4 Keys to Overcoming Data Visualization Problems

Organizations are embracing data visualization as more than a tool to “see” trends and patterns in data but as a pathway to a dynamic culture of visual data discovery.

As with any type of cultural shift, there are going to be a few bumps along the road as innovative ways to transform data into actionable insights through the power of data visualization are sought.

However, with a few considerations kept top-of-mind in the early stages of data visualization adoption, common problems can be avoided.

  1. Take Time for Due Diligence Upfront - Avoiding data visualization problems begins with being sure you’re bringing the right tools into the mix in the beginning. No tool is a magic bullet for every data need, especially out of the box. It’s the organization’s responsibility to look beyond bells and whistles and thoroughly evaluate tool offerings to assess strengths and weaknesses in line with the needs and expectations of the organization before purchasing. Visualization tools are part of a larger data ecosystem, and knowing how a tool fits will help you maximize the value of your investment, streamline implementation, and minimize surprises. Additionally, it is important to have a clear idea of how users will apply the tool to solve business problems and where use cases, training opportunities, or requirements may exist that affect how tools are compared. Crafting a vision statement prior to embarking on tool evaluations will ensure tools are evaluated in line with the organization’s plans today and for the future.
  2. Establish a Common Glossary - Establish a common business glossary to facilitate a shared understanding of what data visualization is and how it is used that is understood, trusted, and utilized within the organization. Variations in definitions cause confusion, inconsistency, redundancy, and reduce the validity of analysis activities. Conversely, standard definitions promote consistency, streamline visualization activities (for example, a single visualization or dashboard can be shared by many rather than each group recreating the same visual), and can facilitate fruitful conversation because everyone is working on the same set of assumptions and meanings of the data, its business use, and its context. Additionally, expectations for version control and updating of the business glossary should be established, so that there is a level of assurance that this tool will continue to generate value within the business. Determine how often and who will be responsible for leading efforts to make sure the vocabulary stays up-to-date and accurate.
  3. Know Your Data Stewards - Using data visualization to nurture the democratization of data highlights the need for clear boundaries and data tracking and monitoring, especially among those who have the ability to make changes to the data or the visualization through their visual discovery activities. Data stewards do not necessarily own the data; they are the go-to people for questions, concerns, or doubts on the data’s use, quality, or context. They are also the people that can be counted on to contribute a meaningful definition of the data back into the business vocabulary. These stewards are often intermediaries between business and IT and speak to both sides. They understand the business drivers and needs and how the data supports them and are also versed in the entire lifecycle of the data, spanning generation or acquisition, where it lives in the data architecture, how it is administered, its security and access controls, and how it’s leveraged as a visual information asset in the business.
  4. Respect the Limitations of Self-Service - The demand for self-service is heard throughout the tiers of data users. However, it is necessary to exercise caution. There is a wide gap between empowering users with self-service and those users becoming self-sufficient. Different users will use data visualization for an assortment of needs, through a variety of form factors, and will bring to the table varying levels of expertise in visual design, data visualization best practices, and even data storytelling. There will be users who desire to be hands-on and deeply involved in building data visualizations, and there will be those who prefer to consume visual assets. Thus, it is important to know both the audience of data visualization and whether they will need to be presented data either within in a hands-on or consumption environment.


Subscribe to Big Data Quarterly E-Edition