Corporate AI Adoption: 70% of Success Depends on Processes, Not Models

AI adoption in a company depends 70% on reworking business processes, not on choosing a model, said Dmitry Chudinov, Deputy CEO for digital transformation at Russian Post, at the CNews FORUM conference.

Why are there so few working AI projects?

A year ago, about 50% of Russian companies said they were adopting artificial intelligence. Today it is almost 100%: AI has firmly entered corporate strategies, budget cycles and board reports. The technology comes up at every industry conference, and companies allocate budgets and form teams around it.

But the growth in declarations has not produced growth in projects that actually work. Chudinov observes that functioning AI deployments are still extremely rare. The market is full of pilots, demo stands and "AI on slides", but not of systems that run reliably in production and deliver a measurable operational effect.

The key reason is misplaced priorities. Companies start by picking the "right" model: a domestic LLM, a foreign neural network, an in-house build or a fine-tuned open-weights solution. Resources, vendor negotiations and most of management's attention go to this technical choice. The question of how real workflows will change, including regulations, the distribution of responsibility and decision points, moves to the background or is never asked at all. Companies imitate transformation instead of carrying it out: they buy the technology, report "AI adoption" and return to their familiar operating routine.

What "70% is processes" means

Russian Post describes a different approach: operational transformation first, technical choice second. This means that before choosing a supplier or a model, a company must answer basic questions: what exactly is changing, how the effect is measured and who is accountable for the result.

According to Chudinov, 70% of an AI project's success is determined by organizational work: rethinking who makes decisions and how, which steps are fully automated, and where AI is built into the work cycle as a full participant. This is the stage where a company finds out which data it needs, who owns it and how the human role in the process loop will change.

The remaining 30% is the technology choice: model, infrastructure, integrations. It is an important stage, but it should follow operational design, not precede it.

For Russian Post, with its nationwide scale of hundreds of thousands of employees and a logistics network across the country, this is not an abstract principle but a practical requirement. Every AI deployment must deliver a measurable result: speed, accuracy, lower operating costs. Without reworking processes first, the model is dropped into the old scheme and reproduces its inefficiency.

What this means

The gap between "almost 100% of companies talk about AI" and "only a handful get real results" is not a failure of the technology. It is a failure of management: organizations buy a tool without changing the environment in which it has to work.

Chudinov's thesis is simple and uncomfortable at once: for AI to work, the company has to change itself before the neural network goes into production. Technology amplifies what is already there: if a process is well built, AI speeds it up; if not, it exposes the chaos. That is why AI transformation is a management project, not an IT purchase.

How many Russian companies adopted AI in 2026?

A year ago, about 50% of Russian companies said they were adopting artificial intelligence; today it is almost 100%. AI has firmly entered corporate strategies and budget cycles.