I am currently working on a project about energy consumption dashboards, and I’ve been experimenting with adding more interactivity like filters, hover effects, and drill-downs by region and energy type.
At first, it felt like these features made the dashboard more engaging and flexible, but now I’m starting to wonder if I might be overdoing it. Too many options can make users lose focus or struggle to interpret what’s most important.
So I wanted to ask, when does interactivity stop being useful and start hurting comprehension? And are there any guidelines or studies that help decide how much interactivity is “too much”?
Would love to hear your thoughts and how others balance this, especially when the goal is to make the visualization both exploratory and clear.
Great question. In my experience, this is a good time to take a step back and work on your user stories, eg. who are they (sketch up a few personas/preferably behavioral archetypes). This is a prerequisite to user stories. Then ask what tasks they would want to complete on your dashboard, and make sure that that is possible. For this I recommend looking up ‘user stories’ within agile/software development. Ps. you will have a lot of user stories to work with. This is where you seemingly complex dashboard might actually start to make sense.
When taking the user perspective, more things will make sense, and perhaps you will find that some functions or features are redundant, because they serve noone. And then you have some answers to your question.
Another approach would be to do a quick and dirty ‘desirability test’ or a ‘think out loud’ with anyone. You can always go deeper into more controlled user-involved processes, but just to get you started.
If you still feel that your dashboard is too crowded, start using all of the graphic design design principles, and embrace ‘white space’ to help the eyes navigate a bit.
Sounds like an interesting project, you have. Enjoy the proces!
This is a great question. My short answer is that any interactive or visual feature should directly answer a user need. It is always tempting to add more filters, drill-downs, or hover effects, but more features than necessary can also overwhelm people. A bit of quick piloting or interviewing with users can already show you which interactions really help and which ones distract. Prioritising what matters is key.
There is also the option of progressive complexity, where the basic view stays clean and the more advanced interactions are offered on demand. This keeps the interface focused without removing functionality.
@Alexandra suggested giving you pointers to our prior work on interaction: https://hal.science/hal-02197062/document
One way to think about “how much interactivity is enough” is in terms of the degree of flexibility, meaning how many and what types of actions users are allowed to perform in your system. In our theoretical paper, we organised actions into seven categories: input, processing, mapping, presentation, meta, social, and interface actions (see their descriptions in Section 6). This can be a practical way to organise your features and check whether each one serves a clear purpose for your users.