Comment pouvez-vous améliorer les visualisations grâce à l’interactivité ?

For me interactivity helps to reduce lengthy dashboards and makes the experience for users worthwhile. I make use of filters, slicers, bookmarks and so many other features to enhance my dashboards and engaging so users can make decisions and drill down without my help. Most users have this questions prepared and through KPI’s this interactivity can be maxmized. For example one of the KPI’s is date by month so you have a date column, create a calculated column for this date column and you put a slicer for the calculated column which the user can use to discover more insights.

I believe that static dashboards are useless. They assume that the dashboard has what the consumer will need, but this is rarely the case. There are always multiple ways to slice and dice data, and static dashboards cannot accommodate all of these possibilities. Interactive dashboards, on the other hand, are not limited in this way. They allow users to explore the data in any way that they want. This makes them a much more powerful tool for data analysis and decision-making. For example, let's say you have a dashboard that shows sales data for your company. A static dashboard might show you a line chart of sales over time, but it wouldn't allow you to drill down into the data to see sales by product, region, or sales rep.

In a pivotal project, I embraced interactivity in data visualization, experiencing its transformative power. Going beyond static graphs, I found that interactivity enhanced user experiences. Enabling filtering helped users unearth insights, while zooming and drilling down engaged them deeply. This journey turned passive viewers into active participants, leaving lasting impressions. Moreover, interactivity showcased my data storytelling skills, requiring creativity and attention to detail. Each interactive element became a deliberate choice, reflecting my grasp of the data and audience needs. This firsthand encounter revealed that interactivity unlocks data's true potential, bridging understanding gaps, and creating memorable experiences.

In my experience, interactive dashboards reduces many redundant reports. Furthermore you can present your data at multiple granularities so both technical & executive users can consume the same dashboard. It also aids in quick decision making as cross filter or cross highlighting provides correlation etc.

While thinking about visualisation of data, it is important to structure the flow in order to tell a story. Business leaders need to go through many dashboards with tons of data daily. It is very easy to miss out on critical information, or mis-interpret it. This is where interactive dashboards make things interesting and user's engagement with the data helps with their interpretation and understanding. It is crucial to also highlight the key critical insights upfront to the end-users with a guided approach to point them to the most actionable areas of the dashboard, and this can be personalised to the user's needs and preferences.

If you spend some time finishing off the visualization with formatting and pay attention to ensuring you add filters that can work across multiple sheets and panes, you can save your users time and draw their eye to key metrics. Adding Tooltips and multiple highlight actions can also make working with the vis much mote intuitive.

I want to share my perspective about adding interactivity using Tableau. Tableau has certain features which lets the user interact with the charts in the playful way. Some of the features are: 1. Navigation button - helps you navigate between different dashboards in a workbook. 2. Show/hide button - helps you to enable/disable a particular item inside a container 3. Actions - you can filter, highlight and even use sets and parameter to add interactivity. 4. Dynamic Control Visibilty - game changer feature in Tableau. Based on a parameter, we can control any object visibility in Tableau.

One more, popular solution is Looker Studio. As I know Power BI and Tableau has more features than Looker Studio, but the last one is much easier to learn and it has many native data connectors. It also has many interactivity options, like filters, sorting, additional metrics for sorting, comparison visualizations, page level and report level calendars and data sources.

There are loads of good visualisation tools - but most companies have Excel in place, so that can often be a good first place to start - especially if you need to prove concept for analysis.

In my experience, we also use dynamic titles based on filters. It comes in handy while interacting with the dashboard. For example, when you change the filters, the line chart or bar chart title keeps changing based on the filter value and return to default-all values view when no filter is used.

When diving into interactive visualizations, it's crucial to keep user-friendly design principles in mind. Start by pinpointing the purpose and the people you're targeting. Knowing this helps you decide what type of interactive features to use. Don't go overboard with interactivity – only include it where it truly adds value, clarity, or excitement. Lastly, help out users who might not get the hang of the data right away. You can do this by providing clear instructions, using icons, labels, and colors to guide them through the visualization. Just remember: user experience is key!

Plan the user's journey through the visualization. Understand how users will interact with it, where they will start, what actions they might take, and what insights they should gain.

Interactivity is most effective when it follows and anticipates the end users' curiosity. Start from your initial data view, list which user questions might be generated from it, then prioritize each based on their needs or workflow. Interactive elements that can answer these question add user value; those that don't may cause distraction instead.

Once you've built your interactive visualization, it's crucial to make sure it works well and achieves what you wanted. You can use special tools to find and fix any mistakes or issues in your code or data that might affect how your visualization works. This helps you see if there are any problems or things that can be made better in terms of how easy it is to use, how fast it is, how much people like it, and how well people with different abilities can use it. Plus, there are tools that help you gather and study data about how people look at, share, or talk about your visualization. This helps you understand how much your visualization affects the people who see it and whether it achieves the goals you had in mind.

Test your visualization with individuals and small groups' before scaling it to the larger organization. Beyond simply testing data accuracy, ask users to explain how they are interpreting the data and what initial conclusions they draw. If they're coming to different interpretations than intended, address and re-test. Data misinterpretation causes as many headaches as data inaccuracy. Catching these issues ahead of organization-wide rollout will save you time and minimize the publicity of tweaks and rollbacks.

Another element that I have found useful is to understand how the users will access your visualizations. Many users would also be using their mobile devices which demands that the visualizations and their legends/filters / etc.. are all responsive to your design. Many tools do offer options to tailor your entire dashboard to how it should be displayed on a mobile device which is great. Along with this comes the Accessibility aspect. Make sure the interactive features are accessible to all users, including those with disabilities, by providing keyboard shortcuts, alt text for visuals, and other accessibility features.

Most studied and verified factors that are true for media usage (such as websites) are also true for data visualization. Take into account ease of use, viewing comfort and speedy, clear discovery of information. Often times less is indeed more.