Anthony Deighton is Chief Marketing Officer for Celonis.
Organizations are collecting and storing huge quantities of data, and most companies are finally turning that data into cash. The Economist Intelligence Unit conducted a cross-industry survey of 476 executives and found that 60 percent of businesses are already generating revenue from their data, while 83 percent have used data to make existing products or services more profitable. The challenges associated with a data-driven strategy, however, are familiar: Issues like regulatory uncertainty and budgetary constraints mean that business and IT leaders need to be meticulous in picking where they invest their budget.
Business Intelligence (BI) tools are used to transform raw data into meaningful and useful information for business purposes, help to identify trends, monitor customer behavior and detect significant events. More often than not, Business Intelligence sits within a visual analytics dashboard, providing employees with insights based on data averages. This is an important distinction to make because oftentimes insight is lost in averages. Businesses often need to drill deeper into their data because data trends are not enough to answer more specific, nuanced questions, or to pinpoint exact problems as they take place on an individual basis.
As a result, businesses need to move beyond the BI dashboard and invest in technology that provides a deeper level of understanding into how an organization actually operates. One such technology, process mining, analyzes all IT-driven processes from event logs already in a company’s system to provide a “full body” scan of a business in order to pinpoint exactly when, where and why things may be going wrong. This detailed, bottom-up view helps businesses gain a deeper level of understanding of how the organization operates. Decision-makers may find that what they thought was a business intelligence initiative to collect and analyze data themselves is really a process discovery initiative that can be conducted in a fraction of the time to jumpstart business transformation efforts. Here are a few issues to consider to help for determining the approach:
Why Assume All Processes are Operating as Planned?
It’s dangerous to assume that a process functioning at peak performance is optimized and functioning as well as it possibly could be. BI tools monitor Key Performance Indicators (KPIs) in order to measure how a process is being executed, but they don’t catch every inefficiency. BI tools are useful, but their nature is to simplify information, which means you inherently lose insights in the averages. If process KPIs fall outside the optimal range, there’s no way to identify the root cause or the corrective action that needs to be taken. Unlike BI, intelligent solutions built on process mining don’t just visualize data averages or assume that all processes are operating correctly. Instead, they highlight all process deviations and allow for granular scrutiny, and the most advanced solutions even recommend a roadmap to improvement. The idea is to focus less on seeking out marginal performance from known process improvement levers, and to focus more on re-imagining the process entirely, without the guesswork.
Spend Your Time on Transformation
The adage “you don’t know what you don’t know” holds especially true for process improvement and operational excellence initiatives. BI can improve decision-making based on what is relatively known, but solutions built on process mining reveal the unprescribed processes taking place. This is an important distinction, because the latter enables decision makers to make more informed business decisions that don’t lean on intuition or guesswork.
Process mining evaluates sequences of events over time and puts data into the context of a process, while BI tools simply analyze and visualize data. BI can determine the average customer acquisition cost by dividing the total acquisition cost by the number of new customers in a specific timeframe, but it falls short when analyzing and improving complex processes. Intelligent systems build on process mining analyze existing event-logs from IT systems, visually recreating the different variations without bias, and then automatically prescribing recommendations for improvement. These objective representations give businesses the capability to quickly examine enormous quantities of data, identify where root causes exist and then immediately start working toward a solution.
Actions that Create the Most Leverage
The real key to driving improved operational performance is determining which process changes will require the least amount of effort and yield the greatest result. BI tools can measure a KPI that illustrates how a process is operating within the organization, but they don’t answer the most important question: If you needed to make a single change to your process to improve it, what would it be? For your best customers or most expensive vendors, BI can be useful, but solutions built on process mining allow you to re-imagine what you would do differently within the process, to understand which parts of the process are breaking down, and to realize which steps to take to remedy process problems. Fundamentally, the difference stems from using prescriptive instead of descriptive analytics to realize actionable insights.
Traditional BI solutions use known information such as historical data and KPIs in order to help companies make better, more informed business decisions. While this is important for a top-level analysis, it provides very little actionable insight at an operational level to improve process performance. Intelligent systems built on process mining move beyond the analytics dashboard and use event-based approaches to dynamically visualize workflow-specific insights, generating quick ROI and assisting with strategic planning. It’s a no-brainer: These types of approaches create a more efficient way to empower employees with data-driven decision support, taking organizations a few steps closer to achieving true operational intelligence.
Opinions expressed in the article above do not necessarily reflect the opinions of Data Center Knowledge and Informa.
