5 Data Tooltip MUSTS
We’ve had enough of you, default data tooltips!
I’ve seen a lot of tooltips that look like this floating around in production lately (or some variation if you’re using a different tool).
Using a couple simple tricks, it takes about 2 minutes to clean this up to a more polished version that will not only improve usability, but also make your users think you spent more time refining the aesthetics than you actually did.
There are certainly a lot more tricks and considerations when it comes to tooltips, but since they would take more time, we’ll focus on those later.
For now, here are 5 super quick and super simple hacks to immediately improve your tooltips — absolutely no new skills or additional time required.
1. Make it readable.
I prefer tooltips that read like a sentence and I also prefer whole numbers, but that style is not always appropriate depending on the dashboard. But if it is, it could look something like:
A field-based list can also work with clear field names, logical ordering of the fields that reads somewhat intuitively, and tighter alignment/spacing for better readability. After all, people will tend to read the tooltip like they would read a sentence.
Consider the following auto-generated default tooltip example:
Here is one revised version:
Or even:
There are a few ways you can do this to allow for some flexibility, but it is already looking better.
2. Opt for “on hover” over “responsive” tooltips.
This means a user will need to deliberately hover over something for the tooltip to appear, rather than tooltips “responsively” popping up left and right like whack-a-mole as you’re just navigating across the dashboard. I will caveat that some people may prefer responsive because it is a hair faster. That might be fine for dashboards that don’t have a lot of tooltips that could appear. The biggest issue is trying to navigate across a dashboard and having numerous tooltips appear simply because you traversed their airfield in passing.
3. Disable the Command Bar.
In Tableau, this is the Keep Only, Exclude, and the other “techy-looking” buttons that no user of mine has ever wanted to futz with (particularly the Set and Group buttons). It ends up cluttering the tooltip and makes it look more like a development tool than a refined product. If your audience isn’t analyst or technical folk, they’ll be none the wiser and it will look much more polished and approachable. There are also other ways you can add this functionality if needed. I’ve found that [in general], the fewer things a user can accidentally filter, restrict, modify, or change, or limit, the better because it can be challenging for users to back out of them without also losing your deliberate selections.
4. Rename or customize your field names.
This should really be standard practice. We often have to add a lot of different and/or duplicates of fields into the dashboard view, so they make their way to the default tooltip. While you and I know what the field [avgCountrySales_notfixed_tablecalc (copy)] is, our users don’t and — to be honest — you and I will forget by the next time we open it up 2 months later. Think of this as an additional semantic layer and have it read like a descriptive label that you, your users, and future developers can understand what it represents. Clean up any field names that aren’t intuitive (i.e. “distinct count of Csid” is the “number of unique patients”).
5. Remove extraneous fields/information.
Again, we often have to add a lot of different and/or duplicates of fields into the dashboard view, so they make their way to the default tooltip. We don’t need every bit of descriptive data in the tooltip, especially since a lot of it may already be visible on the initial data view. Users will typically hover over your tooltip to get more detailed information about the data point of interest— only include info that is valuable and helps to further inform answers with a bit of more granular or detailed information (note: I will touch on graphical/embedded viz tooltips in a different post because they take more time and have their own considerations).
Okay, here is a 6th bonus option: when in doubt (and when appropriate), disable the tooltip altogether. There are some cases when this might work: there is no additional information to drill down to, the viz is small or self-explanatory, there is limited monitor real estate, you’re trying to guide them to the next story slide/data view rather than providing more information all on one screen, etc.
As analysts and technical builders, we tend to want to include as much information as possible so our users can make informed decisions. But that can be counterproductive and often cause our users to drown in information, not knowing where to start. Every added bit of information or decision that a user has to make or decipher while exploring our dashboard contributes to perception of complexity. We have a duty to make that as easy as possible while understanding what will inspire the right questions, how our tools can be designed to help guide (but not pigeonhole) users to the answers they need, and what belongs on prime time or ancillary real estate.
Your stylistic preferences, particular use case, userbase, and end goal for the dashboard will influence how — or if — you approach some of these suggestions. They may not all be appropriate for every dashboard. But try some on your next dashboard. Your users will thank you and you will probably spend less time explaining what things mean.
Disclaimer: I use Tableau examples because it is my primary and preferred dashboarding tool (and for most of colleagues’ and clients’ as well), but the concepts are universally applicable.
