Everything You Need

for Clear and Efficient

Data Visualization

A 4-Part Series on Data Visualizations

Data visualization is the discipline of communicating
01
Dashboard Planning & Design
In part 1 of our data visualization series, we
outline the steps needed to plan and design
proper dashboards. While each data dashboard
02
Choosing Appropriate Visualizations
In part 2 of our data visualization series, we begin
to choose the appropriate visualizations for the
data we want to display. Data visualization are
information clearly and efficiently using graphs, plots, lines, has its own requirements, limitations, and goals, intended to be more than mere eye candy – they
there are certain guidelines that are almost should serve a specific purpose and convey
bars, and other tools. Data visualization can be seen as always relevant for dashboard creation. Follow specific facts in a more effective way than the
both an art and a science, and in today’s world of internet this guide to get started. basic tabular format. Read this part to understand
which visualization will be more effective for the
and Big Data, being able to visualize your data in the right types of information you are trying to relay.

03 04
way is becoming increasingly important.
How to Assemble Final Dashboard Design Worksheets
Dashboards, Design Formats, & Checklists
This 4-part series will show you how to clearly and and Common Pitfalls In part 4 of our data visualization series, we
efficiently visualize your data. Start with the planning, In part 3 of our data visualization series, we include worksheets and checklist for initial
choose the right visualizations, assemble your dashboards, start to structure visualizations with the proper planning, visualization review, structure, and final
formatting. Follow these points to make sure your checklist. These sheets will ensure consistency
and then use the checklists to make sure you haven’t visualizations are effective and understood. Don’t and completeness of your visualizations, and it’s
overlooked anything before publishing your work. miss the common pitfalls section and make sure a great way to start your visualizations off on the
to avoid them in your visualizations. right foot.
Dashboard Planning
and Design
#1 in a Series of 4

Principles For Creating Effective Dashboards What are the key rules for creating such a dashboard?
Clear
Design with purpose - whether your dashboard is for The business user should be able to understand the meaning of what is being
displayed without much interpretation.
internal or external (embedded) users. Every design
choice should contribute to the goal of the dashboard. Easy to use
Business users should be able to locate filters and other functions without assistance.
‘How’ can we tell the story of our data, ‘what’ is the
Purposeful
relevant information we need to present and ‘why’ is the Every object should have a reason for being available to the business user. It should
add visual appeal, serve a data function, or enable exploration and navigation.
information we visualize relevant to tell this data story.
Aesthetically pleasing
All these are key questions that need to accompany us Inviting, abstract, and representative of the designer’s style.
while creating a dashboard in order to create a good, Correct and consistent. With a visual tool, business users will tend to dwell on
display issues, typos, irregularities or inconsistent formatting, even if the results are
clear, readable message. correct. Although the issues might be known by the developer to be minor display
considerations, they will affect the user’s perception and adoption of the dashboard
and the insights.
Basic design pointers
Keep it simple. Eliminate lines, labels, axes, charts and any extraneous components that
Dashboard Design
are not needed to understand the data story. Extra elements clutter the design and The goal is to use visualizations that meet the
distract the user.
business requirement.
Create a visual hierarchy. Text and graphical elements should use size, color, and other While you may have an in-house brand, this does not mean that the same
styling options to give the essentials a ‘heavier’ appearance to draw the eye. By making visualizations should be present on all worksheets. When designing a dashboard, the
design choices to highlight critical elements, the dashboard designer creates balance developer should consider:
and guides the reader to each point that serves the visual’s purpose.
Who is the audience?
To create an easily readable dashboard, you should also keep in mind the basics:
make sure you use the same color palette, font type, colors and size throughout the What are they interested in knowing?
entire dashboard. You should not use more that one font type and 2-4 font size.
What is the best visualization for communicating the information?
Use these same design principles even if you are white-labeling your embedded
analytics for external users.
Create a dashboard that is simple in its design with the goal of highlighting the
Use meaningful metrics. Be certain that the audience is familiar with metrics, so they
information that supports its message.
can correctly interpret them within the context of the dashboard. The data should also
be verified so that users can be confident in its validity. Generally, the eye travels from the upper left corner to the lower right corner.
Relevant components of a dashboard should be arranged accordingly. The designer
Plan for maintenance. Whether it is data refreshes, functionality updates, or the addition
should also strive to achieve balance across the dashboard so that no one area has too
of new users, planning for upkeep should inform data and dashboard design choices.
much weight and causes the aesthetic to become unappealing. Notably, both symmetric
If it is too difficult to maintain, a dashboard may be too cumbersome and not worth
and asymmetric dashboards can be well balanced.
the investment.
The aim should be to fit the worksheets on the business user’s screen, without their
Limiting the scope: attempting to accomplish too many goals in a single view can lead
needing to use any navigation ribbons to see any part of a worksheet or a worksheet’s
to information overload when there is too much information for the audience to
data. If this is not possible with the existing layout, it may be necessary to rethink the
process effectively.
structure of the dashboard, perhaps by using more filters or by splitting the worksheets
across more dashboards.

