While Agentic AI definitely means a turning point in the interaction of man-computers, passing from using a tool for cooperation, the next step is the integration of these agents and actual values. In VentureBeat’s Transform 2025Matthew Kropp, managing director and senior partner at BCG, offered a plan for the evolution of work flow, adoption of employees and organizational changes.
“Companies that are at the top of this curve – what we call the future built, those that are the most mature – see significant results: 1.5 times greater increase in revenues, 1.8 times higher shareholders,” said Kropp. “There is a value here, but we are early.”
Implementation, transformation, invention
To exploit and create a value with artificial intelligence and agents, the company must determine where to focus using implementation, transformation, inventions. AI is already being implemented in every enterprise and will have agents in the next few years. But if you give a chatbot employee, you haven’t changed your work. You need to think about work and transform functions, departments and flows of work, identifying where you can automate people’s work.
“We are now advising companies to focus on your three or four large rocks. If you have a large organization of customer service, you should use artificial intelligence in customer service. This has a huge impact. If you have a large engineering organization, you should use tools such as Windsurf to change the method of engineering, software development.”
The invention is still at a very early stage, but enterprises should think about how to exploit AI’s ability to be innovative, reason and plan. Look at the services and products and the way of interaction with clients: can you discover it again with these possibilities? “
For example, L’Oreal makeup launched a virtual beauty advisor to scale this exclusive service outside their retail locations, inventing the way they think about interaction with gigantic -scale clients.
Thinking beyond the basic cases of exploit
It is also very vital to think about how Ai changes your business. Over the past few years, a lot has focused on reducing costs, replacing employees, but this is not great thinking. AI strengthens the employees you currently have, radically increasing their performance.
“That’s what we see in creating the software,” he said. “I do not think that companies release their programmers. We will see a huge explosion in terms of the possibilities and functions that software companies build.”
In a BCG study with Harvard, Wharton and myth, they asked 750 knowledge of knowledge to write a business and marketing plan with generative artificial intelligence and without them. Participants using GPT4 performed 25% faster, performed 15% more tasks, and the quality of their production was 40% better. And when they received LLM, the lower performers on the base line were done as well as the best performers.
“This brought the performance of everyone, which is very powerful, because in most organizations new carpentry are less effective than more experienced people,” he said. “It has the ability to increase time for proficiency.”
AI can also exceed the human scale, even open novel applications that were not previously possible. For example, in the medical space, the results for patients are significantly improved along with preoperative and postoperative observation from a nurse, but the implementation of this was correct-and the arrival of AI nurses who can take this task for a gigantic patient population.
Overcoming the biggest obstacle: Adoption
Although these tools are fantastic, people do not exploit them. BCG tracks the adoption of Github Copilot and performance indicators for organizations from about 10,000 software engineers. 5% engineers doubled their performance within four months, while 60% showed zero improvement, because they simply did not accept the tool at all.
Why don’t people accept? There are three reasons. The first is the ignorance of the ability. Second, habitual inertia. The third is an identity threat and it is the most hard to overcome. Developers ask: “If this artificial intelligence can write a code for me, who am I? What is my value?”
“It will be the real work of the next three to five years,” said Kropp. “It makes people use agents.”
Strategies overcoming reluctance
There are several valuable ways to overcome these challenges. Of course, getting the right tool is the first step and integrating it with the way people work while training them explicitly. It is also extremely vital to measure and celebrate adoption for these employees actively using the tools so that everyone else begins to see that they have to get to this fashionable place.
Another vital step is to raise the deficiency – this means taking resources so that employees have to make more out of less. At the same time, it is necessary to redesign work processes with employees on the front line. Do not identify labor -intensive processes in which manual work can be automated – specify the parts in which people bring value.
“We minimize effort and maximize joy,” said Kropp. “We remain with a much more efficient process, a much more efficient company, a much more productive working force and work in which people like to be.”