Measure

Once tags have been deployed (see Adhese Setup and Implementation), they start collecting data immediately.

What does Adhese record?

Actions and parameters divide the measurement process of Adhese into two categories:

• An action is 'something that happens';
• The parameters are 'all the information that is collected together with the action.'

Actions

Adhese measures actions. Actions are behaviours that a visitor can perform, such as visiting a web page, reading an article, viewing or clicking on a banner ad, watching a video, playing a game, submitting a form, purchasing a product, etc. Each time a user performs a pre-defined action, the client sends a request to the Adhese servers.

The most basic action of the ad-serving process is impressions. This means that a specific format (such as a leaderboard) was loaded onto a particular location (for example, the homepage or an article page) of a publication (for instance, a newspaper's website). It depends on the chosen implementation method when the counting of an impression takes place. Please refer to www.github.com/adhese/sdk for a complete description of the different implementation methods.

User actions: beyond impressions and clicks

Adhese allows for tracking of user actions beyond impressions and clicks. Additionally, Adhese enables tracking of multiple actions simultaneously.

The standard syntax for an action tracking call enables the insertion of three elements:

• A campaign identifier, for example, Volvo2015
• An identifier, like "info request" or "request for demo"
• A free value, such as the value of a purchase

This tracking feature requires a custom setup of your account. Contact Support if you wish to implement this feature.

Let's look at some examples of user actions.

• Within the ad environment If Adhese serves a splash ad or interstitial, it is possible to track the visitor's behaviour. Does the visitor click outside the ad or the close button inside the ad to continue to the underlying webpage?
• Outside the ad environment Adhese can also track events that occur outside the online advertisement environment. For instance, if a user clicks on your ad, Adhese can track their visit and determine if they made a purchase as a result of your advertising. This allows for more accurate attribution models, as many events can contribute to an online purchase or lead.

Viewability

Adhese has a tracker that can measure viewability if (legacy) document.write is used (see Parameters for templates and Advar templates).

If you use JSON, your logic can be implemented to measure viewability and then report it to Adhese via the viewableImpressionCounter parameter, which is available in the JSON response (see List of JSON response structure object fields).

Parameters

A set of parameters accompanies every action or request. Parameters offer a greater insight into who is doing what and where and how these actions occur. The parameters consist of predefined data, such as screen size, browser and operating system, URL and referrer, or custom user data, such as age and geographic location.

Custom parameters can be implemented based on your specific needs and possibilities. Refer to the Request Target Parameters appendix for a list of readily available parameters. The available target parameters and their prefixes are determined by the configuration of your Adhese account. Adhese will use this information to build your inventory and provide insights to match campaigns with your audience and vice versa.

Remember that the unique identification of visitors and the storage of personal and other data are subject to national, European and international legislation. It is the exclusive responsibility of the publisher to comply with the law and adequately inform the user of his rights. Adhese can help and advise publishers here. More on user privacy is available in the GDPR section of the documentation.

Impression measurement

Impression measurement is an essential aspect of ad serving. Understanding how and when Adhese counts an impression is important. There are two scenarios:

1. An ad is requested and visualised.
2. An ad is requested and could be visualised, depending on rules like minimal screen size, device rotation, viewport, etc. The impression is counted later (i.e., deferred impression).

The second scenario results in two types of requests:

• An ad request
• A track request

An ad request is a request that returns an ad, the first scenario. The second scenario executes a tracking request after the initial ad request. Therefore, it acts as an impression request.

If you are using track requests to measure impressions, these numbers will build up your inventory. The actual inventory is reported and used in forecasting.

The above explanation about impression measurement applies to any tag (See Implementation of the Adhese ad tags).

The Campaign actions report per position reports the number of requests, tracks, and other events.

Third-party discrepancies may arise when the publisher's report does not match the advertiser's report. Adhese counts an impression the moment the ad request is made, while the advertiser may count an impression when the ad is delivered to the user. For more information about third-party discrepancies, refer to Troubleshooting.

Retention period of data

Historical summarised data, such as daily unique visitors, is available anytime and can be archived at the client's request. The data can be accessed through reporting, data mining, our API, or on-demand as a custom report.

However, there are some restrictions that you need to take into account:

• Legislation: certain information cannot be saved or must be deleted after a short period. The unique identification of users and personal and other data storage is subject to national, European and international legislation. Adhese can assist and advise publishers in the matter. However, it is the exclusive responsibility of the publisher to comply with the law and adequately inform users of their rights.
• Common sense: the amount of collected data can be huge for some (combinations of) parameters. You should consider whether storing such a large amount of data is necessary or desirable. The answer to this question will differ from client to client. Where there is a lot of data, it is often recommended that only aggregated data is maintained.

Predictions are usually based on historical data. The type of data saved, the frequency of saving (whether daily, hourly, or otherwise), and the duration of storage depend on the client's needs and strategy.