If Key Performance Indicators (KPIs) are available, it is sometimes best practice to

present them at the top, rather than the business user first seeing the details.
It is often a good idea to start with a design on a sheet of paper the size of the
dashboard and map out the structure. The following are some typical layouts:
Dashboard Sizing
For dashboards used internally, it is advisable to set the dashboard size to the
standards set by your business. There should ideally be a style guide that prescribes
layout, format, etc.

For embedded visualizations, the size of the canvas should be determined with your
business user depending on their requirements.

There can be a temptation to cram everything onto one large dashboard. However, bear
in mind that Sisense provides multiple means to allow the consumer to explore their
data down to the row level, such as:

Drilling

Filtering

Jump to dashboards with a greater level of detail

Choosing Appropriate
Visualizations
#2 in a Series of 4

Not Just Another Visualization

White Paper
This isn’t just another white paper about visualizing data
or how a BI company thinks data should be visualized.
We’re offering a unique perspective using real-world data experts who use BI tools
every day to deliver meaningful business data to busy executives at
companies just like yours.

In this second part of our four-part series, we share the collective wisdom of
thousands of data experts. The ones who are in the trenches (or offices) using
Sisense widgets to display their business data. The widgets they most commonly
use are selected for their ability to visually display meaningful business information,
not to create pretty pictures (although sometimes it’s a side-effect).
The Top Data Visualization Widgets
(and why our customers are using them)

As we onboard our customers, we spend a great deal of time understanding their

needs and guiding them towards the right visualizations for their data using the best
practices listed in this paper. Our goal is to help them achieve the best possible insights
from their data by using the right widgets.

As a data company, and in order to achieve the above, we look at how our customers
are using dashboards and widgets. Here are 2 column charts and a bar chart illustrating
the top widgets our customers are using:
 A Dashboard for Every Goal
Rather than selecting visualizations at random or
inconsistently across dashboards, the goal should use
standard visualizations for specific requirements, unless
there is a compelling reason to diverge from them.

KPIs – Indicators and Gauges

Numeric Indicator - Single Value
Indicators are numeric values or gauges that are good for visualizing key performance
indicators (KPIs) or important business information. Indicators visualize a single value
such as a company’s total annual revenue-to-date and can be supplemented with a
second value and title that enhances the data, such as the percentage of a target value
As you can see, there are seven widgets that are used most and that make up most of that has been reached or the same value for the previous year. Indicators showing KPIs
our customer’s dashboards. This points to the same widget being used multiple times should be refreshed as frequently as possible.
to represent different forms of data. Armed with this information, we will dive into the
best practices for your visualizations focusing on the specific requirements (KPIs, Indicators are used to show a single significant value.
comparisons, individual data points, time comparisons, etc.) and not on the
It is important that the value should have sufficient
widget itself.
context to be meaningful to the dashboard consumer
This will help you find the right widget to reflect the data goal (even if it is used in
Often not necessary to add a title
multiples on the same dashboard).
Numeric Indicator - Two Values Often not necessary to add a title.

Indicators are used to show two significant values. Consider using conditional formatting, e.g., set the gauge to be the color red for
an indicator that is below target.
The values should be related.

It is important that the value should have sufficient

context to be meaningful to the dashboard consumer.

