Ask HN: How can I make a dashboard that doesn't suck?

It is a question about useful information. What problem are you trying to solve? What is useful to the user? And: is the information actionable?

Information that is just displayed, or something to be informed about does not necessarily need a dashboard. This is called reporting.

Something that screams "Action!" is useful as a dashboard. It should help guiding the decision making process.

You might also consider automation. If a machine can do what a dashboard asks for, an action step, then you actually don't need a dashboard instead you can do reporting ("X steps done during Y period").

Build something useful and have fun!

Read Edward Tufte.

Checkout paid themes from wrapbootstrap or themeforest or similar sites.

Think about what different user groups there are, whether you'll be making 1 dashboard for all or different dashboards for each. Then delve into what kind of questions they want answered (maybe also figure out priority?), and then use that to inform the information architecture for the first version of your dashboard. Time permitting, you could just make a mockup in Sketch and do user interviews where they go through it and think out loud to tell you what they're looking for, what they see, and you can use that to iterate and re-design where appropriate.

Observe your users and figure out what they look at the most. Take the top 3 - 6 actions / data and make them viewable / easily accessible on the dashboard.

Get to know the user for it, ask them what they want (granted that isn't always the best thing to design exactly as they want it but it helps to know), what they want to accomplish, any pain points.

If possible (depends on use case / user) give them some level of customization, moving widgets, filters.

Use a free theme from uifort.com

> The users don't seem to mind what's on it.

Why are you going to build a dashboard if the users don't seem to need a dashboard?

I would strongly suggest perusing through the resources of Stephen Few over at Perceptual Edge.

Library: http://www.perceptualedge.com/library.php

You could do a lot worse than becoming familiar with Few's "Information Dashboard Design" book, and the Perceptual Edge blog contains a wealth of material, some of which ended up in his subsequent books.

Older posts on his blog also include reviews of different author's books, often having to do with cognition or analysis related topics which are also informative. Seems he has retired from his consultant business and now blogs at: http://www.stephen-few.com

Here's one of his whitepapers you may find helpful: https://www.perceptualedge.com/articles/Whitepapers/Formatti...

I believe it helps to define how "doesn't suck" is evaluated. Is it visual? UX?

One suggestion is to use something like hotjar to see how users use the dashboard (once you get something up and running), and analyze what they look for, how they use it, and what they spend a lot of time doing. This might give some ideas on how to make it easier for them to do their jobs. IMO that's the main purpose of a dashboard, so starting there makes sense to me.