Enterprises recognize machine learning as an important tool to be more competitive.
We have written and spoken a lot recently about the benefits of machine learning—artificial intelligence (AI) that enables computers to learn without being explicitly programmed by adapting when exposed to new data—for self-optimizing network systems used by telecommunications operators and service providers. But the utility of this technology is not limited to telecom; in the enterprise/business world, it is also being recognized as an important tool to gain competitive advantage, with similar adoption challenges and benefits.
For example, a recent report based on a joint survey by MIT Technology Review Custom and Google Cloud, concluded that although most companies are struggling to apply machine learning, a few early adopters are already realizing genuine ROI from this technology, thanks to their hard work on strategies that merge big data and analytics with AI.
The “Machine Learning: The New Proving Ground for Competitive Advantage” report looks beyond the buzz around machine learning in the enterprise/business space to examine the real-world potential for turning “raw data into useful, predictive tools for business.” The survey it’s based on—conducted in late 2016—was sent to companies (mostly in the U.S., Canada, and Europe) in the tech industry as well as business and financial services, with representative businesses ranging in size from very small (1-person) to very large (more than 3,000 employees). Respondents—C-level executives, enterprise developers, and VP-level executives—were asked to characterize their organizations’ current and planned machine learning strategies.
Current state of machine learning adoption
Key themes uncovered from the survey include:
- Machine learning is already happening; more than half of companies surveyed have already implemented strategies for this technology, and almost a third said they’re at a mature stage with those initiatives.
- Machine learning gives companies a competitive edge; just over a quarter of respondents said the technology already provides that benefit.
- Companies are investing significantly in machine learning, with just over a quarter of respondents reporting that more than 15 percent of their IT budgets are dedicated to initiatives using this technology.
- Early adopters of machine learning are already realizing the technology’s potential to extend data analysis efforts and increase data insights; almost half of respondents reported success in this area, with demonstrable ROI.
Why machine learning?
The most common machine learning projects are image recognition, classification, and tagging; emotion/behavior analysis; text classification and mining; and natural language processing.
Underlying the adoption of machine learning is a need to “gain better understanding of customer and prospect behaviors, needs, and desires,” the report noted. (No doubt that goal applies to the telecom industry as well as business generally, in terms of a potential application for the technology beyond assuring network and service performance.)
And, while it’s often difficult to achieve ROI with traditional online analytics and visualization tools, tying those in with AI to create machine learning appears to hold more promise is already reaping benefits for companies using it.
Indeed, this survey’s results reveal the “remarkable speed with which respondents are able to demonstrate ROI with their ML initiatives, which …. was not the case with big data analytics. Within the early-stage group, more than half report they are beginning to see a demonstrable ROI, and within the mature-stage group, more than have had demonstrated ROI.”
The report cautioned that, as of yet, those findings may not be representative of the greater business world, given that the companies surveyed were mostly innovation-focused companies and thus are further along than average with their machine learning strategies. These early adopters are leading the way for companies that are still unsure how to approach machine learning, and how to get started.
Getting a head start with machine learning
Regardless of what stage of machine learning development or implementation companies are at, the report concluded that there is an overarching motivation to incorporate the technology into their operations.
The report concluded: “These are still early days for ML, and many questions remain. But early adopters make it clear that business relevance and the ability to compete will increasingly hinge on working with machines to interpret and learn from data.”
That’s as true in the telecom space as any other industry, and forward-looking companies that adopt machine learning now will no doubt gain market advantage and move ahead of peers that are holding back.