Can (And Should) Management Dashboards Tell Stories?

Before going any further, let me clarify one thing:

Dashboards are not used to tell data-driven stories.

Wait? What?

Indeed, dashboards are not typically designed to convey a single, tailored story, and I believe forcing them to do so at all costs is unnecessary.

The essence of storytelling involves delivering a one-off, targeted message aimed at achieving a specific outcome. For such purposes, presentations, memos, or documents are more suitable. Restricting a dashboard to a single narrative would diminish its broader utility. It’s akin to using a tank to destroy a hornets’ nest – while possible, the collateral damage to the surrounding area would be disproportionate and the whole endeavor impractical.

Yet, management dashboards are crucial in modern organizations as they support operations and enhance decision-making processes. They offer insights into key performance indicators (KPIs), consolidating data from multiple sources into a single interface. This integration supports informed decision-making and automates data collection, saving time and minimizing errors. Additionally, dashboards facilitate communication by aligning team knowledge. Finally, dashboards, if accessible on various devices, allow users to access critical information from anywhere.

These capabilities render dashboards indispensable tools for building operational efficiency and promoting data-driven decision-making across different management levels.

Last year, I discussed the techniques to enhance a dashboard’s storytelling capabilities, drawing from my experiences as a dashboard designer and a user.

Leveraging management dashboards for storytelling: a viable pathway?

I would like now to extend the concepts presented in that older post. I think embedding some of the storytelling principles can significantly improve the experience of dashboard users and increase their engagement.

Therefore, in this post, I’ll guide you through designing dashboards that navigate users through various narratives, ensuring information is absorbed efficiently and effectively. This will be achieved by incorporating storytelling techniques into dashboard navigation, page layout, and individual Visualization design.

Why should dashboards tell stories?

Storytelling can significantly enhance the functionality of a management dashboard:

• Storytelling can help simplify complex data. A well-thought single-page layout and intuitive navigation through multiple dashboard pages can considerably streamline data exploration. Top this up with a few data-driven storytelling design rules applied to visuals to further strengthen that effect.
• Storytelling has the transformative power to elevate dashboards into practical communication tools. Use the power of titles, tooltips, or interactive commentaries to increase understanding and channel the interpretations.
• Storytelling can facilitate seamless transitions between the helicopter view and detailed exploration. Start from that one KPI everybody in the organization talks about and then go deeper, adding detail, dimensions, measures, and data sources.
• The integration of storytelling and interactive features in a dashboard creates a powerful tool for data exploration. Unlike static reports, dashboards enable users to interact with data in ways that suit their specific needs or interests. Users can compare historical and current performance, as well as customize date ranges, which adds a dynamic and engaging element to the data exploration process.
• Storytelling provides valuable guidance on effectively using color coding in dashboards. Color serves as a potent tool for conveying information and enhancing narrative within the dashboard. This approach enables immediate and intuitive data interpretation, often without the need to closely examine the numbers. By strategically employing colors, dashboards can communicate insights more clearly and effectively.

So, why don’t dashboards do that?

As mentioned earlier, dashboards are typically not designed primarily for storytelling. As Brent Dykes would likely argue [1], their primary function is storyframing – helping users identify and extract insights from data. And I will not attempt to transform this purpose. Instead, I will demonstrate how creatively applying techniques derived from storytelling can enhance the storyframing function of dashboards, making them even more effective in revealing and communicating key data insights.

Before we delve deeper, let’s consider the common shortcomings of many dashboards we encounter daily, which hinder their ability to tell stories and uncover insights within the data effectively.

