As a part of The Data School training, we worked on 8 different client projects. These projects usually start on Monday with an intro session with the client, and end on Friday, with the presentation of our findings and dashboards. Five days is not much time, and it’s important to make sure that the final dashboard is correct. To save time (and mental energy) at the end of a project, I complied the dashboard checklist below.
Are your measures displayed in the correct format (i.e., currency symbols, units)?
Do you use the correct aggregation? Is the aggregation clear for the user?
Are your measures clearly labeled on the chart?
Do they show correct measure and aggregation?
Are they visible?
Are they properly labelled? Are there too many (or not enough) labels?
Are they fixed? Should they be fixed?
3. Syncronised dual axis
If you are using a dual axis chart, make sure to synchronise the axis to display the data at the correct scale.
4. Numbers make sense
Check that the numbers make sense: i.e. add up to 100% where they should, reasonably large or small depending on your data.
Did you set up the action filters correctly? Do they connect correct worksheets?
For quick filters on the dashboard:
Does the user know which filters apply to which charts? Do they see what’s shown in the chart now and which filters have been applied to it?
Does the top X filter shows the correct top x dimensions?
Data source filters:
Do you need to use them? If yes, check that they are setup correctly.
Do they help the user to understand the chart?
Do they provide additional information that is helpful?
Are they formatted properly?
Did you exclude command buttons option?
Did you use viz in tooltip in a proper way (not too big, or detailed)?
Check that your fonts are large enough (but not too large!), legible and have enough contrast. Think about the use case for the dashboard, and which screen it’s going to be viewed on.
Are there are too many colours on the dashboard?
Are your colours too bright?
Is it clear what each colour represents? Did you include colour legends?
Is there one colour representing different things on the same chart / dashboard?
Does the colour reflect correct direction (higher/lower value)?
Is your colour palette colour friendly?
Do the colours have enough contrast and easily distinguishable?
9. Marks size
Is it easy to understand what the size represents? Is there a size legend?
Is it easy to distinguish between larger and smaller marks on the chart? Are small marks large enough to see?
Does the size reflect correct measure or dimension?
Are there larger marks that overlap smaller ones?
Does the size reflect correct direction (higher/lower value)?
Are there too many different shapes on the chart?
Is it clear what each shape represents? Is there a shape legend?
Is it easy to distinguish between different shapes?
11. Double encoding
Make sure that you are not visualising the same metric twice on the same chart, for example displaying the size of sales on a scatter plot with the size of a circle as well as the intensity of the colour of the circle.
12. Titles & subtitles
Check that your titles and sub-titles are clear and help to understand the charts. Make your titles and subtitles dynamic, where possible, to reflect the filter selection.
13. Instructions on how to use the dashboard
Sometimes navigation or additional functionality of a dashboard is not intuitive to a user, so make sure to include any additional instructions or directions to individual charts or at the top of the dashboard. Information buttons with instructions in the tooltip can be a good idea. Another option is to add an overlay instructions layer as described in this post by Marc Reid.
14. Data source reference
Make sure to include the source of your data and the link to the website where you got it (if applicable) at the bottom of the dashboard.
If you are using third-party images or icons in your dashboard, double-check the conditions of using these media, and include references to the original source. Do not include images of other people, logos or trademarks to your dashboard unless you have an explicit permission to do so.
Remember to add your name at the bottom of the dashboard if you publish it to Tableau Public or any other public website or platform.
16. Spelling mistakes
The last, but definitely not the least point. Take a break from the dashboard, and look at it with fresh eyes to make sure that you are not publishing a ‘Stales overview, 2019-2020’ dashboard.
I hope this list will help to keep your dashboards error-free. Do you have more points to add to this list? Let me know and I’ll keep this post updated.
Illustration by Storiset