Current trends in the development of ecosystems and the forecast for their development in Russia for 2024 and the midterm

The 2023 results gave reasons to assert that the main Russian ecosystems have generally completed the processes of adaptation to new geopolitical and economic conditions, while these changes have caused no significant damage to any of them, and they have not been critically weakened relative to their competitors. In the midterm, we can expect a return to the usual development dynamics, but taking into account new opportunities and restrictions that are almost the same for all players, in particular:

• continuation of import substitution in a number of markets, such as entertainment content and video services, various software and digital services, advertising, hotel booking, after foreign players left this market,
• legal transfer (redomiciliation) of companies to Russia, as well as revitalization of the domestic IPO market, which opens opportunities to raise additional funds for both large companies and relatively small ones. This creates opportunities, on the one hand, and increases the level of competition, on the other hand.

It is especially important to note the e-commerce segment, the potential of which in Russia is far from being exhausted (penetration is several times lower than in the most advanced markets in this regard, such as, for example, China, South Korea, Great Britain), so Wildberries and Ozon, as the two largest players, may well begin expansion into new areas, that is, to start generating their own ecosystems. The example of Ozon’s fairly successful promotion of bank cards shows the potential strength of large marketplaces, while the development history of Amazon in the USA can become a role model in general. This trend towards ecosystems can be strengthened by the fact that the composition of sellers and the range of leading marketplaces will converge, the number of pick-up points and parcel lockers will most likely level off in the midterm, and the price competition will be constant. Thus, it is quite possible that none of the marketplaces will be able to achieve a critical gap in terms of business scale, which will stimulate a desire to enter new markets. Entering new markets can happen quite quickly — based on the solutions and infrastructure of partners (for example, the fintech-as-a-service model for the implementation of built-in financial services, the white label approach for services for purchasing air tickets and travel, etc.).

Key development vectors: technological and business trends

Regarding trends in various areas and business directions that are relevant for digital ecosystem companies globally, we would like to note the following points.

In the field of telecom:

• 5G continues to develop rapidly around the world and overall penetration is increasing faster than it was in previous generations of communication. Nevertheless, the introduction of this new standard has not yet provided operators with extra revenue. Thus, from a commercial point of view, the business of 5G operators is rather an evolution than a revolution,
• the eSIM technology has also not become widespread yet (compared to the expectations) and has not had an impact on the competitive situation in the mobile communication markets,
• however, disturbances are possible not in the mobile communication markets, but at a lower level, namely in the telecom equipment segment. This is where the Open RAN approach is gaining real traction among operators, despite resistance and skepticism from major vendors regarding this new network architecture. For Russia, the Open RAN approach is especially attractive in light of the need to develop and produce network equipment within the country,
• the current technologically successful development of Starlink still has a greater impact on the traditional satellite communication market (in the countries of operation) along with putting significant price pressure on it. However, the company’s ongoing deployment of satellites with DTC functionality (direct to cell, direct satellite-to-smartphone communication), taking into account its scale and capabilities, creates the prerequisites for mass introduction of this service in a number of countries in the next couple of years. Herewith, we do not expect increased competition or emergence of significant extra revenue for mobile operators; rather, it will be a PR service for tourists and professional consumers.

In the field of smart devices, we would like to note the following:

• a certain crisis in the development of smart speakers and voice assistants manifested, inter alia, in the reduction of correspondent investment by Amazon and Google. In fact, smart speakers remained rather narrowly functional tools (listening to music, asking for weather forecast, news, searching for information, alarm clock) and did not become the core of smart home control system. However, there is a chance that the possible progress of voice assistants with the introduction of LLM (see below) will increase the popularity of this category of devices,
• Smart TV penetration (significantly higher than that of smart speakers) creates the prerequisites that this category of devices will be the center not only of home consumption of video content as it is now, but also the actual center of the entire smart home. This may intensify competition in the smart TV market, including through competition of operating systems for them,
• developers of AR and VR devices are still trying to find real needs for their mass application: the idea of metaverses has still remained a buzzword, while gaming is too narrow a niche.

In the field of machine learning and artificial intelligence (AI):

• mature solutions based on the machine learning technology have already emerged in a number of areas, such as image recognition and video analytics (including biometrics),
• the focus of development now is generative models, such as Large Language Models (LLM), text-to-image models, combined (multimodal) models; over the past two years, they have transformed from being an interesting toy, as to the general public, into a horizontal technology setting the basis for a new category of tools for increasing labor productivity in many industries, which, in a historical context, may be comparable to the processes of automation and computerization,
• further expansion of the capabilities of generative models, in particular LLM, may slow down for a number of reasons, such as the depletion of high-quality training samples and the increasing need for computing resources for training models (diminishing returns),
• at the same time, qualitative growth is possible due to new approaches and architectural tools (long-term “memory”, automatic addition of factual information relevant to the request, new approaches to learning, etc.), which will make the behavior of models more predictable (as well as acceptable as a business tool), while the models may become more compact (less resource intensive),
• overall, although there are few developers that are major market players in the world and in Russia, the risks of monopolization of this technology for the benefit of individual players are not so great: basic models (for example, in the form of open source projects) will be available to everyone, although, probably, lagging behind the advanced ones for 1–2 generations (which will give a certain head start to the corresponding players in the B2C markets).