Preparing for the Digital Services Act: a survey of internet-based services and platforms in Norway

30 October 2023 | Regulation and policy

Harald Wium Lie

Article | PDF (3 pages)

"Identifying intermediary services according to the DSA definitions is not easy; specific difficulties arise for online platforms and in particular in distinguishing those where the intermediary service is ancillary to the service provided."


In light of the new EU Digital Services Act (DSA), Analysys Mason conducted a survey of internet-based services and platforms in Norway on behalf of the Norwegian Communications Authority (Nkom).

Online platforms can have a significant influence on users and their decision-making patterns. Some online platforms with large economies of scale and scope offer mass-market services that are chiefly provided by one or more of the global ‘hyperscalers’ (for example, Alphabet, Amazon, Apple, ebay, Meta, Microsoft, X etc). Partly as a response to this, the EU has adopted the DSA and the Digital Markets Act (DMA). The DSA will be applicable to all digital services in the EU from 1 January 2024. Through the European Economic Agreement (EEA), Norway is included in most EU legislation and will be subject to the DSA. 


The DSA is a targeted set of mandatory rules that apply to providers of intermediary services with a substantial connection to the EU. It includes different provisions for different categories of intermediary services. The purpose of the DSA is to give better protection to internet users, establish transparency and accountability, and provide a single, uniform framework across the EEA, creating rules for a “safe, predictable and trusted online environment that facilitates innovation and in which fundamental rights […] are effectively protected.” 

Types of intermediary services in the DSA

According to the DSA, an intermediary service is a service that is provided remotely, via electronic means and at the specific request of one user (as opposed to being disseminated to many users simultaneously and without prior request). The DSA divides intermediary services into three broad categories: mere conduit, caching and hosting services. Under hosting, the DSA also distinguishes between (non-disseminating) hosting services, online platforms and very large online platforms (VLOPs).

  • Mere conduit services are intermediary services that transmit or deliver information to recipients of the service without any alteration or modification of the transmitted content and without involvement in the communication between the senders and receivers (for example, internet service providers (ISPs)).
  • Caching services are intermediary services that transmit or deliver information to recipients of the service while engaging in automatic, intermediary and temporary storage of that information for efficiency-related purposes (for example, content delivery networks (CDNs)).
  • Hosting services are intermediary services that store information provided by, and at the request of, a recipient of the service. Online platforms and VLOPs are two additional types of hosting services to which additional obligations apply. Examples of hosting services include web hosting services, cloud computing services and email services.
  • Online platforms are hosting services that, at the request of a recipient of the service, store and disseminate information to the public (for example, online search engines, online marketplaces, app stores, social media platforms and collaborative economy platforms).
  • VLOPs, which include ‘very large online search engines’ (VLOSEs), are online platforms that have more than 45 million average monthly active service users in the EU or more than 10% of the population. Examples include Google Maps, Google Search, Snapchat, TikTok and YouTube.

Method used in Analysys Mason’s survey

We found no single data source that could identify all relevant companies, and we do not believe that such a source exists. Instead, we based our Norwegian survey for Nkom on 21 data sources to identify DSA subjects. These data sources were used to investigate the most popular websites and applications. In addition, we used a manual, bottom-up approach combined with several expert interviews. However, surveying the market using this approach does not ensure a collectively exhaustive list of services that are subject to the DSA. Moreover, the final database contained many duplicate values for services identified through different sources. For instance, the online marketplace was identified through seven different sources.

We created a database of 1565 internet-based services that target Norwegian users. These services were mapped into 34 service classes and five DSA categories. For each service, we also collected data on 12 other attributes such as company registration number and the size of the active user base.

When assigning a DSA category to each service identified, we used the main feature of that service. For instance, we only assigned one category to Facebook, namely VLOP.

Findings from the survey

Of the 1565 services we considered, the largest DSA category by number is mere conduit with 423 services, followed by online platform with 250 services, and hosting with 213 services. Also, we found 587 services that we do not believe are subject to DSA. 

The classification of intermediary services according to the DSA definitions was not a trivial task, for several reasons. Firstly, we identified a large number of services as non-intermediary services partly because we used a large number of data sources, which contained many services outside the scope of the DSA. Secondly, some service categories were more difficult to identify than others. While mere conduit and caching services were relatively straightforward, but the task of collecting data on intermediary services proved harder for hosting services. Specific category-specific comments are as follows.

  • We found mere conduits to consist of one large group of services, namely those from ISPs, on which data was readily available from Nkom.
  • Likewise, we found the largest group of caching services to be content delivery network (CDNs).
  • Hosting services were more heterogeneous: the hosting and online platform service categories were not dominated by a particular type of service. As such, mapping out the hosting services category was done using a significantly larger number of sources.
  • The DSA highlights in Recital 29 that an intermediary service can be provided in isolation, or as a part of another type of intermediary service or simultaneously with other intermediary services. In other words, one part of an internet-based service’s offering may be classified as an intermediary service and thus be affected by the DSA while another part of that same service may not be. For instance, private groups on Facebook, which require an invitation to join, may not be subject to the online platform obligations that other parts of Facebook are, because the “dissemination to the public” requirement probably does not hold in these closed groups. Such considerations further complicated the survey work.
  •  Finally, online platform services that are ancillary to the main service may not merit an online platform designation. For instance, comments sections on newspaper sites are ancillary to the primary service (publication of news), and may thus not merit an online platform designation. We interpret this to also apply to online stores with user-written reviews under their products. However, a question then arises as to when an ancillary service stops being ancillary. If reviews sections are a critical part of an online store’s sales strategy, such sections may merit an online platform designation after all.

Although we surveyed more than 1500 services, we may not have discovered all applicable services in Norway. This is especially true for online platform services, which often have no physical (infrastructural) footprint in a country and are as such less bound by national borders. In any case, this leading-edge survey establishes a valid starting point for further study of the digital services market.

To read the full report, see Nkom’s web pages: If you have any questions about this project, or are interested in conducting a similar exercise in another Member State, contact Analysys Mason’s project manager, Harald Wium Lie.

Article (PDF)