1. The information users want to get is difficult to find. First, a user must open the correct dashboard, select the right page or tab, choose the metric, and set up the filter combination. That takes several clicks, assuming he or she makes all these steps error-free. And then, every filter change causes the whole thing to reload, which takes several seconds…
2. The dashboards do not address the users’ needs. For example, presenting a complex, multi-source dashboard filled with numerous small visualizations to company executives is problematic. Such individuals often don’t have the time to figure out how to use that dashboard effectively if at all. As a result, they become discouraged and revert to spreadsheets or emails. Similarly, providing a simple, straightforward dashboard that tells a specific story to data analysts isn’t helpful. Data analysts usually prefer to discover detailed insights through exploring unique combinations of data sources and trends. And what if we have an executive who likes to drill through the data? But has no time nor patience to perform all the steps above? Or no technical skills to set up a preference dashboard view?
3. The dashboards lack focus. There are essentially two problems here. One is the so-called "data bloat". It refers to a situation where, accidentally or not, we display the same information on the same page (or dashboard) using various visualizations. The second problem is when we show a lot of unrelated visualizations on one page based on different, unaligned data sources, causing confusion and unnecessary questions, resulting in an overarching lack of confidence in our current and future work [2].
4. The dashboards are not relevant to the organization’s activities. Many dashboards fail to address the organization’s business needs. Given the rapid changes in business environments, only the fortunate few can use the same dashboard and data sources without needing updates. Effective dashboards must adapt to remain relevant. They should tell stories that align with current company situations, strategies, and goals.

From that perspective, dashboards can be categorized as:

• Relevant dashboards: Those that meet specific, ongoing business requirements.
• Outdated dashboards: Initially relevant, these no longer serve a clear purpose as their original business needs have been forgotten or have evolved.
• "Experimental" dashboards: Created to explore new data or tools rather than to meet a specific business need.

The beginning letters from the list above form the acronym "ROE," . Surprisingly, that’s also the acronym for one of the most critical KPIs for every organization: Return on Equity. This metric measures the profit a company generates from its investments. The principle of profit generation applies to all activities, assets, or tools a company utilizes. This includes management dashboards as well. In other words, they must provide the highest possible return on the effort to create them. For that, they must be relevant, up to date, and serve a real business need.

Dashboard design tips from data-driven storytelling

A good dashboard can be compared to a comic strip. This means that while it may not offer the extensive text and detail found in a book or the three-dimensional experience of a movie, it effectively and engagingly conveys the story. It won’t be perfect; some story elements will be simplified, and not all nuances may be visible. However, according to Pareto’s principle, it will achieve most of the desired effects efficiently.

And yet, most of the dashboards are like the collage pictured below. The only feeling they cause in users is panic (images and story in this paragraph inspired by [5]).

Okay. So what do you do to avoid such a mayhem? The first principle is to design your dashboard with a data hierarchy in mind.

How to design a dashboard with data hierarchy in mind? [3]

Designing a dashboard with the hierarchy in mind brings lots of benefits. People love to listen to stories regardless of the topic. However, they can only follow the story if there’s a logical order of events. A well-organized dashboard will tell a story explaining the causes and consequences of activities impacting performance.

The most straightforward approach enables users to analyze data step-wise, from most general to most detailed. Users should be allowed to pause at the level of granularity they require or can comfortably understand. Simultaneously, the same dashboard can cater to diverse user needs. Executives might require a high-level overview, mid-level managers can delve into more detailed explorations, and analysts can access data-source extracts, sometimes with options for self-service customization. Ideally, users should be able to start their analysis from their preferred level of data granularity.

Allow me to exemplify these ideas. I will show you howe they could work using panels with simplified mock-ups of dashboard designs.

In the first panel, I’ll demonstrate how to structure narrative across multiple dashboard pages. We should **** begin with an opening page featuring a straightforward visual, such as a card or plain chart highlighting one or two key performance indicators (KPIs). Users can then navigate through the dashboard by clicking buttons or visuals, progressively accessing more detailed visualizations, often accompanied by additional metrics. The final "Details" tab typically allows for some level of self-service. Ideally, this would resemble a pivot table, enabling users to manipulate combinations of dimensions and measures freely.

Examples of card visuals that could be used for the launch tab/ page of the dashboard are presented below. Also, here is an interesting article on effectively working with this visual.

Regarding the single-page layout, there are some rules you could follow, as exemplified on the second panel.

