Intel Corporation considers how the business side of the house can think about its maturity with respect to advanced analytics.
Whatever your organization is looking to achieve from analytics, its ability to be data-driven depends largely on the stage it is at in its analytics journey. In Getting Started with Advanced Analytics, we map out five steps to maturity, each of which builds upon the one before. As organizations evolve, they move from a focus on historical “what happened and why” questions to more forward-looking outcomes, as follows:
- Descriptive – answers questions about what happened in the past
- Diagnostic – offers insight into why things happened
- Predictive – provides insight into what could happen in the future
- Prescriptive – suggests actions that a business could take
- Cognitive – augments or automates decisions using human-like analysis
While these levels offer a useful gauge, in fast-moving and complex business environments it can be hard to pinpoint the level an organization has reached particularly if it is still at an earlier, more descriptive or diagnostic stage. You could engage a consulting firm who would be able to conduct an exhaustive analysis of each line of business, but this could be overkill. So, how can you establish a pragmatic picture of your organization, quickly and effectively, so you can identify the steps necessary to get you to the next stage?
When we work with customers across a range of industries as they start out on the road to better analytics, we tend to find that lines of business think in terms of how they use the tools, rather than what value the tools can bring. Specifically, lines of business tend to fit one of following models:
- Traditional, descriptive reporting – for example looking at weekly or monthly reports about sales figures, customer satisfaction indices and so on. Even with Business Intelligence (BI) reports, decision makers operate more on instinct and experience than data-driven insight.
- Enhanced reporting – reports are more accurate, timely and granular, and may incorporate external data such as social media reports. However, data sources still tend to be silo-ed and backward facing.
- Big data dashboards – insights are delivered to ‘the point of need’ through dashboard-based tools, offered directly to managers and staff so they can act more quickly. For example, in retail, product discounts or re-ordering can take place based on demand.
These models loosely fit the first three maturity levels of descriptive, diagnostic and predictive, respectively. From a business perspective, the job is one of moving from the passive use of data and towards more active data use, through real-time dashboards. This is what we mean when we talk about data centricity not as a technical term but as a business goal: put simply, do your people have the numbers and facts they need to do their jobs?
The goal of analytics is to deliver on this business goal. A wealth of technical options is available, from in-memory databases to cloud-based solutions: each has different benefits and constraints, which need to be considered first from a business perspective – what sorts of data do you have? Who is using it? What type of questions do you need answered, and how often? In the case of sales data, for example, it might be most useful to deliver a report of the previous day’s sales ready for the morning meeting, rather than the (more expensive) by-second reporting; other data, for example, service call response levels, may be needed in real-time.
As these examples illustrate, moving from passive models towards more (pro-)active use of data also requires the business to follow practices that can make the most of it. There is no point offering real-time data feeds to a line of business that wants to stick to a model of weekly cycles, or which is run by an executive that operates by command-and-control. In our experience, this is another reason why maturity needs to develop over time. It is very difficult to incorporate higher-order analytics into standard business practice, without lower-order mechanisms, skills, and understanding in place.
Of course, we would love to see all the organizations we work with benefit from machine learning and artificial intelligence to deliver on their business transformation goals. Even as we look to innovate across the sphere of analytics solutions, a big part of what we do with our enterprise customers is understanding where they are, then helping them build the infrastructure, knowledge, and competencies they need to advance.
These competencies, and above all goodwill and trust between the business and technology sides of the house, are essential elements of end-to-end analytics success. You can learn more about how advanced analytics can help you transform your business, and what you can do to make it happen, by downloading Intel’s new eGuide: Putting Data Analytics in the Driving Seat.