IBM Buys IRIS Analytics To See Into The Future of Fraud

IBM has just announced the acquisition of German analytics firm, IRIS Analytics. Analytics acquisitions aren’t new for IBM of course; it has invested an estimated $17B+ in analytics alone in recent years. Big data and analytics, along with cloud and security, underpin the service provider’s strategic plans towards moving into higher value business services and creating new value for the enterprise.

Leading the Winner’s Circle in HfS’ 2015 Enterprise Analytics Services Blueprint, IBM continues to invest in industry relevant analytics. IRIS Analytics brings technologies for a broad range of banking and financial services clients: card issuers, payment processors, payment service providers and wallets. As the payments industry diversifies with more mobile and e-wallet payment adoption and other solutions, fraudsters aren’t standing still either. Fraud techniques are getting more sophisticated, making it harder to keep track of and counter fraud in a timely and controlled manner. The Nilson Report suggests that fraud losses worldwide reached $16.31B in 2014, and are estimated to reach $35B in 2020, with the U.S. making up almost half the worldwide losses. With that in mind, IRIS Analytics is an important addition to IBM’s fraud analytics offerings because:

• Real time is a real need for counterfraud. The IRIS system intercepts at-risk transactions to prevent fraud before it occurs, flagging and reporting cases for offline investigation. It is flexible enough to be configured for whatever level of authorization autonomy is requested, and since decisions are computed in real-time, there is minimal disruption to response times.
• Machine learning will become the most scalable way to predict fraud. IRIS uses machine learning to find new fraud patterns as they occur, towards modifying its predictive models and creating ad-hoc ones. These AI and cognitive techniques will become crucial as financial institutions try to reach greater speed and scale with their counterfraud modeling and mitigation.
• Talent augmentation is a key aspect of good analytics engines. The IRIS system focuses on augmenting the decisions taken by fraud analysts by providing them the best fraud predictions and recommendations. Further, IRIS is not just a technology buy for IBM but also a services play with teams that develop and manage decision support for clients, on the back of the product. Counterfraud learnings are fed back by these operational teams into product development, making them as invaluable, if not more, than machine learning progress.

Real-time decision engines are definitely the need of the hour at the rate that fraud and financial crimes are growing for an expanding range of payment types and providers. IRIS Analytics is a great niche buy, but what will be interesting to watch is how IBM integrates IRIS’ capabilities with IBM Watson.

 Integration has been a challenge in the past for IBM. To address it, last year, IBM announced the IBM Analytics group to bring together a tighter integration of the units that touch analytics—GBS, Software, Systems & Technology and Research. HfS is watching how this group is able to be a capability broker for its internal units and broad acquisitions like IRIS to provide clients with more meaningful analytics engagements that speak to their business needs and outcomes. As an example, IBM has the lucrative opportunity of exploring the applicability of this technology to other industries plagued by fraud, waste and abuse—insurance and healthcare are prime examples where enterprise buyers are seeking new capabilities beyond traditional predictive modeling.

HfS sees a renewed interest in the analytics marketplace for the banking and financial services industry, spurred by new technology advancements. IBM’s acquisition of IRIS is one such example. We will publish a Blueprint for BFS Analytics Services in March 2016 to further evaluate this competitive landscape.