A good data visualisation is an accessible data visualisation. The more accessible a visualisation is, the more likely it is that the visualisation will convey the message of interest. For that reason, before you share your data visualisation with the world, it is always good to make sure that you comply with accessibility guidelines.
The best guidelines I have encountered so far are from the Government Analysis Function website in the UK (based at the Office for National Statistics). For accessible charts, they have even created a checklist of the basics.
As general recommendations, among other things, it is advised to use a maximum of ten light grey gridlines, right align values on the y-axis, and that charts should match the width of the text around them (the latter is something I often see papers in academic journals get wrong).
For bar charts, they recommend – among other things – to rank bars by value, that gaps between bars should be narrower than the width of a single bar, and to avoid stacked bar charts when categories have negative values. For other chart types, they recommend things such as to aim for a maximum of four lines (for line charts) and to rank categories by size with the first at the 12 o’clock position (for pie charts).
You will find a lot of good recommendations in the material. There is also an R package called {afcharts}, which tries to implement a lot of the guidelines. As always, it is challenging to make good guidelines work as a default, but it is also worth checking out.