, and result. Review campaign quality alongside performance, rather than waiting until a problem becomes expensive.
The aim is a more trustworthy view of what is happening, with practical controls that protect the customer experience as well as the business.
Suspicious traffic needs investigation. Label findings carefully: confirmed tracking issue, poor audience fit, unexplained anomaly, or evidence of invalid activity.
