Key Takeaways From ‘The State of Machine Learning in Fintech’ Report
The use of machine learning in fintech is growing. You’ve likely noticed an increasing number of companies using it to provide personalized services, transform customer experience, or improve fraud detection.
Across market segments from insurance and real estate to personal finance and wealth management, companies of all sizes are using machine learning to improve their businesses and gain a competitive edge. But how exactly is it being used? What is the current level of adoption? And what hurdles are companies facing?
Being the data-driven bunch we are at Netguru, we wanted to know the answers to these questions and more. So we asked the industry.
Between the end of 2019 and the beginning of 2020, we surveyed 80 representatives from the fintech community around the world to get an up-to-date picture of machine learning in the fintech space. You can see the full results in our latest report, The State of Machine Learning in Fintech, and here, we share a snippet of our key findings.
Adoption of machine learning is growing – rapidly
One of the most significant takeaways to come out of the research was that almost three-quarters of the companies we surveyed are actively engaged in machine learning initiatives. Levels of maturity vary, but more than 47% of companies are currently running projects at an intermediate or advanced level.
These adoption rates show a sharp increase in comparison to a survey by O’Reilly, which reveals that in 2019, just 54% of organizations were evaluating AI and 27% were in the “mature” adoption phase.
Over 80% of advanced companies have been investing in data or machine learning for more than three years, while at the intermediate level, all companies have been working on it for at least a year.
Advanced analytics tops the use cases
Fintech companies are pursuing a wide range of objectives through the use of machine learning – from specific tasks like churn prediction and credit approval to broader goals such as task automation and increased operational efficiency
In our research, advanced analytics emerged as the top use case, with more than half of the respondents indicating that their companies were working on it. This result ties in with the finding that 54% of companies are adopting machine learning to extract better information from their data.
“The more accurately we predict, the better. We can save money and lives and improve the performance of every worker by being more efficient and making the right calls,” says Alvaro Mantecon-Rodriguez, CEO at TRIA UK.
The second and third most popular use cases were forecasting and fraud prevention, respectively.
While almost half of all companies reported results that were stable and in-line with their expectations, 14% felt that the results they had achieved were outstanding. Only 1% observed negative results.
Skills shortages are a major challenge
Our findings showed that fintech has not been immune to the global shortage of machine learning skills. Over half of the companies surveyed cited a lack of suitable talent as a key obstacle in adoption, which is representative of findings in other recent surveys, such as Gartner’s Research Circle Survey.
Almost a third of companies also reported low data readiness and budget limitations as key challenges, dampening the acceleration of machine learning initiatives in the fintech industry.
While you might expect larger companies to be leading the way in data readiness due to greater resources, surprisingly, 35% of large companies indicated that they are working with data but felt that they are not mature.
The future looks bright
“The future will be a combination of more complex algorithms, more data science, and a lot more process automation,” explains Mark Cheng, Management Consultant at Sia Partners. “If machine learning can take the error from existing processes and learn from them on its own, then the next processes will improve based on the prior changes.”
Our research indicates that companies are convinced of the benefits of AI. Almost 90% predict that their use of machine learning will increase in the next 12 months, with 45% expecting a significant increase. Advanced analytics remains the number one use case, and its popularity is set to increase – 62% of companies have earmarked it as a future project, versus 56% currently working on it.
“Our idea is to use AI to identify, select, record, and channel cases to the right departments and people,” says Piotr Kosior, Director of Global Expansion at AC S.A. “In general, we need to improve the quality and flow of service cases. Our second idea is to implement AI to support the effectiveness of remote training for our partners.”
AI is key to success
Whatever the future holds, it’s clear that fintech companies of all sizes are committed to extracting value from the explosion of big data in their businesses. And with the level of machine learning adoption rising rapidly, doing so successfully will be a key factor in remaining competitive.
AI technologies could potentially deliver up to $1 trillion of additional value in global banking each year, according to an estimate by McKinsey. The management consulting firm sums it up thusly: “For many banks, ensuring adoption of AI technologies across the enterprise is no longer a choice, but a strategic imperative.”
For more insights and further detail on our key findings, download The State of Machine Learning in Fintech.