Tag Types and Environments

FraudSensor uses detection tags to collect post-bid signals and analyze impression quality. Rather than taking a one-size-fits-all approach to detection, we tailor our signal collection to provide coverage in a variety of scenarios. Each tag type offers different strengths and weaknesses, including the number and quality of signals collected, which ad formats are supported, and the environments where the tag can be deployed.

Tag types

There are two main types of detection tags: JavaScript-supported detection tags and non-JavaScript pixel tags.

JavaScript tags

JavaScript tags are FraudSensor’s primary tag type. In order to function properly, these tags must be deployed in a JavaScript-enabled environment.

JavaScript tags include the following components:

• URL: The URL of the server that delivers the tag.
• analytics.js: A script that loads FraudSensor's main payload.
• Parameter list: The tag parameters built into the body of the tag, included as key=value pairs separated by an ampersand (&) character. These parameters should be populated by your ad server.
• async: A Boolean that dictates whether the tag should load asynchronously from other page content. If its value is set to true, loading the FraudSensor script won't slow down overall pageload times.

Pixel tags

Pixel tags offer a “slimmed-down” version of FraudSensor. Pixel tags collect basic information using a server-side GET request and the parameters passed through the tag. Although these tags offer limited functionality, they're useful in situations where JavaScript tags aren't supported.

Pixel tags include the following components:

• URL: The URL of the server that delivers the tag.
• analytics.gif: The actual pixel image (in GIF format).
• Parameter list: The tag parameters built into the body of the tag, included as key=value pairs separated by an ampersand (&) character. These parameters should be populated by your ad server.

---

Feature comparison

Although JavaScript tags and pixel tags both collect signals in support of FraudSensor's traffic analysis, each tag type functions differently and offers unique capabilities.

Maximum session time

• JavaScript tags: FraudSensor grants JavaScript tags a maximum of two minutes to fully collect signals for each impression. However, most JavaScript tags will finish collecting signals in far less time. (Also note that JavaScript tags load asynchronously, so the amount of time it takes to collect signals shouldn't impact overall site performance.)
• Pixel tags: Pixel tags finish collecting signals as soon as the tag is loaded.

UTM codes

• JavaScript tags: JavaScript tags automatically collect UTM data if it's included in the URL of the impression’s referring domain.
• Pixel tags: UTM data must be passed as part of the pixel tag's parameters.

Payload size

• JavaScript tags:
  • Initial: ~2 KB (the detection tag)
  • main.js: ~33 KB (the JavaScript function)
• Pixel tags: No payload.

Signal strength

• JavaScript tags: JavaScript tags collect the richest and strongest signals to power HUMAN’s analysis. We recommend using JavaScript tags whenever possible.
• Pixel tags: Pixel tags collect fewer signals than JavaScript tags. As such, we only recommend using pixel tags when JavaScript tags aren't supported.

Environments

• JavaScript tags: You can deploy JavaScript tags in desktop and mobile browsers for most ad types, including VAST 4.0 video ads. You can also deploy JavaScript tags for certain mobile app ads, like banners and rich media; however, native ads in mobile app environments aren't rendered in a browser or web view and therefore don't support JavaScript tags.
• Pixel tags: You can deploy pixel tags in any environment, including desktop browsers, mobile browsers, mobile apps, and CTV devices.

---

Environments and ad formats

To get the most out of FraudSensor, it's important to choose the right type of detection tag to deploy alongside your ad. The following table lists our recommended detection tag type for each ad format based on environment.