My personal favorite is the "Z" principle, which uses the fact that people in most cultures read a page following the Z letter: they start from the left corner, then go to the right, then bottom-left and bottom-right. And that’s how you should order your visuals or data – with the most important (or the one you want the users to read first) in the top-left corner. Another idea is the horizontal layout, which somewhat rests on the same rule as the "Z" principle. It allows users to read data logically (e.g., increasing the detail level from left to right), but it is more handy if you want to show different measures or categories on a single page.

You may also consider the vertical layout if you have more categories to present. In my personal view, choosing this option is not ideal, though it may be suitable for multi-product, segment, or tier businesses. Quite often, such layout stems from a direct request from the dashboard users.

How to underscore data hierarchy with visual aids? [4]

Effective navigation and layout can significantly improve the effectiveness of data exploration in the dashboard. To foster engagement, however, it is vital to differentiate between the elements, enabling users to go from more to less essential visuals.

The weaponry we can use for that purpose is not extensive. However, if it is used thoughtfully, it’s powerful. Our three tools to be used here are:

• size,
• contrast, and
• color.

How do they work? We can manage user attention with size. Typically, a bigger element draws attention first. Contrast works similarly. If something is darker and more transparent, people tend to look at it first. Lastly, use a different, contrasting color to underscore more critical information. Examples of an application are shown on the third panel below.

One thing is essential to make this technique effective. We must stay consistent with the way we highlight more critical elements across the entire dashboard. If you use different methods on different pages of your dashboard, people will not only see the story but will get confused.

Another prerequisite for effectively using the techniques discussed is correctly ordering data (as shown in panel 4). Randomly applying highlights, varying sizes, contrasts, and colors will undermine their effectiveness, which is unfortunate because a well-executed combination of these elements can significantly enhance control over how people read reports or dashboards. You can deliberately guide the viewer’s attention by thoughtfully organizing these visual elements. This approach allows for the strategic presentation of information, including subtleties that might otherwise be overlooked – although one must consider the rationale behind emphasizing such details.

The last tool that aids attention management is highlighting. Sometimes, we need to use that single page to convey information. That could result from the specific needs of the dashboard’s stakeholders. Making some elements highlighted vs. others makes memorizing critical information easier, enabling those interested to get familiarized with the details they need. The worst thing you could do is use the same plain color without highlighting or underscoring. The dashboard becomes dull and challenging to analyze, especially for busy professionals (see panel 5 for reference).

This last function can be effectively implemented with the right BI tool. For instance, PowerBI, as exemplified below, can manage to highlight upon the user clicking on an item across multiple visuals.

Data visualization tips & tricks for dashboard storytelling

Okay. Now that we have designed our dashboard and planned the navigation and page layouts, we can focus on the individual visualizations.

A significant characteristic of compelling visualizations facilitating data-driven storytelling is how quickly users can read the information we want to convey. For that purpose, we may use several techniques: functional titles, relevant visualizations, friendly data presentation & discovery.

Functional titles

What characterizes data-driven storytelling is how it uses titles. Instead of saying "Sales revenue", we may say, "Our revenue dropped/increased because of…" See? Not only do we say what’s in the visualization, but we also suggest what’s interesting about it and why it happens.

This approach may not entirely apply to dashboards, but at least we may make these titles dynamic to improve the user experience.

For instance, we can add dynamic titles to our visualization that change with the critical filter.

Another addition could be adding value, especially if we want to capture the total and some detail level in a single visualization.

Lastly, if we work in an international organization, we can consider adding various title versions according to the language set in the operating system of a given user.

Relevant visualizations

Another tactic for improving user experience significantly is a selection of relevant visualizations.

This tactic rests on the fact that not every type of data will work with every kind of visualization. It may mean that we may (unintentionally, of course) present data confusingly and misleadingly. We don’t have time for that. If we want to stay safe, the answer is one: for most needs, we need only three visualization types: bar chart, line chart, and pie chart.

