OPINION: The path to industrial strength analytics
Paul Franks, Director, Financial Services, SAS
Innovation remains top-of-mind for many financial institutions as they seek to remain competitive in a lower growth environment. Yet in focusing on innovation, many remain ill-equipped to make better decisions in a world of increasing constraints and regulatory limitations. So how have leading organisations addressed the challenge of making better business decisions? Leading organisations recognise that effective decision making is an amalgam of the right people, processes, information and technology. They are placing more reliance on analytic models to support executives and line managers and help them make better decisions, and this is both positive and to be encouraged. Analytical models are at the heart of many critical business decisions whether finding new opportunities, managing uncertainty, assessing and scoring new or existing customers or pricing risk. As such, the number of models used to support real-time decision making can be expected to increase very considerably.
With analytic models being used in this way, they should be treated as the high-value assets that they are. Their creation should come from robust and industrial-strength processes and be managed for optimal performance throughout their entire lifecycle. Business analytics and technology teams need a repeatable and efficient process and a reliable architecture for creating, testing and deploying predictive analytic models into production systems.
In recognising analytic models as essential corporate assets, leading organisations seek to create the best models possible. However, few fully manage all the complexities of the interactive and iterative model life cycle. This is a multi-faceted task for large institutions as they manage through the key stages of problem identification, analytical data preparation, data exploration, model development, model validation and testing, deployment and assessment monitoring. In an effective analytical environment, data is created and accessed in the correct structure. Models are then rapidly built, tested and deployed into a production environment to generate trusted output. Their performance is constantly monitored and underperforming models are quickly replaced.
Analytics means more than creating a powerfully predictive model. It is about managing each of these lifecycle stages holistically across the entire portfolio of models. This is no easy task when analysts are developing multiple and competing models which are at various stages of development and customisation for different products and business units. An analytical life cycle is iterative and interactive in nature and models are continually revised and updated during testing and as new results and data become available. People from business and technology streams with differing backgrounds and capabilities are involved at various stages and success requires more than relying solely on the technology element. For the best results, organisations must look at the processes and have people with the right skills in place and working collaboratively in their designated roles.
The analytics ‘model factory’
The growing complexity and magnitude of the task – of managing potentially hundreds or even thousands of models in flux – puts organisations at the cusp of an information revolution. The old and inefficient “hand crafted” approach must evolve to a more effective “factory” approach. A predictive analytics factory formalises ongoing processes for analytic data preparation, model building, management and deployment – with particular attention to managing. As demand for predictive models rises, a structured approach enables an enterprise view of the organisation’s portfolio of models. A formal model management framework enables analysts to register, validate, deploy, monitor and retrain analytical models in the shortest possible time. A predictive analytics factory makes it far easier to document models and collaborate across internal and external teams. With a mechanism for feeding results back into the process for continuous improvement, it becomes clear which models continue adding value and which need to be retired. By bringing cohesion to a fragmented process, a predictive analytics factory enables more strategic thinking about models and how you can treat them as corporate assets.
Analytics projects and talent can evolve from the current technical focus into a stronger focus on business drivers and the understanding of problems in business terms. By starting with a decision in mind, the business and analytics teams are encouraged to think about how to operationalise the model, integrate it into businesses processes, and determine when it has outlived its original purpose. With a formal model management framework, the “best” models get into production faster to start serving the business sooner. The organisation can generate more models, and more sophisticated models, with a large variety of analytic methods, with fewer resources. Analytical models are continually monitored and refined, so they remain up-to-date and accurate. The process is also more transparent, making it easy to explain analytics-based decisions to regulators and business leaders.
Reliable processes, positive outcomes
As more and more organisations are discovering, a predictive analytics factory approach delivers a host of benefits covering efficient development, faster deployment, faster scoring, active monitoring and management, and reduced risk. Such benefits are already being realised by leading organisations that have been able to reduce model development and promotion cycles from three months to just days, decrease data preparation times by 40 per cent and improve analyst productivity by 50 per cent. These are impressive gains and very much aligned to the productivity agenda. Predictive models use your data to tell you about the likelihood of future events. Since nobody knows exactly what’s going to happen in the future, managing predictive models is about managing the uncertainty of future outcomes. That’s important enough to deserve rigorous process controls – a predictive analytics factory approach to analytical lifecycle management. For those seeking to become more innovative by harnessing their data assets for competitive advantage, such an industrialised approach is fundamental to commercialising new ideas.
- Paul Franks, SAS, Innovation, analytic models,
- AB+F Online
- Article Posted:
- May 01, 2013
Review this content
Fields marked with an asterisk () are mandatory.