A quick Google search on the term “Analytics” yields varying definitions. Oxford Dictionaries defines it as “the systematic computational analysis of data or statistics” or the “information resulting from the systematic analysis of data or statistics.” Wikipedia further describes it as “the discovery and communication of meaningful patterns in data.” For data driven organizations, analytics is essentially the activity and process that takes, assesses, aggregates and analyzes the data – to create meaningful business insight.
Analytics is one of the hottest topics in both business and technology today. Whether it is the news of the business executive wanting to better understand and analyze their customer buying patterns, leveraging electronic health records to improve healthcare and health policy, or the hype around big data technologies and the promises of Hadoop – analytics are everywhere.
In a recent survey conducted by DATAVERSITY and in partnership with First San Francisco Partners, four out of five respondents (80 percent) said they would be investing in predictive analytics over the next five years. This is on top of the 17 percent that indicated they have already embarked on an analytics program several years ago. Assuming there is no overlap in respondents, then only three percent of enterprises are either undecided or not planning to invest in analytics at all. Clearly, money is being spent and that trend will only continue.
However, there has been an equal amount of indication in the press recently that many analytics programs – particularly those involving big data analytics – are having challenges, and are failing. In fact, an article in Forbes last year predicted that half of all big data projects will fail to deliver against their expectations.
If we revisit the definition above, this means that most organizations are failing to execute on that continuum from gathering data, to assessing/aggregating/analyzing, to achieve critical business insight.
There are a number of factors that have been shown to play key roles in the success and repeatability of an actionable analytics program. However, the five following capabilities are critical to turning data into insight.
1. Business alignment
First and foremost, it’s important to understand something about the insight you are seeking, in order to be sure you are looking in the right place, investing the appropriate amount of money and time, and are able to identify the insight once it is found. Business alignment is the understanding of the business purpose for the activity and assessment and recognition of the value that the activity provides to the organization.
This might not be perfectly quantified – although it is better if it is – but it is important that there is some understanding and agreement about the purpose for which you are placing the effort. When the goals are identified, it is possible to then articulate the information that is needed to achieve that goal and produce the insight.
2. Data understanding
Once there is an articulation of the information requirements, then it is possible to break them down into the specific data needs. Data understanding is the knowledge that you have about the data, its content and location.
Data understanding is expressed in organizations as business glossaries, data dictionaries, metadata or other places where information about the data is stored. It is important to have an understanding of what the data is and its meaning, when it was last updated and therefore how timely it is, where it can be found so data can be pulled, how to access it, whether it’s appropriate to use the data based on its sensitivity or level of privacy, etc. Many times it is also critical to understand who is accountable for that data in case there are questions or points of clarification needed to be sure it’s the right data.
3. Data quality
Next, it’s necessary to be clear on what it means for that data to be accurate for the purpose for which it is being used. After all, if the data is not of sufficient quality, where to find it and how to use it, it still may not serve the purpose needed to deliver the insight you are seeking.
Many times, data quality is vaguely described as “good” data. However, the definition of “good” may vary based on the usage of that data. For example, a marketing manager may want to do trending analysis for the number and types of products sold in an area. For this purpose, they rely on the postcode to be accurate. However, a shipping department needs the street address and number to reflect a building to which the product can be delivered. In this way, what is accurate for marketing at the postcode level is not considered accurate enough for shipping.
There are also many dimensions of data quality, including timeliness, relevance or accuracy among others. Data quality is a point that pulls together the business goals and alignment with the data understanding. The business goals will define the desired data quality dimensions and respective requirements, and the data understanding will tell you whether the data meets those requirements. If it doesn’t comply, it will also indicate whom to contact to determine the steps needed to improve the quality based on your needs.
4. Data-centric processes
You may be thinking all of this sounds very reactive, and that data understanding and data quality should be considered even before there is a requirement for analytics and insight. And you would be right.
If an organization already has data-centric processes, data understanding is gathered as new data is created, and data quality is measured when that new data is created and monitored as part of operational processes. Data-centric processes can occur in many departments where data is viewed as an important part of the process, not just as an output of the process. A project management process can be data-centric when the data needed is an input to the definition and design of the project. A systems development process can be data-centric when the data is considered not just the functionality, but also the UI, the infrastructure and the hardware.
Lastly, an operational process can be data-centric when it becomes equally meaningful to create new data in an accurate standard way, as it is meaningful to execute the process within a specific time frame. When an organization implements data-centric processes, the data is of higher quality and better understood, which enables the analytics process to be more agile and the quest for insight to be more real time.
5. Data-centric resources
Establishing data-centric processes also drives the need for better data-oriented knowledge and skills among the staff. In order to execute on a data-centric process, it will be necessary to know more about data management practices and develop those capabilities to create, manage, model and share data in efficient and effective ways.
Over time, the staff’s knowledge and capabilities to manage the data increases; thereby, creating an inherent knowledge and desire to implement data-centric processes throughout the organization. People are data-aware.
Tightly associated with that is when processes are data-centric, employees become more focused on the importance of data and the value it provides to the organization. They become data driven.
Creating a data driven organization means incorporating data understanding, usage and accuracy into all activities and behavior. This is very similar to creating a LEAN organization, or following Six Sigma quality processes.
Becoming data driven is also developing the aspiration for decision-making based on facts and information – data – rather than intuition or instinct. In many organizations, decisions are still based on instinct and “gut feel,” sometimes veiled as “experience” and then data is used to justify a decision. A data driven organization may still consider intuition, instinct and experience, but it would be in combination with data and information that provides facts. In this way, being data driven reinforces the goal of turning data into insight. In a truly data driven organization, the demand for insight is based on trusted, understood data.
As the need for analytics continues to grow, organizations will maximize their investment in an analytics discipline by focusing on the business purpose and goals, and the underlying asset that is used to produce the analytics – the data. And by understanding the data, ensuring it is of high quality and usefulness, the process to turn that data into insight will be more effective, efficient and repeatable.