Often not necessary to add a title.

Best Practices:

Numeric indicators are the most easily read visualization for KPIs

Gauges are helpful for indicating the “health” of a value or seeing a value as a
percentage of a target value.

Enhance gauges with color coding to indicate KPIs that are good, bad or may
require attention

Use widgets for only the most critical information. Overuse clutters dashboards
and makes them difficult to understand or use.

Gauges

Indicators are used to show one or two important metrics.

The metric should be related.

It is important that the metric should have sufficient context

to be meaningful to the dashboard consumer. Typically, it
might make sense to show the gauge to show a percentage
vs. an absolute value, as it minimizes the mental effort
required on the part of the consumer to interpret it.
Comparison of Aggregated Values / Totals
Bar Charts, Column Charts, Tree Maps, Point Geographic
Maps, Pie and Donut Charts
Bar Charts
Column charts can be used for comparing items and comparing between
different types of products or brands (categories). Charts can include multiple
values on both axes, as well as a breakdown by categories displayed on the
horizontal X-axis. Data is graphically represented by vertical bars next to each
other lined up on the horizontal axis. Each bar represents a different category, Use for metrics across single dimensions.
and the height of the bar correlates with numbers on the vertical Y-axis.
Generally, does not require an axis.
Bar charts are used for comparing values, items or data. They are like column
charts with their axes reversed: values are represented on the horizontal X-axis, All bars should be the same color.
and the categories are on vertical Y-axis. Bar charts are commonly used to
compare different values, items, and categories of data. Bar charts are not
usually used to show trends over time.
See example dashboards:
GoFigure! FIFA World Cup Dashboard

GoFigure! Healthcare Search Dashboard

Comparison of Aggregated Values / Totals Cont. Tree Maps

Useful for drawing the eye to relative sizes.

Column Charts
Tree maps should be used instead of pie charts where there are more than
two-dimension members.

Label dimension members and metrics as much as possible.

Use for metrics across single dimensions.

Consider using an axis.

All bars should be the same color.

Best Practices:

When showing a single item over a long period, line charts are better for
revealing trends.

Add a second set of items for side-by-side comparison.

See example dashboards:
Use a unique color for each value and add additional colors to your palette
GoFigure! Healthcare Search Dashboard
as required.
GoFigure! e-commerce Dashboard
Use bar charts instead of column charts when categories have long names.

Displaying lines and bars together can give you different views of the same data.
Comparison of Aggregated Values / Totals Cont. Pie and Donut Charts

Point Geographic Maps

To be used sparingly, if at all.

Do not use if more than three members in the pie chart dimension.

Do not use if the business user is going to use visualization to make fine distinctions.

A point map indicates that the metric shown exists only at a singular point.
This format is best for raw numbers such as sums or counts, or when the data
represents a discrete location.

Do not attempt to show so much information the business user is overwhelmed

(several metrics etc.).

The smaller geographic regions must be easily visible.

In Tooltips, provide additional details, including the full names of states, etc.
Comparison of Aggregated Values / Totals Cont.

Donut chart

For more Pie exapmple see:

Sisense GoFigure e-commerce Dashboard

Sisense GoFigure e-commerce Dashboard

See example dashboards:

GoFigure! Baby Names Dashboard

GoFigure! e-commerce Dashboard

 Column Charts
Comparisons of Statistical Functions
(Average, Min, Max, etc.)
Column Charts and Bar Charts, Tree Maps, Pie and Donut
Charts, Filled Geographic Maps, Pie and Donut Charts

Bar Charts

Use for metrics across single dimensions.

Consider using an axis.

All bars should be the same color.

Use for metrics across single dimensions.

Generally, does not require an axis. See example dashboards:

All bars should be the same color. GoFigure! Bitcoin Dashboard

GoFigure! Ebay Auction Dashboard

Comparisons of Statistical Functions (Average, Min, Max, etc.) Pie Charts
Pie charts show the breakdown of a complete entity into several parts. Each slice
of the pie represents a static numerical value or category, and the sum of all
Tree Map
categories is equal to 100%. Each slice is given a unique color to make it easily
distinguishable from the other slices or categories.

Useful for drawing the eye to relative sizes.

