- Almost every organization has encountered challenges in implementing AI and advanced analytics initiatives.
- But there are some basic steps to follow to increase the chances of success.
- Insider Intelligence publishes hundreds of Connectivity & Technology industry insights, charts and forecasts with the Connectivity & Technology Brief. You can learn more about subscription here.
AI is difficult to understand and difficult to implement. Almost all organizations have encountered challenges in implementing AI and advanced analytics initiatives, according to Ericsson IndustryLab research.
A full 91% of organizations surveyed said they face issues in each of the three challenge categories studied, including technology, organization, and culture and people. Of these, humans remain the biggest hurdle to overcome.
AI tools and technologies are still evolving rapidly compared to other IT products and services, but the challenge of adopting AI is not simply waiting for a new version of an algorithm or a new powerful cloud service hits the market. Adoption of AI and advanced analytics is as much about installing new software as it is about the humans using it.
According to the Ericsson report, 87% of respondents said it was more difficult to overcome challenges with people/culture when it comes to AI adoption than challenges presented by the technology or their organization. Of the top 10 critical AI implementation challenges, seven are related to people, culture, and the organization’s budget and staff. The top three people challenges:
- People prefer to stick to proven routines.
- Employees are afraid of losing their jobs if these technologies take over.
- Lack of qualified employees for tasks related to AI or advanced analytics.
A large number of companies are lagging behind in terms of the breadth and depth of AI adoption. As many as 49% of respondents said that AI or advanced analytics tools had been fully implemented in their companies; these companies have been ranked as “AI leaders” by Ericsson. However, the second largest group was ‘AI enthusiasts’, with 41% having partially implemented AI or advanced analytics tools, followed by ‘AI newbies’ who were only ‘implementing tools.
Digging into the data, there is wide variation in the number of AI projects companies have completed, with a small number of respondents saying their company has completed more than 100 projects. The majority – 60% – have completed between six and 20 projects. Rebecka Cedering Ångström, Ericsson Principal Researcher Noted that even AI leaders faced many challenges. Although they started their initiatives an average of four to five years ago, 60% said they had encountered difficulties with the projects.
Despite the challenges faced by AI projects, enterprise adoption is growing and addressing the human factor in AI will be key to a successful AI implementation. The challenges surrounding the training and motivation of management teams and employees are significant; however, experts interviewed by Ericsson noted three main strategies to help de-risk AI projects:
- Organize technology workshops, training sessions or training courses for employees.
- Encourage employees to develop new skills related to AI or advanced analytics.
- Educate employees on the benefits of changing work routines.
The shortage of skilled data science workers is one that has been well documented in recent years, however, and will be an ongoing issue for companies looking to grow their talent pool.
Want to read more stories like this? Here’s how you can access it:
- Join other Insider Intelligence customers who receive this briefing, along with other connectivity and technology forecasts, briefings, charts, and research reports, in their inboxes every day. >> Become a customer
- Explore related topics in more depth. >> Browse our coverage
Current subscribers can access the entire Insider Intelligence content archive here.