Today MEJOIN would like to point out the current state of machine learning in companies based on a study by Algorithmia. This study includes interviews with 750 decision-makers who are actively developing machine learning or have just started to do so. Primarily, the research presented seven key findings that reflect the current atmosphere in the industry.
The first finding is that there are still relatively few data specialists in many companies. Half of the companies have stated that they employ up to 10 data scientists, and around one-fifth already employ up to 50 of them. In contrast to the same study in 2018, these figures have already improved, although there still seems to be miserliness of human capital.
Therefore: Data scientists create the basis for successful machine learning. They are, therefore, an equally important reason for investment as AI applications!
The second topic deals with the use cases of machine learning. Here, the focus of the companies is on reducing costs and generating customer insights, which account for almost 40%. Improving the customer experience (34%) and improving internal processes (30%) also remain essential reasons for using AI. The survey also measured the reasons for using ML concerning the number of employees. It was found that for companies with less than 100 employees, client insights are the most crucial reason, whereas cost reduction is the most important reason for companies with more than 1000 employees.
Therefore: The cost reduction/improvement of internal processes is an enormous strength that goes hand in hand with machine learning. But don’t forget the possibilities that can be realized about your customers. After all, these are the ones that generate your revenues!
The third part of the study deals with the degree of maturity of machine learning/ knowledge in the companies. Here it is almost an equal distribution over the different options, only the early adopters, who have already more than five years of experience, are less represented with 8%. The „newer“ players, who have less than four years of experience or are just developing use cases, are divided among the remaining respondents (between 15 and 21%). Here again, the results were listed with the dependence of the number of employees. Here it is apparent that larger companies tend to have started development/planning somewhat later, whereby all company sizes overall have little or no experience in the implementation of ML use cases.
Therefore: There are many synergy effects through the exchange with other, more experienced companies, although no precise information is given here. One to two years of productive experience in machine learning can already serve as a suitable template for many companies!
The second part of the study evaluation will be available next week, with the subsequent four key findings…