Bar charts are excellent for comparing aggregated data (e.g., annual values or product totals). The line chart is suitable for displaying progress and catching any unexpected (or expected) trend changes. Pie charts are suitable for displaying data distribution.

What if we must present precise values or display multiple scenarios on a dashboard? The solution is straightforward: use a table. Tables are practical storytelling tools when used correctly. For instance, prioritizing data by placing it in the first column or row helps organize information effectively. By applying color coding (see next paragraph), we can highlight more critical information. Additionally, incorporating so-called traffic lights in our table allows for quick interpretation and adds a decorative element, enhancing functionality and visual appeal.

If you want to learn more about visualization tips & tricks, I encourage you to see my other post from last year.

How to Talk About Data and Analysis to Non-Data People

Friendly data presentation & discovery

I want to introduce here the 3×3 rules for dashboard design:

• Three actions – the desired number of actions users should perform to get to the required information.
• Three filters – the maximum number of configurations (lists, drop, or select boxes) to be included on a single page.
• Three digits – the maximum to present any number (i.e., instead of $3,341,456.01, we should show $3.34 million.

How do these guidelines apply in practice? Take the three-action rule, for example. A user logs into the dashboard, selects the desired page or tab, and applies a preferred set of filters, ideally loading these filters from saved settings. This setup should be streamlined, assuming the filter combination can be pre-loaded or is user-generated. This rule pairs with another principle involving the use of three filters. Ideally, one of these filters should span multiple visualizations and pages, which removes the need for users to reapply the filter each time they change pages. This unified filtering enhances both usability and efficiency.

Another strategic tool used in data-driven storytelling is color coding. I wrote an extensive post on that subject last year.

How to Use Color in Data Visualizations

Regarding the use of color in dashboard design, I recommend three strategies for effective implementation. They might be grouped under a single "3xC" rule.

First, color usage must be clear. This means that if colors differentiate elements, it should be immediately apparent which item is emphasized and which is not. Subtle shade variations, driven purely by aesthetic preferences, will not work. Second, the approach must be consistent. For instance, if a color is chosen to show a positive change, this should be applied uniformly across the dashboard. Lastly, color use should be conservative. This involves using the fewest colors necessary to achieve the desired impact. Ideally, no more than three or four different colors should be used. For example, consider using black for the basic dashboard font, grey for basic visualization fills (such as bars in a bar chart), green to highlight positive changes, and red to indicate negative changes.

Conclusions

In this post, I showed how to incorporate a few storytelling elements to enhance dashboards’ functionality and user engagement.

While traditionally dashboards were not designed as storytelling tools, integrating narrative techniques can transform them into more effective communication and decision-support tools. By carefully designing navigation, layout, and visual hierarchy, and by employing rules such as 3×3 and 3xC, dashboards can provide not only plain data but also contextual insights. These enhancements can help users navigate complex information landscapes more intuitively, making the dashboards not only more engaging but also more aligned with varied user needs across different levels of an organization.

Ultimately, the successful fusion of storytelling with data visualization in dashboards can lead to greater clarity, increased stakeholder engagement, and more informed decision-making in fast-paced business environments.

---

This post was drafted using Microsoft Word, and the spelling and grammar were checked with Grammarly; I reviewed and adjusted any modifications to reflect my intended message accurately. All other uses of AI (image and sample data generation) were disclosed directly in the text.

---

Did you like this post?

Consider a subscription to get notified about my new stories, follow me, or leave a 👏 .

---

References:

[1] Brent Dykes, The Real Reason Most Dashboards Don’t Tell Data Stories

[2] Stu Eddins, Most Analytics Dashboards Fail. Will Yours?

[3] Kobe W., How To Structure Visuals On Dashboards?

[4] Kamil Moon, Creating a visual hierarchy in dashboard.

[5] Zach Gemignani, How to Apply Data Storytelling to Dashboards.

[6] Isabelle Bittar, Next-Level Dashboard Design With Power BI’s New Card Visual With Reference Labels