Data Visualization at Riskified

Introduction

This guide provides a visual outline for data visualization practices at Riskified. It will walk you through the most common chart types and will give you instructions on how to make your chart’s appearance conform to Riskified’s design language.

General Guidelines

Keep it simple and avoid decoration.

What goals do you hope to accomplish with the data visualization? Try to emphasize important information and eliminate unnecessary data points in order to communicate your main idea as clearly as possible.

Do's and Don'ts

1. Don't overload the chart

Adding too much information to a single chart eliminates the advantages of processing data visually; try removing or splitting up data points, simplifying colors.

2. Don't use 3D

The projection of 3D objects into two dimensions distorts the data.

4. Connect data points with straight lines

In case of linear representations, polygonal shapes are preferable to curvy lines.

3. Avoid dual-axis charts

The scales of dual axis charts are arbitrary and can therefore mislead readers about the relationship between the different data series.

5. Grid

A grid should support the readability of a chart, but not distract from the data. Color the grid lines in a light grey and, when appropriate, remove vertical lines.

6. Don't truncate bar charts

The numerical axis needs to start at 0 to avoid distortion. Our eyes are sensitive to the area of bars on a chart. If those bars are truncated, the viewer might draw the wrong conclusions.

7. Zoom into relevant range when appropriate

A line chart reveals trends or changes over time rather then drawing attention to the absolute magnitudes of it's values.
Zoom into the relevant range to better emphasize these changes.

Typography

Google Sheets Text Styles

Chart Title

Roboto Bold, 20pt

Legend

Roboto Regular, 14pt

Chart Title

Roboto Regular, 12pt

Numbers/ Categories

Roboto Regular, 12pt

Lables

Roboto Regular, 10pt

Colors

Qualitative Comparism

Principle Colors

Red
#DF375F
Teal
#217883
Aqua
#7EDFEC
Purple
#BA97FF
Blue
#5A4CFF
Green
#00BEA6

Extended Colors

Additional categorial color options - to be used in a moderate amount, only when needed

Dark Blue
#283593
Dark Purple
#7F1FF7
Tan
#D07A5F
Yellow
#F8DE81

Color Attributes

Approved
#00BEA6
Declined
#DF375F
Chargeback
#BA97FF
Canceled
#B9C2CA

Qualitative Comparism

Diverging Palette

Data that progresses outward from a middle value should be represented by a diverging color scheme.

Click on a color swatch to reveal it's HEX code.

#7F1FF7
#A064FC
#BA97FF
#D8C4FF
#EEE5FF
#D4F9F2
#A5EFDD
#6CDAC6
#00BEA6
#1E9F9A
Neutral A
#7F1FF7
Neutral B
#1E9F9A
#DF375F
#F56882
#FF9CAF
#FFD2DC
#FFECED
#D4F9F2
#A5EFDD
#6CDAC6
#00BEA6
#1E9F9A
Negative
#DF375F
Positive
#4C3EFF
#DF375F
#F56882
#FF9CAF
#FFD2DC
#FFECED
#F1F5FF
#C5D6FF
#97B3FF
#637FFF
#5A4CFF
Negative
#DF375F
Positive
#5A4CFF

Monochromatic Palette

The monochromatic palettes are good for showing heat maps and ranged values

Click on a color swatch to reveal it's HEX code.

#D4F9F2
#A5EFDD
#6CDAC6
#00BEA6
#1E9F9A
#217883
#00525B
#002B33
Light Green
#D4F9F2
Dark Green
#002B33
#F1F5FF
#C5D6FF
#97B3FF
#637FFF
#5A4CFF
#283593
#0F1862
#080E3F
Light Blue
#F1F5FF
Dark Blue
#080E3F
#EEE5FF
#D8C4FF
#BA97FF
#A064FC
#7F1FF7
#4D179F
#36136B
#1F0547
Light Purple
#EEE5FF
Dark Purple
#1F0547
#FFECED
#FFD2DC
#FF9CAF
#F56882
#DF375F
#B72358
#6D0C4D
#440035
Light Red
#FFECED
Dark Red
#440035
#F3F5F7
#E7EAEF
#CFD6DD
#B9C2CA
#9EA6AE
#868E96
#595D69
#000000
Light Grey
#F3F5F7
Black
#016249
#D1F7FA
#7EDFEC
#27C1D1
Light Aqua
#D1F7FA
Dark Aqua
#27C1D1

Chart Models

The charts below are a curated set of data visualizations for common needs at Riskified. Click on a chart to view it’s underlying data and setup in Google Sheets.

Line Chart

A line chart reveals trends or changes over time and can be used to show and compare between different categories of data. You should use it when you chart sequential data sets, for example the Riskified approval rate vs SLA approval rate.

Keep in mind: Avoid dual axis charts. The scales of dual axis charts are arbitrary and can therefore mislead readers about the relationship between the different data series.

Column Chart

A column chart is used to show a comparison among different items, or of items over time. Use vertical column charts when you’re graphing sequential data sets, for example recovery rates among bank-declined orders.

The numerical axis needs to start at 0 to avoid distortion. Our eyes are sensitive to the area of bars on a chart. If those bars are truncated, the viewer might draw the wrong conclusions.

Bar Chart

Use horizontal bar charts when you’re graphing nominal variables. Sort nominal variables from greatest to least or from least to greatest.

Stacked Bar Chart

A stacked bar graph is used to show how a larger category is divided into smaller categories and what the relationship of each part to the total amount is. For example, use a stacked bar chart to show how many chargeback disputes Riskified has won, relative to the total chargeback population.

Pie Chart

Pie charts are used to show parts of a whole. A pie chart represents numbers in percentages, the total sum of all the divided segments must equal 100%.

Keep in mind that here should not be more than five slices and that the largest slice should start at the top and fall to the right.

Map

Cartography is used to display geographical data. Use a map to show, for example the world-wide distribution of orders by billing country.

How-to: Google Slides

This chapter is providing you with instructions on how to stylize charts in google slides.

Customize colors

...

Text styles

...

Grid lines

...

Highlight single data points

...

Additional materials and guidance can be found at the links below.