We live in the early stages of a data analytics revolution. Data is driving the world and transforming the global economy. In a report published in December of 2016, McKinsey estimated that “45% of work activities could potentially be automated by today’s technologies and 80% of that is enabled by machine learning.”
While big data and AI investments are on the rise, many companies are struggling in their efforts to be data driven. A 2019 HBR article found in a study that “An eye-opening 77% of executives report that business adoption of Big Data/AI initiatives is a major challenge, up from 65% last year.”
To better understand business analytics and why companies are struggling, I recently attended an executive education course at the Carlson School of Management. The class covered a range of topics including artificial intelligence (AI), machine learning, big data. As a consultant who manages IT projects, I wanted to learn about what it takes to deliver analytics projects.
What Is Business Analytics?
Business analytics is an exploration of an organizations data, with emphasis on statistical analysis. It enables companies to make data driven decisions that help gain a competitive advantage. Companies that deliver business analytics projects treat data as a corporate asset. Below are the four main types of business analytics:
- Descriptive Analytics – Descriptive analytics answers the question, what happened? It is an exploratory analysis that provides visualization and BI dashboards. This can help companies to use key performance indicators to understand the state of the business.
- Predictive Analytics – Predictive analytics answers the question, what will happen next? It analyzes trend data to predict future outcomes. Data mining, machine learning, model lifting, and forecasting are techniques used.
- Prescriptive Analytics – Prescriptive analytics uses past performance to generate recommendations for the future. Optimization, simulation and rules are used in prescriptive analytics.
- Causal Analytics – Causal analytics answers the question, did x truly cause y? Before business decisions can be made based on analytics, you need a high degree of confidence. It’s not enough to go by gut feel, data trends or correlations. Casual provides the confidence using a/b testing, econometrics and experimentation.
Why Most Business Analytics Projects Fail
Gartner CIO research reports that more than 50% of analytics projects fail. Why do they fail? In short, most fail because companies neglect to connect the analytics to business value. They focus on analytics output, tools and technologies verses business outcomes. Some common pitfalls include not knowing what the problem is, force fitting a solution, or trying to boil the ocean. Poor data quality and a lack of technical expertise are also big issues.
Another reason business analytics project fail is due to the way they are managed. The traditional plan driven approach does not work for analytics projects. “Business analytics projects are often characterized by uncertain or changing requirements and a high implementation of risk. So, it takes a special breed of project manager to execute and deliver them.” (Viaene, Van Den Bunder, MIT Sloan Mgmt Review).
Setting Up Analytics Projects For Success
The best way to setup analytics projects for success is with framing. The goal of framing is to define the problem boundaries, break down the problem and determine the analytics method. This helps to reduce ambiguity and the risk of project failure.
There are different frameworks that can be used throughout an analytics project life cycle. Below is a description of a common framework that can help:
Situation, Complication, Key Question (S, C, KQ) – The S, C, KQ framework helps gather the information needed to provide a clear and holistic problem statement. The Situation provides the information relevant to the problem. It is the context that sets up the complication. The situation must lead to the complication. The two must connect.
The complication clarifies the need for change. It specifies why a change is needed and a decision must be made. The key question then follows from the situation and complication. It is the one focus of the project and it makes clear the decisions to be made. The key question should use the SMART guideline – Specific, measurable, action-oriented, relevant and time bound.
Below is a basic example of what the S, C, KQ framework might look like in a project definition sheet. The example provided is to solve a business problem for a health club:
The example gives provides a basic idea of the S, C, KQ framework. Once the problem statement is clear, the analytics project team can decide on which type of model to use. The final step in the framework process is to provide an outcome that is measurable and actionable. The outcome explains what business decisions will be made based on the results of the analysis. The outcome also explains how success will be measured.
With a solid framework in place, a cross-functional analytics team can then develop a model. A typical deliverable from an analytics project team might include a web dashboard that provides predictive or prescriptive analytics. Data engineers and data scientists write algorithms and build the models. They also perform data cleansing, aggregation, integration and transformation.
Summary
Companies need to treat their data as a corporate asset. The ability to use data to predict future performance provides a competitive advantage. While business analytics is growing in importance, many companies are failing in their efforts to become data driven.
A key to success with business analytics projects is to use a framework that will align the analytics with business value. Analytics projects must be managed in an agile way that adapts to changing requirements. An iterative and experimental approach to project delivery is best.
Often the biggest roadblock to business analytics success is the business imagination. Companies must find ways for human intelligence and machine intelligence to work together.
A special thanks goes out to the Carlson Executive Education program and professors Ravi Bapna, Gedas Adomavicius, Ellen Trader, and De Liu. Below is a picture of our business analytics class filled with business leaders from the Twin Cities.
About the Author: Mike MacIsaac is the founder and principal consultant for MacIsaac Consulting. Mike provides leadership as an IT Project and Program Manager as well as an Agile Scrum Master/Coach. Follow Mike on Twitter@MikeMacIsaac or subscribe to Mike’s blog.