Tree maps should be used instead of pie charts where there are more than To be used sparingly, if at all.
two-dimension members.
Show clear labels.
Label dimension members and metrics as much as possible.
Do not use if more than three members in the pie chart dimension.

Do not use if the user is going to be using this visualization to make fine distinctions.
See example dashboards:
GoFigure! e-commerce Dashboard

GoFigure! Bitcoin Dashboard

Comparisons of Statistical Functions (Average, Min, Max, etc.) Cont. Filled Geographic Maps

Donut Chart

Use no more than 6-7 subcategories in a pie chart so information can be easily
understood.

For many types of data, use a tree map instead of a pie chart.

Labels are better than keys for identifying each slice.

Data should be sorted by size so that slices are also ordered according to their size.

Include the values inside each slice for easy comprehension.

A filled map implies homogeneity of the data over the area to which it applies.
Therefore, the metric should always be normalized, as an average, percent,
or per unit.

Do not attempt to show so much information the business user is

overwhelmed (several metrics etc.).

Show a clear legend.

The smaller geographic regions must be easily visible.

See example dashboards: In Tooltips, provide additional details, including the full names of states, etc.
GoFigure! Google Play App Store

GoFigure! World Happiness Dashboard

 Sparkline
Time Series
Line Graph

Shows trends in a series of values, such as seasonal increases or decreases,

economic cycles, or to highlight maximum and minimum values.

Make appropriate use of axes notation.

Datapoints should be clearly marked.

Individual Datapoint Details Best Practices:
Pivot Tables Increase understanding by breaking down data into subcategories, such as
sales per quarter.
Pivot tables are one of the most useful ways to simplify, summarize and visualize large
amounts of data in a table or Excel spreadsheet that can be difficult to understand. Enhance the table with additional features such as color formatting and data bars.
For example, if you have a table with the source data for all sales orders in 2017,
it might be hard to read, especially if it includes all the following: order ID, order date, Use conditional formatting to reveal areas of success or areas that require follow-up.
customer name, address, shipper name, time and date of delivery, products ordered,
Consider how end-users use the table and pre-filter the pivot table to show only the
quantity, revenue, salesperson, payment type, shipping costs, etc.
most important data. When embedding, control access to users with system settings
A pivot table could give you pertinent insights from this mass of data, like that control access to data and make sure users see only the data they need.
understanding how much each salesperson sold during the year.
Use color to highlight critical data and ease understanding.

All the data we wish to view should be visible without using ribbons.

The eye should be drawn to significant values. In the example, above:

There is a manageable number of rows.

Utilizing conditional formatting when possible.

Using a data bar to display the difference between records.

Composition Comparisons - Percentage Stacked Column to 100% example:

Weightings
Stacked Bar to 100% example:

The bar should be ordered with the largest values at the base.

Label all dimension members.

See example dashboards:

GoFigure! e-commerce Dashboard

GoFigure! Refugee Trends Dashboard

Ranking Comparison Year-over-Year Comparisons
Dual Axis Bar Charts

Bump Charts

This is useful when tracking changes over time.

The core member we are interested in tracking should be easily distinguishable

by color.

While it may be tempting to assign a different color to each member of the

group, the number of colors can be confusing. It typically makes more sense
to use the same color for all the non-core members and have a selectable l The advantage of using bars for both time periods is that it is very clear we are
egend to highlight individual members. looking at the same metric.
Include Name, Rank, Details, etc. on tooltips. This type of chart requires a clear legend.
 Provide additional data in Tooltips.
Relationships Between Two Dimensions
Scatter Charts and Bubble Charts
Or Metrics Scatter charts display the distribution of two variables on an X-Axis and Y-Axis. The
colored circles scattered across the chart represent the categories being compared
These charts are useful for seeing a cross-section of the data. The chart type should (indicated by circle color). Bubble charts are variants of scatter charts and show the
reflect the granularity. numeric volume of the data (indicated by the circle size). A scatter chart can reveal
Heatmaps are useful for less granular dimensions. trends, clusters, patterns, and correlations between data points.

