Understanding Impression Measurement

Mobile advertising and attribution today revolve largely around click-through conversions. The direct correlation between a click event and a subsequent conversion (like installing an app) may seem inherently stronger than the indirect correlation between an impression (like a page view or ad view) and subsequent conversion.

Although impressions and ad views may not always lead immediately to conversions, and the intrinsic value of an impression can be difficult to quantify, don’t overlook the insight and value gained from impression measurement (especially in context of the full performance advertising lifecycle).

With the multi-touch attribution in Attribution Analytics coupled with impression measurement, advertisers and their advertising partners (publishers) can see not only the clicks that immediately preceded a successful app installation, but also any impressions that played a role in the attribution lifecycle (particularly those that led up to a conversion).

By default, Attribution Analytics measures and attributes ad clicks for all of its partner integrations. This topic describes how you, as a partner or ad network, can use Attribution Analytics to measure and attribute impressions as well. If you’re already familiar with impression measurement and ready to start, you can jump straight to Enabling View-Through Attribution.

Note: If you enable view-through attribution for an advertising partner, the partner must be measuring impressions with us for view-through attribution to occur.

Before explaining how to implement impression measurement, let’s review some terminology:

Term Description
Impression      An event that occurs when a user sees an ad. For banner ads, this event equates to when the ad is displayed. For video ads, this event equates to when the first frame of the video is displayed
View See impression above.
Click-through An event that occurs when a user clicks an ad and lands in an app store, on an advertiser web site or landing page, or on a specific screen in an app.
View-through An event that occurs when a user views an ad and lands in an app store, on an advertiser web site or landing page, or on a specific screen in an app. View-through conversions occur when there is only an impression and no click. Therefore, view-through rate is calculated as the ratio of ad impressions to conversions.

The measurement and reporting of digital video impressions is governed by the IAB Broadband Video Commercial Measurement Guidelines. The measurement and reporting of the companion ad is governed by the IAB Interactive Audience Measurement and Advertising Campaign Reporting and Audit Guidelines.

Click-through vs. View-through Conversions

The classic scenario for click-through conversions involves partners notifying Attribution Analytics of only clicks and not impressions. The following flowchart shows the general process for attributing click-through conversions.

Impression_tracking_view-through_vs_click-through_-_New_Page.png
However, when partners notify Attribution Analytics of impressions in addition to clicks (requires both a click URL and an impression URL to be associated with the ad), two possible scenarios arise:

  1. A user views an ad AND clicks on it, so attribution can occur via click-through and/or view-through.
  2. A user views an ad but does not click on it, so attribution occurs via view-through only.

In Scenario 1, the last engagement by the user with the ad is a click event, so Attribution Analytics attributes installation and other events as click-through conversions. The following flowchart shows how Attribution Analytics accommodates both impression-based and click-based events.

Impression_tracking_view-through_vs_click-through_-_New_Page__1_.png
In Scenario 2, the last engagement by the user with the ad is an impression, so Attribution Analytics attributes installation and other events as view-through conversions. Since the user did not click the ad, Attribution Analytics can only attribute installation and other events back to the impression. The following flowchart shows how Attribution Analytics attributes view-through conversions (with no clicks involved).

Impression_measurement_view-through_vs_click-through_-_New_Page_2_

View-through/Impression Attribution

For view-through attribution, Attribution Analytics uses a shorter attribution window than for click-through attribution due to client/advertiser expectations around the value of impressions versus clicks (they perceive clicks to be more valuable than impressions, thus an extended attribution window for clicks).

Similar to click-through attribution, the attribution window for view-through attribution may also vary depending on which attribution method Attribution Analytics uses. The following table organizes the default attribution windows (for both click-through and view-through attribution) by attribution method:

Click-through (Click) View-through (Impression)
Identifier Matching 7 days 24hrs
Google Play Install Referrer 30 days N/A
Device Fingerprinting 1 day 1hr
Open URL w/Click ID 7 days N/A

By default, view-through attribution is disabled. Once you’ve enabled the advertising partner and that partner supports impression measurement, you can choose to enable view-through attribution.

Attribution Analytics supports custom attribution windows for view-through conversions as long as they are measured by identifier matching. Attribution Analytics does not support custom attribution windows for attribution by device fingerprint matching.

Impression URLs

The structure of impression URLs is almost identical to click URLs. The Base URL takes the format of:

https://XYZ.tlink.io/serve?

* any of the following domains can be used and all resolve to the same server deployments.

  • hastrk1.com
  • hastrk2.com
  • hastrk3.com
  • api-01.com
  • api-02.com
  • api-03.com
  • api-04.com
  • api-05.com

The following table shows the required parameters to include for impression URLs:

Parameter Description
action impression‘ to log ad view events
publisher_id Your account ID with the Attribution Analytics-defined advertiser
site_id The ID of the mobile app for your client
offer_id The campaign ID for your client in Attribution Analytics (same as campaign_id)

TUNE links for integrated partners generally include their own internal macros for device identifiers mapped to Attribution Analytics internal device identifier parameters. If you include a device identifier in the URL to measure impressions, then Attribution Analytics performs attribution by identifier matching instead of device fingerprint matching.

Consider appending the following applicable device identifier parameters to your TUNE links:

Device Identifier Description
ios_ifa The Apple advertiser identifier with iOS 6.0+. Also referred to as IFA or IDFA.
device_id The device ID of the Android device: IMEI for GSM phones, MEID for CDMA phones, and SerialID for non-cellular/WiFi-only tablets.
android_id The AndroidID for Android devices only. It is a 64-bit number (hex string) that is randomly generated on the first boot of the device.
mac_address The MAC address of the WiFi adapter on the phone.

For more information about the device parameters that you can use in your TUNE links (for impressions as well as clicks), visit unique identifiers in attribution.

TUNE links can also include one of our sub parameters to allow clients to gain more valuable insight into their advertising campaign performance. The following table shows the sub parameters that you can add to your TUNE links:

Sub Parameter Description
sub_publisher Name or ID of the publisher of the partner.
sub_site Name or ID of the partner site or mobile app.
sub_campaign Name or ID of the campaign for which you are advertising with the partner.
sub_adgroup Name or ID of the ad group for which you are advertising with the partner.
sub_ad Name or ID of an ad for which you are advertising with the partner.
sub_keyword Name or ID of the keyword specific to Google AdWords integration and other search campaigns.
sub_placement Name or ID of an ad placement (position of ad on a web site or in an app).
ref_id Internal reference ID parameter that you can use for your own ID values (such as impression or session ID). This reference ID is typically used by partners for reconciliation with their own third-party system.

Impression URLs support all of the same parameters as those defined in Effective TUNE Links.

The following URL shows an example of an impression URL with a reference ID for a specific partner.

https://XYZ.api-01.com/serve?action=impression&publisher_id=200&site_id=1000&offer_id=1500&ref_id=10000000000000000001xxxxxx1

When creating a TUNE link in Attribution Analytics, the platform will generate both a click URL and an impression URL based on your selections (as seen in the example below). Note: The platform will only generate an impression URL when the Partner you select during creation supports and is enabled for impression measurement.

No Comments

Leave a reply