Going Global with Predictive Analytics

Predictive analytics holds true promise for advancing enterprise performance. It can:

• Help increase forecast accuracy and business predictability thus allow for better planning and enterprise teaming.
• Boost profitability by identifying any product failures in advance, analyzing trends like seasonality and using big data to figure out what products are demanded by which customers.
• Drive cash flow by better assessing customer risk, using analytics to better understand supplier relationships and focus on specific customer segments.

A recent study conducted by Gilles Bonelli and Stephen Ferguson from The Hackett Group, found that what stands between full delivery of this tremendous potential and the current state of affairs is not the lack of technology solutions. The problem is that at most companies advanced analytics projects are handled on a targeted, local basis and should be scaled up to a global level to achieve their full impact.

The study concluded that while there is not yet a broad quantitative measure of value that show a correlation between the use of predictive analytics and top-performing enterprises, there is plenty of anecdotal evidence to support the case that using predictive analytics has a positive impact on cash flow, predictability and profits.

The study focused on specific case studies where predictive analytics were yielding real business value. In one case, a natural resource company was using predictive analytics to make smarter decisions about extending credit to customers. The analytics group built a model that used internal information about the customer payment behavior, external insights about the customer’s industry and also publicly available information about the customer’s financials to predict likelihood of default.

Overcoming the Obstacles

The Hackett Group research shows more companies are going to adopt advanced and predictive analytics tools in the next couple of years. Our 2017 key issues study revealed that while currently only 8% of finance organizations use advanced analytics as mainstream application, that figure is destined to jump to 35% between 2017 and 2020. The mainstreaming of advanced analytics will include the expansion of targeted pilots into enterprise-wide implementations.

On their path to bring predictive analytics to a global scale, companies will face the following obstacles:

1. They will have to define the purpose of the project. The objectives should relate the role of analytics to the key business decisions to improve predictability, profit and cash. Too often there are multiple opinions on what the project is supposed to deliver. The project will work when participants agree on a single shared vision.
2. They must decide on everyone’s roles and responsibilities. The study found there is disagreement where that analytics function should sit. To make it global, analytics must be its own function or a function with global reach and a common data governance structure.
3. They should hire or develop staff with critical data science and analytical skills who would be able to use the tools, identify patterns, ask the right questions of the data, put it in a useful business context and communicate about it effectively so it leads to action.
4. The must review their existing technology landscape. Legacy systems make it hard to pull together the data to feed predictive models. And the availability of multiple vendors further complicates the choice of the right solution. The first step is to review and close any foundational gaps. Then judge vendors on both their merits and preexisting relationships and how they integrate with existing architecture. 

Conclusion

The road to a full scale global analytics capability is long, but The Hackett Group research shows there are three things you can do:

1. Develop the business case for going global. Be specific about how predictive analytics can increase performance as you present to senior management.
2. Draw an operating model, i.e., go beyond the theoretical. How will the new function deliver services to its stakeholders? Who will be responsible for what activities
3. Assess the technological landscape and come up with a realistic roadmap for implementing new technologies and a future state vision.

It’s important to remember that new use cases and applications will continue to emerge for predictive analytics solutions, and that the solutions themselves will continue to evolve with the adoption of more artificial intelligence and cognitive computing capabilities. So, whatever the plan, it needs to be flexible and focused on continuous learning and improvement.