Database Reference
In-Depth Information
a retail organization analyzing poor sales in a region may want to drill down
to a store that's performing particularly poorly, check the store's performance
trend over the last year, and bring in a list of any marketing campaigns run in
that store for that period.
numBer oF dATA PoInTs In A VIsuAlIzATIon
The suggested number of data points in a visualization is four to seven. Of
course, this doesn't mean that you should only have seven dates in a line
chart; instead the limit is suggested so that you only have a maximum of
seven series plotted over time. This limit applies to any kind of visualiza-
tion and is based upon the human ability to keep only a certain number
of items in focus at any given time.
Be careful not to confuse this data discovery portion of glance and go with
data exploration!
The key difference between the data discovery based on glance-and-go visu-
alizations is the intent. In glance and go, you know what question to ask—for
instance, “Have sales targets been reached?”—and typically which follow-up
questions need to be asked, such as the following:
uu
Which stores missed their sales targets?
uu
Which product categories are performing poorly?
uu
What marketing campaigns were run in that period?
In data exploration, the questions are unknown, and the visualization is
explored until a pattern is discovered. The patterns to be discovered may be
correlations such as similar products being bought together or seasonality
of purchasing patterns.
uSIng ColoR In VISuAlIzATIonS
Preparatory to discussing the use of color in visualization, it's important to
understand the way color is perceived in the human brain. Despite the popu-
lar education of red/green/yellow as primary colors, the three primary colors
in the human eye, and not coincidentally, in the computer monitor are red/
