The Adoption of AI in Organizations: Mindset and Cultural Change

26/10/2021 207

We already know it, artificial intelligence (AI) is reshaping businesses and organizations, although not at the speed some thought. Companies that are successfully adopting AI aren’t just launching a few projects to evaluate results.

These companies are changing organizational structures, changing how people and teams collaborate, and changing how managers make decisions.

Although technology changes and evolves exponentially, the capabilities that we develop for one of them, for example AI, will be there and will be necessary for the next technology. In other words, we are building new organizational capabilities to understand, take advantage of and exploit new technologies.

Whether it is the ability to manage and leverage data for decision-making, the building of trust in multidisciplinary and remote teams, or the ability to lead change based on technologies.

In a study by McKinsey, led by Tim Fountaine and his colleagues, and published in Harvard Business Review, only 8% of the companies surveyed applied good practices for a broad and successful adoption of AI.

We are building new organizational capabilities to understand, take advantage of and exploit new technologies.

Most companies had only run ad hoc pilots and applied AI in a single business process. AI faces organizational and cultural barriers within an organization.

Successful AI companies took action early on to break down and resolve those barriers to ensure they could capture AI opportunities. In other words, AI is not a plug and play technology with immediate results.

In a series of articles I will try to develop a series of recommendations and good practices for the adoption of AI in organizations. These recommendations will be based on the premise that the successful adoption of AI requires a correct alignment with the culture, structure, processes and strategy of the company.

We are going to start with culture in this article. Peter Drucker already said it to reinforce the importance of this organizational dimension: culture eats strategy for breakfast.

To widely adopt (on a large scale) AI in an organization, leaders must promote and achieve a change in behaviors, ways of thinking, or ways of acting in organizations. Let’s discuss three of these required changes:

Moving from a siloed organization to an interdisciplinary collaborative organization: AI requires the development of cross-functional teams with a mix of complementary viewpoints and skills. For example, when designing a solution (an algorithm that predicts something) we should also think about the changes that will have to be applied in operations with the client or with internal users. That is why we must include the end users in the team working on the design of the solution.

Moving from an organization that makes decisions guided by the intuition of the leaders to the organization in which those responsible for direct operation make decisions based on data:

It will not be possible to move forward successfully, if those responsible for direct operation must ask to bosses before making a decision based on data and the recommendations of an algorithm that was designed to enable better and faster decisions.

Be more agile, adaptable, rather than risk-averse: From my point of view this is one of the most difficult challenges, especially in some cultures where failure is frowned upon, and is continually avoided.

We cannot be successful in the broad adoption of AI if we must have the support of everyone before, if we must have everything under control before, if we must avoid failure so as not to be questioned or even fired.

An experimentation mindset will allow us to see failure as a way to get initial feedback, get to know the customer better and, in general, as a learning mechanism for success. And here comes the main challenge: leaders must be convinced that this is the right way. Then they can prepare the others.

In summary, in this first post, companies must face and work on the cultural changes necessary to ensure that the adoption of AI can scale successfully.