INFOnline Measurement System Sketch¶
INFOnline Measurement is essentially based on the fact that individual accesses to customers' web pages or within their applications (apps) can be
through technical measures.
The data collected by the measurement is aggregated depending on the order and regularly forwarded to IVW and MMC/agma (if participation consists), who then check and publish it.
Due to the data types and data volume, the collected data do not allow a user to be clearly identified as a person. While pseudonymous measurement procedures still work on the basis of client identifiers, anonymous procedures work completely without such identifiers. The pseudonymous INFOnline Measurement is designed as a pseudonymous system (with client identifiers) and the Census Measurement is designed as an anonymous system (without identifiers).
The so-called INFOnline Measurement Manager has an interface for a central technical test and control logic with which, on the basis of the legal and industrial IAB specifications (e.g. through the use of a TCF 2.0-compliant CMP on the website or in the app of the client), the required measurement sensors for anonymous and/or pseudonymous data processing are automatically integrated and executed in the INFOnline Measurement depending on the degree of a user's consent (=visitor to the web page or application) to one or more processing purposes in order to determine the standard key figures customary in the market.
Anonymous measurement method¶
INFOnline Measurement: Census Measurement¶
(only page impression; no movement data)
In the site (WEB), the client implements the retrieval of the web sensor, which is executed in the user's local browser. The web sensor calls up the service platform managed by the site, which in turn acts as an interface to INFOnline's central server systems. The data set generated by this is a pure PI data collection and is transmitted to the service platform.
No personal data occurs in the INFOnline data centre or the integrated cloud services during the census measurement. The only personal data is the IP address, which is required for communication purposes with the client. In the census measurement, no personal information is processed at all, including in particular the IP address. This is completely removed from communication and processing. There is also no geolocation by means of IP address.
Pseudonymous measurement method¶
INFOnline Measurement: pseudonymous measurement¶
(Page impression: Page Impression; cross-offer movement data in different aggregation levels: Visit, (technical) client, Qualified Client etc.).
The system for the online survey is not operated by INFOnline; there is merely a forwarding to this system.
If the use of an application (app) is to be measured, a software library is integrated by the provider of the application, which INFOnline makes available. The library ensures that all required information is transmitted to the measuring system in the correct format.
The counting pulses transmitted to INFOnline are first temporarily stored as raw data and the IP addresses are anonymized. Anonymization by shortening the IP addresses takes place as soon as possible (for IPv4 shortening by 1 byte). Other unique identifiers of end devices are transmitted exclusively as a hash.
INFOnline offers the option of Opt-Out (https://optout.ioam.de) from the measurement.
Activating the Opt-out in a web browser causes the browser's counting pulses to be discarded in the measuring system. In this case, no further analysis or measurement of the counting pulses of this browser takes place. TCF 2.x also provides an opt-out for the user. The implementation of the opt-out in apps is the responsibility of the respective provider of the app. INFOnline describes how the opt-out can be implemented in apps in the respective Integration Guide of the various platforms.
Examples of the data collected:
Time stamp, IP address (shortened), cookie content, web page accessed, signature of the browser, ID of the end device as hash (for smartphones), event triggered (e.g. "start", "stop") etc..
The data is analyzed and stored in the downstream systems.
The aggregated and evaluated measurement data is made available to IVW e.V. and MMC/agma (if participation consists) via defined interfaces. A standardized interface available on the internet is used for this purpose. The transmission of the data is always encrypted.
Overview of data categories and standard market key figures in INFOnline Measurement¶
Depending on the measurement method and device used, the measured data is recorded with the following data categories and key figures for the respective site:
|Desktop, MEW and CTV||Apps (Android/iOS)|
|base data||Page Impressions, Codes, URLs||Page Impressions, Codes, URLs|
|Advanced data(*)||Page Impressions, Visits, Clients, Qualified Client, Codes, Useragents, Usetime||Page Impressions, Visits, Clients, Qualified Client, Codes, Events, Usetime|
(*) from data collection requiring consent (consent of the user (= visitor to the site) required / assumed)
Explanation Data categories and standard market performance indicators:¶
|Page Impression||A single usage process, e.g. calling a web page.|
|Visit||A Visit describes a coherent usage process. Visits are calculated on the basis of page impressions.|
|Client||The parameter "Clients" is purely a technical parameter and indicates which devices or browsers have used the site.|
|Qualified Client||The Qualified Client is an end device used by at least one person (e.g., stationary computer, tablet, smartphone, hbbTV, etc.) from which digital sites are accessed via a browser or app. Qualified Clients are calculated on the basis of clients.|
|Events||Events/actions transmitted by an app, e.g. "view-appeared" when opening a news article in an app.|
|Useragents||Browsers / devices used|
|Urls||Shortened URL of the web page accessed by the user.|