Start with the first checkpoint you cannot verify. An ad interaction, a page request, an analytics event, an accepted form, a qualified lead and a billing adjustment are different records. Pick the symptom below, collect the named evidence and run one check before changing a campaign.

This hub covers paid-traffic delivery, measurement, lead quality, suspicious patterns and protection decisions. Each route links to a complete diagnostic with its own sources, examples or worksheet. Related free tools inspect a specific part of the evidence; they do not certify every other checkpoint.

Keep the first pass small

Choose one campaign or cohort, one completed period and one business outcome. Record timezones, filters and processing delays. Use controlled synthetic tests rather than repeatedly clicking your live ad. Preserve missing or sampled evidence as unknown.

Choose the symptom you can describe

Follow the main diagnostic first. Use the related free tool only when its input matches the evidence you have. For example, a click total without a client observation supplies no IP address to exclude, and an address list cannot reconcile a billing credit.

Visits and measurement

Start here when the expected page or measurement record is missing. Locate the first failed checkpoint before changing targeting or buying protection.

GA4 traffic looks automated or strangely engaged

Next check: Review collection and engagement definitions alongside real outcomes. A short or engaged session alone is not a fraud verdict.

Bring: Campaign/page cohort, key-event definitions, collection gaps and qualified outcomes.

Bot traffic in GA4: how to spot it, and what to do about it →

Leads and conversion records

Start here when an event exists but the business outcome or destination report is wrong. Keep server acceptance, qualification and attribution separate.

Campaign and traffic patterns

Start here when the issue concentrates in a channel, network, address or location. Compare useful outcomes before attributing intent to a label.

A shared or hosting-labeled IP may be iCloud Private Relay

Next check: Verify the client field and relay range, then compare actual customer outcomes.

Bring: Trusted address, dated range snapshot, request and rule IDs, and accepted or rejected outcomes.

Is iCloud Private Relay traffic being mistaken for bots? →

Exclusions, billing and buying decisions

Start here when you are deciding what to change or pay for. Separate a proposed list, an applied destination action and actual billing evidence.

I need to compare Google native filtering with added protection

Next check: Define the missing control and inspect its actual scope, evidence, action delivery and legitimate-traffic handling.

Bring: Native findings, site-side decisions, supported actions and false-positive review.

Does Google protect against click fraud? Compare the controls →

I cannot tell whether protection pays for itself

Next check: Compare verified costs, contribution and review time without calling flagged requests saved spend.

Bring: Actual ad costs/credits, qualified/paid outcomes, full service costs and comparison limits.

Is click-fraud protection worth it? Test the value →

Review conversion configuration without changing the account

If the symptom concerns what Google Ads treats as a conversion, start with the example below or paste the supported action export. Select the intended business outcome, then compare the findings with the actual goal and action settings. This browser tool cannot observe live tags, decide whether a lead is qualified or apply bidding changes.

Runs in this browser. Your action names and settings stay here.

Use the read-only export script below. Keep the header row. CSV and tab-separated exports with the same headers are accepted.

Check campaign goals to confirm bidding eligibility. Custom goals can include secondary actions. Match transaction or lead IDs to verify repeated events.

The read-only conversion-action export script needs your own authorized Google Ads access. For PMax placement evidence, use the separate read-only PMax placement audit script. A placement export is not a visitor ledger or a conversion-action export. Follow each script’s setup and scope notes.

For another input, use the landing URL checker, IP exclusion-list builder or placement exclusion builder. Proposed outputs still need review, supported destination scope and verification of any change you later apply.

Leave with one investigation plan

  1. Write the chosen symptom and the first unreliable checkpoint.
  2. Save the scope, source records, collection gaps and alternative explanations.
  3. Choose one controlled test or read-only report and define the evidence that would count as success.
  4. Record the result before changing targeting, exclusions, conversion goals or service subscriptions.
  5. Name an owner, next review date and rollback condition for any proposed change.

Download the investigation-plan worksheet.

Inspect the investigation plan
PAID-TRAFFIC TROUBLESHOOTING PLAN

Chosen symptom:
Guide URL:

FIRST MISSING OR UNRELIABLE CHECKPOINT
Ad interaction / website delivery / measurement / server acceptance /
qualification / destination action / billing / other:

SCOPE
Account/campaign, network, device, hostname/page:
Completed period, timezones, export time and processing indicators:
Business outcome and actual source record:

OBSERVATION
What is directly observed, in which source?
What is missing, filtered, sampled or otherwise unresolved?
Which alternative explanations remain?

NEXT CONTROLLED CHECK
One reproducible test or read-only report:
Required data and source permissions:
Synthetic test labels (avoid repeated live-ad clicks):
Success criterion and evidence to retain:

DECISION
No change / repair collection / review targeting / validate application /
review exclusions / investigate billing / evaluate protection:
Owner, review date and rollback condition for any proposed change:

FOLLOW-UP
Result and qualified/paid outcome, if observed:
Actual applied action and billing evidence, kept separate:
Next question if the checkpoint still cannot be verified:

Free browser tools review pasted inputs; they do not apply account changes.
Account export scripts need your own authorized access and are read-only.
No single traffic label, address or counter proves fraud, a charge or a refund.

https://adprotektor.app/blog/google-ads-traffic-troubleshooting/

Where the question is identity or enforcement, the IP/browser-identity guide separates the signals and their limits. The local browser-signal lab tests browser capability. The tracking/protection comparison separates event delivery from enforcement and ad-platform actions.

Move from an observation to a supported action

AdProtektor adds site-side visit evidence and configurable protection once the collection path is working. Check actual decision reasons and supported destination status alongside legitimate business outcomes. Traffic labels, request counts and proposed lists do not determine native advertising billing.

Start the free traffic audit when you are ready to inspect your own site’s visits. Keep the finding, the applied action and the business outcome separate when evaluating what changed.