Scatter charts are useful for finer data. Scatter Chart:

It can be useful to provide multiple choices on the axes, but the list of available options
should not overwhelm the business user, and the possible combinations should make
sense from a business point of view.

Heatmap Chart

Highlight significant data points (such as using an icon for core member).

Label values in cells as much as possible. When deciding on whether to label, consider if the number of labels will
clutter the view.
Provide a legend for the color scheme.
 Data point size or color can be used to show additional dimensions.

Provide additional data in Tooltips.

Tracking Migration of a Cohort
Best Practices: Sankey Diagram

Use a unique color for each dataset.

Use dot size and unique colors to display more information.

Use labels with caution as they may clutter the chart or create visual noise.

Highlight significant data points (such as using an icon for core member).

Bubble Chart

Typically used when looking at a process and we are interested in where a

cohort terminates.

The width of the flow indicates its size, so immediately tells a story.

Provide additional detail on tooltips.

Funnel Chart
Inflows and Outflows
Waterfall

Typically used when looking at a process and we are interested in how

many members make it through each stage.

Used to show the effect of various categories of flows to a metric, such as

contribution to profit by each product.

Use colors to indicate positive and negative flows.

Not restricted to monetary values. Useful in other areas like human resources.
Can be used to illustrate the effect of hires, terminations, and retirement on
headcount.
Distributions of Individual Data Points Distribution of Binned Values
Box and Whisker
Histogram

Useful for showing outliers in the distribution of individual values or differences in

distributions across products.

Include an axis for reference. Useful for distributions.

Use tooltips to show details. Ensure that the distribution tells a useful story.

Provide additional detail on tooltips.

Column Chart with Dynamic Buckets
Frequency of Text Instances
Word Cloud

Useful for distributions. Useful for social media sentiment analysis.

Ensure that the distribution tells a useful story. Make judicious use of text, colors, and sizes.
Provide additional detail on tooltips. Does not have to include all values.

Provide additional detail on tooltips.

Geographical Data Area Map:

Area Maps allow you to visualize geographical data as polygons on a map. You can
Scatter Maps and Area Maps use your data to affect the color of the areas. It’s good to note here that these are
four-dimensional and cannot be done in Excel.
Scatter maps are a variant of the scatter chart and allow you to visualize one or two
types of geographical data as data points on a map. Scatter maps distinguish data
by using different colors and sizes for the data points.

Scatter Map:

Best Practices:

Increase the accuracy of the map by entering precise geographical data such as both
the city and the country or the latitude and longitude coordinates.

Use a variety of sizes and colors to display additional information and metrics.
 Area line charts are very similar to line charts but are much more visual as the areas
Trends under each line are filled in (colored). Classic area charts have overlapping areas while
stacked area charts do not. Stacking area charts are recommended for displaying
Line Charts and Area Line Charts absolute or relative values over a period of time.
Line charts show solid, discrete numbers or data points over time, with every two Area charts are particularly effective for visualizing differences or disparities among
points joined by a line. Line charts let you easily visualize trends as well as rapid multiple trends.
changes in data that are indicated by spike or peaks. Just about any type of data
over time can be visualized with line charts. If you have a series of data points with multiple trends, consider using a Scatter Chart
that displays the data points without the connecting lines
Line Chart
Area Line Chart:

Best Practices:

Data points should be clearly marked and connected by solid lines.

When representing multiple types of related data in a single line chart, use a unique
color to indicate each category.

Do not use more than four types of data in a single line or area chart to minimize
visual noise.

Stacked area charts allow you to compare several trends at once visually.
Stacked Area Chart:

See example dashboards:

GoFigure! Chinese Food Delivery Dashboard

GoFigure! Google Play Apps Store

How to Assemble Final
Dashboards, Design Formats,
and Common Pitfalls
#3 in a Series of 4

Assembling Final Dashboard in the header. The aim is to reduce the amount of effort needed by the business user to
identify exactly what information is being presented.
The dashboard should normally follow a Time Period

prescribed structure. Display the dates being selected prominently.

Dashboard Header If the business user can choose a date band, show Start Date to End Date.

