Data Visualization & The Human Factor

Here we are in the Information Age, and the amount of information available to you, to substantiate just about any point you might want to make, is potentially limitless. But the amount of time you have to present that information isn’t limitless and, in fact, is a key factor limiting what information you present and how. Information must pair with the recipient’s available time and be at the level of the recipient’s ability to find meaning. Visual tools can help us do this, with data visualization as the method. To effectively convey meaning, the data visualization method must be properly matched to the target recipient. 

Whereas years ago, data scarcity was a problem, today’s challenge has become data abundance and the need to prioritize multiple, differing data points. When each point requires scrutiny, arriving at insight and driving to action may slow considerably. The ready accessibility of data across nearly every platform we interact with has created the perception of readily flowing insights. And yet, when we consume data, we need to be wary of a deluge of information without value. As an analyst, this scale has made the ability to both “think data” and “speak human” a premium skill, because working with business clients to uncover “the ask” takes time and planning.

It requires us to bridge the gap between providing data in rapid response to ad hoc queries and providing more strategic, ongoing thought leadership. Responding to business queries with in-depth analysis is the ideal; however, time constraints often force insights to be delivered without the benefit of an analyst’s assistance.

In measurement strategy, we consider how the business objectives we achieve can be demonstrated in observable data—what aspects of the data ecosystem define success, what data to obtain and analyze, as well as what information to present back to the business team. As a recipient, our ability to translate data into meaning may take several forms depending on how we consume information. Information scale is the idea that the amount of information presented must scale to the amount of time the recipient will spend with it, and the recipient’s ability to process and retain meaning (i.e., their data literacy).

While data visualization is intended to convey meaning on its own, guided exploration can still benefit the recipient by directing attention to key elements of the presentation. A dialogue with an analyst can help strike this balance. However, what happens when the user doesn’t have the luxury of having a dedicated resource? Certain linear formats such as pre-defined paths, click-through presentations (Prezi, WebEx) or video/animation can offer that same presentation control.

In contrast, the deeper the recipient is in the subject matter the more autonomous the user can be, and the more they may benefit from a self-directed setting where they can explore the data with fewer pre-defined sets of views. For these users, a layered approach to provide aggregate-level and drill-in functionality draws the user in, inviting them to pursue additional discovery.

Dashboards that combine more than one element (e.g., relevant graphics with words and/or an audio component) encourage deeper understanding—known as the “multimedia effect.” Dashboards are fundamentally utilitarian, whether they drive analysis and insight, or focus on immediate actions. They are able to orchestrate differing data sources, like manual data imports or real-time data feeds, and incorporate them to a single view. While real-time or near-real-time dashboards drive a vision of immediate action, they are most impactful for teams that are able to act in that same near-real-time capacity.

Operational dashboards aim to impart critical information quickly to users as they are engaged in time-sensitive tasks. Analytical dashboards provide the user with at-a-glance information used for analysis and decision-making, but don’t have the same level of time sensitivity as operational dashboards.

Where multiple data visualizations coexist in a single view, prioritization of elements assists users in finding meaning, letting the design reflect the hierarchy/importance of the data, and allowing design elements (white space) to focus the users’ attention. Juxtaposed elements build a balanced understanding of a particular initiative and while we may want to add all data surrounding a matter, brevity in visualization selections may tell a more effective story. When forced to add visualizations into a single view that don’t contribute to a coherent story, utilize dividing elements to set each apart.

Data visualization formats must also convey meaning outside of their initial presentation. When information presentation happens within subsequent conversations, portable static/compiled visualization formats, such as exported images or applications, can retain message control. These formats limit access to underlying data, limiting the chance of misrepresenting the findings.

A good data visualization does a few things. It stands on its own; if taken out of context, the reader should still be able to understand what a chart is saying because the visualization tells the story. It should also be easy to understand. And while too much interaction can distract, the visualization should incorporate some layered data so the curious can explore.

Understanding the impact of storytelling may be the biggest opportunity for analysts today. The audience begins to construct a narrative of meaning internally with the simplest of data visualizations. As we shape the data exploration experience, we must decide how we would like the story to be told. Often, we are looking to present a linear set of points that construct a broader view of understanding. However, when control of the story shifts to the recipient, a multi-path/matrixed experience can be created to provide exploration through the data.

Linear storytelling remains the most recognizable framework, a construct which takes the audience through a structured flow, an introduction, exposition, a narrative hook, and conclusion. Static data charts, combined data dashboards, compiled click-through presentations, and even infographics rely on the storyteller to construct how the information is consumed. Animated and video presentations push the linear aspect further by taking full control over timing.

Alternatively, much of today’s data is available for a self-directed exploratory experience. Web or mobile app-based experiences accommodate the user’s ability to explore within data sets. The framework shifts to a matrix, allowing multiple paths across different dynamic states (either pre-determined or generated by the application on the fly). Platform providers are continuing to push both aspects. For example, Tableau’s Story Presentation provides for a linear set of slides, yet enables digital dashboards to allow exploration underneath aggregate data views.

When it comes to data visualization, it helps to keep in mind that we are rapidly evolving something that has always existed. Human beings have always known how to view and process data, so keeping that notion as central is critical.

Therefore, when we use phrases like big data, predictive analytics and data science, it is important to frame our thinking in a way that is tangible. Everything we are talking about in depth here is an evolution of long-standing methods of communication that had very similar goals. 

Keep the human factor in mind and consider information scale, presentation context, and storytelling model when proceeding to visualize data. Think of visualization, first and foremost, as a method of communication, and not a technology.

We’ve chosen a few key scenarios where you may be asked to convey information effectively in constricted situations, and provided use cases, recommended approaches, and examples for each. Download the full report to dig in.