Survey on enterprise machine learning

2020 State of Enterprise Machine Learning: Cost reduction is the top priority

Maika Möbus
machine learning
© Shutterstock / microstock3D

What is the current state of machine learning and data science in enterprises? For the second year in a row, Algorithmia have published their State of Enterprise Machine Learning report. It shows the top machine learning use cases in enterprises and what trends to look out for in the future.

Algorithmia Inc. have released their 2020 State of Enterprise Machine Learning report. It is based on a survey of 745 data scientists and business leaders in companies of different sizes and different ML maturity levels.

It turns out that an increasing number of companies are in the early stages of ML adoption, compared to the 2018 survey, or as Algorithmia puts it: “organizations are ramping up their hiring efforts to build larger data science arsenals.” The share of companies with more than 1,000 data scientists rose from nearly 2 percent to just over 3 percent compared to last year’s survey—among them are Facebook, Amazon, Netflix and Google.

SEE ALSO: 2020 Predictions – Centralizing AIOps Platforms

And here are the most popular enterprise machine learning use cases and possible future trends.

Top machine learning use cases

Across all company sizes, the three most stated reasons for using machine learning are to cut company costs, generate customer insights and intelligence, and to improve customer experience.

Company size is a determining factor for ML use cases. The top priority for companies with 10,000 or more employees and for companies with 1,001-5,000 employees is to reduce company costs. Smaller companies with less than 100 employees on the other hand place the most importance on ML that can generate customer insights and intelligence.

The reasons differ depending on the enterprise’s industry as well. For example, customer-oriented businesses in the retail, healthcare or manufacturing industry are most interested in ML applications that improve customer service, whereas banks are focusing on other use cases such as fraud detection. The energy sector is using ML to forecast demand fluctuations that may cause power outages.


Future trends

As the report points out, the machine learning landscape has changed over the past year when enterprise ML was “very much in pioneering days” and, according to Algorithmia, this trend will continue. They believe more innovation hubs to drive ML adoption will arise and director-level roles to advance ML will increase. They further state that mid-sized companies will employ more data scientists, whereas larger companies may encounter lower levels of customer satisfaction as they focus on cost reduction.

What is currently keeping many companies from extracting value are different challenges ranging from scaling to deployment, as Algorithmia claim. Therefore, they predict a boom in companies providing these services.

Background and methodology

Researchers at Algorithmia Inc., a serverless infrastructure provider, conducted their survey for the second year in a row to identify trends in enterprise machine learning. Last year’s report was based on data collected in the fall of 2018 and was released under the title The State of Enterprise Machine Learning 2018.

SEE ALSO: What 5 new innovations will open source yield in the 2020s?

The current report is based on survey data collected in the fall of 2019 among 745 respondents in a two-pronged approach. The respondents consisted of business leaders as well as ML practitioners in different company roles, such as individual contributors, managers or directors. Survey answers were given via a self-administered online questionnaire to be completed in an average of six minutes.

Further key findings are available in the report that can be accessed by entering your contact data.

Maika Möbus
Maika Möbus has been an editor for Software & Support Media since January 2019. She studied Sociology at Goethe University Frankfurt and Johannes Gutenberg University Mainz.

Inline Feedbacks
View all comments