Skip to main content

Available datapoints

List of datapoints​

Source​

  • Channel: Top level source, i.e. 'Ad network', 'Referral', etc.
  • Source: Lower level source, i.e. 'Google', 'example.com', 'my_utm_source', etc.
  • Campaign: Either taken from utm_campaign, or campaign name from the ad network.
  • Subtype: Additional source details, i.e. 'Instagram', 'Audience network', etc.
  • Group: The ad group name from the ad network.
  • Medium: Taken from utm_medium, or 'paid' if through an ad network.
  • Content: Taken from utm_content, or the ad asset ID from the ad network.
  • Term: Taken from utm_term, or the keyword from the ad network.
  • Click ID: The click ID from the ad network, i.e. gclid, fbclid, etc.
  • Referrer: The referrer domain name.

Page​

  • Entry page: The landing page for the session.
  • Exit page: The last page the visitor viewed in the session.
  • Hostname: The hostname of the landing page.

Location​

  • Map: The country displayed on a world heatmap (available through analytics).
  • Country
  • City

Device or browser​

  • Browser: The reported browser.
  • Browser version: The reported browser version.
  • Browser detected: The browser detected by Hitprobe's tag.
  • OS: The reported operating system.
  • OS version: The reported operating system version.
  • Device type: The reported device type, 'Computer', 'Mobile', or 'Tablet'.

IP address​

  • IP address
  • Network
  • ISP

Other​

  • Fingerprint: The unique fingerprint computed by Hitprobe.
  • UA: User agent as reported by the browser.
  • Language
  • Timezone

Risk & engagement​

  • Risk level: The risk level assessed for the visit (likelihood of the visit being suspicious or invalid).
  • Engagement level: The engagement level assessed for the visit (how much interaction with the website occurred).
  • Events: Custom events associated with this visit. Either sent via the tag or the API.
  • Risks: A list of risks identified, i.e. 'Bot detected' or 'Poor IP reputation'.
  • Actions: A list of actions taken, i.e. 'Audience exclusion'.

Risk level​

During a session, risks are identified (such as 'Suspect device', or 'High IP frequency'). Then depending on the risks and the channel, a risk level is set.

Here are the risk levels and the qualifiying criteria for each:

  • Normal risk level is the default (normal visits).
  • Unwanted level is a visit from a real person that is unlikely to be useful or relevant to you. It is set for poor referrer reputation and where the visit is from an ad network or source tagged, excluded country, or a crawler (a known search engine crawler/spider).
  • Suspicious level is set for visits from a person trying to hide their identity or true device details. It is set for anonymous visits, visits with a poor IP reputation, or from a suspect device.
  • Invalid level is for unwanted visits that may cost you in terms of resources or wasted ad spend. It is set for all bots, or for high IP/network/device frequency (repeat visits) and where the visit is from an ad network or source tagged.
tip

Even where risks don't change the risk level because the visit is not from an ad network or source tagged, the risks are still flagged and can be filtered on individually if needed.

You can visit risk settings to customize IP address, network, and fingerprint limits, or change the countries that should be excluded.

Engagement level​

During a session, the level of engagement contributes towards the 'engagement level':

  • Bounced engagement level is where the visitor left very quickly after only viewing one page.
  • Normal level is set for visitors that don't bounce, but don't engage with the site enough either.
  • Engaged level is set where the visitor stays for a minimum length of time or views a minimum number of pages. These are generally people interested in the site.
  • Converted level is set where the session receives a conversion event through either the Hitprobe tag or the API.

You can visit engagement settings to finetune the thresholds for these levels.

Risks​

Here's a summary of all the risks that Hitprobe can identify:

  • Anonymous: Visitors using a VPN or proxy, or where we can infer that they are attempting to hide their true location.
  • Bot detected: Sessions where a bot is detected, i.e. a headless browser or automated script.
  • Crawler identified: Sessions that are reported to come from a known search crawler or 'good' bot, i.e. Googlebot, etc.
  • Suspect device: Where Hitprobe detects inconsistencies in the values reported by the device (i.e. the visitor may be attempting to hide their true device information).
  • Poor referrer reputation: Where the referrer domain is found to be poor. This can be a very new domain name, an unreachable domain, or where there is no footprint for the domain on search engines or elsewhere on the internet.
  • Poor IP reputation: If the IP is found to have a poor reputation, for example if it is suspected to be part of a botnet or associated with residential proxies, this risk will be flagged.
  • Excluded country: This risk is flagged where the session comes from a country that is excluded in your risk settings.
  • High IP frequency: Flagged where the IP address exceeds the limits set in risk settings.
  • High network frequency: Flagged where the IP address network exceeds the limits set in risk settings.
  • High device frequency: Flagged where the device fingerprint exceeds the limits set in risk settings.

Actions​

In saved views, you can enable exclusion actions to be taken on new matching visits. The actions datapoint lists any actions that have been taken on the session.

  • Audience exclusion: Visitor added to exclusion audiences through connected ad networks.
  • IP exclusion: The visitor's IP address was added to list of excluded IP addresses in Google Ads.
  • Network exclusion: The visitor's IP network (ISP) was added to list of excluded networks in Google Ads.
  • Domain exclusion: The referring domain name was added to list of excluded sites in Google Ads.