No. 469: Cloud services and AI-as-a-Service – Current and potential application scenarios and market developments
(full version only available in German)
Summary
Cloud computing, i.e. the outsourcing of storage and computing capacities as well as the maintenance of operating systems and software to cloud providers has already achieved a high level of market penetration with significant growth rates. Cloud services combined with artificial intelligence (AI) applications create new potentials, both for the providers of these services and for users who need a low-threshold access to AI.
The aim of the study is to determine the current market situation of these applications (AI-as-a-Service (AIaaS)), to identify business strategies of major providers and to derive the key factors which will influence further market developments.
Analysis shows that their ecosystems and their financial resources are important determinants of market success of AIaaS providers. Ecosystems enable providers to bundle comprehensive services and platforms and thus they increase the barriers for users to switch to other providers (lock-in effects). The necessary financial resources are mostly provided by other or neighbouring business areas (Infrastructure-as-a-Service, Platform-as-a-Service, Software-as-a-Service, together referred to as Anything-as-a-Service (XaaS)). This allows major providers to transfer their initial advantages from other XaaS segments to the AIaaS segment and to continually expand their own services through investments in research and development or acquisitions of smaller providers. Hence they keep their own ecosystem closed and ensure and even increase the attractiveness of their service offerings.
The four major cloud providers, on which this study focuses, follow their individual strategies, ranging from pioneering (Amazon Web Service) and price leadership (Microsoft Azure) to commercialising innovative consumer applications in the AIaaS segment (Google Cloud Platform) and a high-priced full-service offering (IBM Watson). Most fully managed AIaaS-applications from these providers can be assigned to image recognition, speech to text, text to speech, translation, text analysis and data analytics.
These providers are disciplined by the competition between themselves and the econ-omies of scale and economies of scope. This implies that they can only sustain in the long run by achieving a sufficient market share through creating a high customer satisfaction. In addition, there are still many niches in the AIaaS market to be served. Those hold a high growth potential, e. g. for company-specific solutions and there is a potential for further market entries.
As a conclusion, the XaaS market fulfils an innovation function and is sufficiently flexible to adapt to new requirements. The competition between the major cloud providers leads to increased struggle for market share in the short term. In the long term due to economies of scale though, there is a risk that suppliers with high market shares will pull ahead from the smaller suppliers.
Discussion paper is available for download.
