Like many enterprises, you’ve likely made a hefty investment in analytic technology—from interactive dashboards and advanced visualization tools to data mining, predictive analytics, machine learning (ML), and artificial intelligence (AI). But, now that you have all these wonderful tools at your disposal, have you stepped back and assessed whether they have truly provided value and improved business results?
As Susan Athey, Economics of Technology Professor at Stanford Graduate School of Business, notes, organizations acknowledge that their analytic projects have failed to meet their expectations: “Only one in 20 companies has extensively incorporated AI in offerings or processes. Across all organizations, only 14% of respondents believe that AI is currently having a large effect on their organization’s offerings.”1
Limitations of common approaches to analytic projects
To paraphrase Shakespeare, the fault lies not in the technology, but ourselves—specifically in how we develop analytic requirements. Generally, organizations tend to focus on any one or all of these three areas:
- Pinpointing data that they think is relevant: A data-centric approach to analytic projects fixates on the data—and not on the insights—needed to improve decision-making.
- Determining which analytic technology to use: Defining analytic projects only on the basis of the technology to be leveraged and not in terms of the business the decision to be improved will inevitably yield poor results.
- Define the metrics or key performance indicators (KPIs) they want to improve: But analytics cannot magically improve metrics—they can only tell you if your analytic investment was worthwhile.
Focusing on decision-making changes everything
All these elements have a significant role in analytic projects. But, if you don’t take into account the decision-making you want to improve, you’ll derive minimal business value from your analytic technology. A better approach with a much higher probability of success is to use decision modeling to explicitly and clearly define the decision-making you want to improve.
You can start with a simple white board drawing to outline your model, such as this:
Or you can use modeling software to build a more refined and detailed decision requirements model. To make the decision requirements model readable by all teams—business and technical—the Decision Model Notation (DMN) standard is used to explicitly delineate how the proposed analytic knowledge will influence the decision.
There are many advantages of modeling decision requirements, including providing structure, enabling transparency, encouraging buy-in, and much more.
We’ve made this sound simple—and it really is once you get your arms around the decision-making you want to improve and you follow a step-by-step methodology. Imagine actually deriving measurable business impact from all the great analytics technologies you’ve invested in. We can show you how to accomplish this. And you can pass your wisdom and experience on to your colleagues.
Download our white paper, “Framing Analytic Requirements” to get more details and begin your journey to better decision-making today.