David Menninger's Analyst Perspectives

I’ve previously written about the analytics continuum, which spans a range of capabilities including reporting, visualization, planning, real-time processes, natural language processing, artificial intelligence and machine learning. I’ve also written about the analysis that goes into making intelligent decisions with decision intelligence. In this perspective, I’d like to focus on one end of the analytics continuum, which I’ll label advanced analytics.

Broadly speaking, I include statistical analysis, regressions, artificial intelligence and machine learning, optimization, goal-seeking, risk and sensitivity analyses in what I consider to be advanced analytics. Many of these analyses are conceptually simple, but mathematically difficult. In previous perspectives, I’ve railed about the fact that we’ve made analytics and business intelligence too hard and therefore not accessible to the majority of the workforce. But sophisticated analytics do have a place in an organization’s information architecture. They are not tools that everyone will use. In fact, I assert through 2026, optimization analyses will remain independent of business intelligence tools requiring 9 in 10 organizations to maintain multiple, separate skill sets. However, the output of these more sophisticated analyses can be valuable to many in an organization.

For example, in sales and marketing, organizations could make many different offers to many different customers and prospects. They must evaluate the likelihood of responses to different offers given the cost of each offer, the cost of the medium where the offer will be made, the margin that will be returned if the offer is accepted and the availability of resources to produce and deliver the goods or service in question.

In production planning, organizations must evaluate the trade-offs between producing different products, where to produce those products and when to produce them. There are many constraints to consider, including production capacity, product costs, available personnel, product demand and service levels. In supply chain management, there are a limited number of trucks, drivers, warehouses and an inordinate number of routes that could be driven. In addition, there are constraints of fuel costs and delivery commitments.

In healthcare, providers must evaluate the trade-offs of different treatments for different patients, with consideration of likely outcomes, potential side effects, insurance coverages and cost to patients. They must also maximize the utilization of facilities and equipment, evaluating the length of time of each procedure, the personnel required to administer the procedure and the demand for different procedures based upon constantly changing patient demographics.

In all of these cases, advanced analytics can be performed to help inform decisions associated with each scenario. Typically, there is no silver bullet, but a well-informed decision that includes an assessment of the risks and trade-offs is better than a decision based on instinct or intuition.

I encourage you to consider which of these types of analytics your organization may require. There is a long list of vendors of optimization software in Wikipedia. Very few of these vendors also market traditional analytics and BI systems. To take advantage of these tools will most likely require specialized skills. To maximize the value of these tools, it is also important to understand how these tools can be integrated into your information architecture. The goal is to help your organization become more efficient and effective by making more informed decisions and sharing that information as broadly as necessary.

Regards,

David Menninger