Data Strategy
A data strategy is the essential foundation for any company looking to make better data-driven decisions. This strategy needs to include processes for data collection, data curation and management, understanding data integrity, and data governance. Without this foundation, companies have no framework or policies to leverage effective data usage. They cannot leverage data assets without any repeatable and documented process. This lack of process decreases productivity and also increases technology spend. In addition, the employees of the company are not enabled to properly leverage and take advantage of the data.
Analytics Layer
With your data strategy lined up, you now need data literate employees and analytic tools that provide the necessary capabilities to give you the necessary insights into your data. Depending on what the goal is, there are different types of analytic techniques that are required.
Typically, analysis starts with descriptive analytics, which is designed to give you an overview of your data. This type of analysis includes looking at things like frequencies, measures of central tendency, and measures of variance. Next there is exploratory analytics, which can be used to find connections and unknown relationships in your data. From there, depending on what you want to do, you may look at performing inferential analytics or predictive analytics. Inferential analytics can be used to help understand what causes certain outcomes. This includes drawing conclusions about a larger population. It is the basis for data mining and machine learning, and examples include marketing to customers, finding new markets, improving operational efficiency, and analyzing supply chains. Predictive analytics analyzes data to make predictions about future events. This could be, as an example, forecasting sales data, segmenting customers, or predicting future buying patterns.
These analytic approaches, combined with more advanced machine learning, aides in a critical component of data-driven decision making called augmented intelligence. Augmented intelligence helps humans become faster and smarter in their decision making.
Decision Making
Data will only make an impact if it is properly leveraged in the decision-making process. Companies can have quality data and skilled analysts who use analytics tools to present insights and recommendations, however if those are not communicated properly to the decision maker or the decision maker does not have the confidence or a consistent process to make the decision, then everything else is useless.
There are a variety of decision making models that can be leveraged. We will discuss them in a future blog post, however they all have similar components. One of those key components is making sure that you are not too focused on your mental model and you are open to thoughts and perspectives outside of them. Mental models are frameworks that include our underlying assumptions that come from our social interactions, values, and our experiences. Another key component is making sure you are aware of any and all bias in your decision making, whether it is intentional or not. Bias cannot be eliminated fully from your data or decision-making process, but it can be understood to help you make decisions with confidence.
Culture
This framework has to be wrapped in a company culture that supports data-driven decision making. Think about the questions below and see if your organization has the right culture to support this:
- Is there a specific data strategy within your organization that maps to your organizations vision?
- Is that strategy implemented so that data is organized, accessible, high quality, and governed?
- Are the employees within your company data literate?
- Are opinions and assumptions things that can be changed based off data?
- Does your company know how to arrive at answers and make decisions using data that empowers them to act?
- Does your company know how to evaluate their decisions after they are made?
In my next post, I will do a deeper dive on data-driven decision making, including a review of various models and frameworks that can be used.