FraudSensor FAQ

Below you’ll find a list of frequently asked questions (and answers) about FraudSensor. If you have a question that isn’t answered here, please reach out to a HUMAN representative for assistance. We’re always happy to help!

What percentage of my traffic should I tag?

Before you start serving detection tags alongside your ad impressions, you should work with a HUMAN representative to calculate the correct FraudSensor sampling rate for your unique needs. This rate tends to be inversely proportional to the average number of impressions you serve per day; for example, if you serve an extremely high volume of impressions, we may recommend only tagging a small percentage of your traffic.

Unless specifically directed to do so, you should not tag 100% of the impressions you serve.

Which type detection tag collects the most/strongest signals?

JavaScript tags collect the most signals of any FraudSensor tag type, but they aren't necessarily the best tag type for all situations, since these tags only work in environments that support JavaScript. To learn more about the different types of detection tags and their respective strengths and weaknesses, see our Tag Types and Environments guide.

Can I deploy multiple detection tags on a single page?

Yes! If multiple detection tags are deployed on the same webpage (for example, if you serve multiple simultaneous impressions to the same user), FraudSensor will recognize that these tags were deployed at the same time and group their data together accordingly. This combined data makes FraudSensor's analysis even stronger and helps us provide an IVT decision that's consistent across each grouped impression.

Can I use the same type of detection tag across all environments?

We don't recommend taking a "one-size-fits-all" approach to detection. FraudSensor works best when each ad impression is served alongside the tag type best suited for that ad's environment.

Although you can technically use pixel tags in almost any environment, JavaScript tags yield better signals and should be used whenever possible. Conversely, not all environments support JavaScript tags, which means that pixel tags are the only way to reliably collect signals for certain impressions.

How does FraudSensor handle inline frames?

FraudSensor's detection tags function inside of iframes without issue. If an ad is rendered inside an iframe, that ad impression's detection tag must be deployed within the same iframe.

Sandboxed iframes require an additional specific token. If you're running ads within a sandboxed iframe, please reach out to a HUMAN representative for assistance.

What should I do if FraudSensor is mistakenly flagging traffic as invalid?

If you think that FraudSensor is making inaccurate IVT decisions, please reach out to your Technical Account Manager and request a false positive check. False positive checks can help determine whether valid impressions are mistakenly being flagged as invalid.

After you’ve requested this review, your Technical Account Manager will work with HUMAN’s Detection team to gather more information and verify the accuracy of your results. Your Technical Account Manager will then help you take further steps as needed.