Dashboard Title
Use an explicit title that says what the dashboard addresses. Using the KPIs
word dashboard in the title is sometimes redundant. KPIs should provide a summary of the dashboard’s main points:

Icons They should generally be shown at the top of the visualization area.
Have client logo and Help icons if necessary. They should be of a
Standard fonts and colors, and numbering conventions should be used, but they can
consistent size and aligned with each other, and consistent with other
generally be larger than on individual visualizations.
dashboards.
Icons can be used to make results clearer.
Displaying Core Metrics in Header
If the business user can use a parameter or filter to show a core metric for For embedded analytics, try inserting some general KPIs and leave the option for the
the dashboard, the name of the selected metric should also be embedded user to change and insert new ones.
 Make sure labels are unambiguous. Avoid abbreviations where possible.
Formatting and Display Standards Show total amounts in titles if relevant.

Using a standard font and color scheme provides a Is it possible to misinterpret the values?

recognizable environment for the business user and Axes, Borders, and Gridlines
saves the developer having to develop a new scheme Axes and borders can distract from the visualization.
for each dashboard. Axes should be only used where it is necessary to understand the visualizations.
Font
Gridline and borders should be avoided, for a cleaner feel.
There should be standardized fonts and sizes that comply with a style guide.
Numbering Conventions
Avoid excessively small text.
Numbering conventions should be intuitive and assessed in terms of the likely
When embedding, ask the client for a copy of their brand and style guide in order values to be shown and the amount of precision that is likely to be needed.
to incorporate into the dashboards. Dollar amounts should be shown with currency symbols.

Color Scheme Rounding should be intuitive.

The developer should use the standard palette. For values of less than 100,000, consider showing absolute numbers rounded
to the nearest whole number. i.e.; $64,567.33 is shown as $64,567
If there is only a single dimension, all members of that dimension should be the
same color. For values of 100,000 to 999,999, use the K convention to one decimal place,
i.e.; $564,567.33 becomes $564.6K.
Make sure the color legend is explicit, prominent and easily identifiable.
For values for over a $1 million, show to a million, with two decimal places,
Are colors consistent for the same dimension on different visualizations?
i.e., $9,564,567.33 becomes $9.56m
Labeling Avoid showing values in terms of ‘000s.
Are all data points labeled? Showing 5,550 with ‘000s in the title, is far less clear than just showing 5.55m.
Is it clear what data point is referenced? By using the “m” notation, the business user doesn’t have to do mental
arithmetic to convert the value to millions.
Highlighting the most relevant values
Highlight values that change with updated selections in bold and italics. Common Dashboard Pitfalls
Clarity of Language Overly involved dashboard
Are there any abbreviations that could be misinterpreted by the non-specialist Do not make the dashboard overly complicated. A single dashboard does not have
business user? to answer all the questions the business user might ever want to know of the data.

Are there any typos or spelling mistakes? While the idea of a single dashboard might be attractive in principle, the effort the
business user might have to put into understanding how to use it can make it unwieldy.
Does the title make grammatical and logical sense? It may, therefore, make more sense to give a series of more structured dashboards
broken out by subject area.
Tooltips
Too many or too large visualizations on a single dashboard
Tooltips should give additional details. They should not be substitutes for labeling.
Keep the dashboard limited to 3-4 main charts, so the business user is not scrolling
The other standards regarding font, numbering, color scheme, clarity should also be to view components.
applied to tooltips.
Unclear language or unnecessary abbreviations

Natural language should be used as much as possible. Abbreviations should only

be used if there is no alternative.

Average rather than Avg.

Amount rather than Amt.

Redundancy in naming convention

Typically, this is found when adding in data fields from the source without changing
the names. The data source naming convention may not be suitable for visualizations.

Using the name “Prod Desc” rather than the simpler “Product.”
Too many different types of information on one visualization Combining magnitude and composition on a visualization
Do not overload a single visualization with too much information (e.g., including, Avoid trying to show both relative size and composition simultaneously on
sales, demographic information, comparisons, etc.). This can overwhelm the the same visualization.
business user. Instead, consider tooltips or separate charts.

Not using the appropriate visualization

Do not to use a complicated visualization just because it is available. A visualization
should only be used if meets the business requirements. For example, using a pie
chart may not make sense with many dimension members.
The same data is far more easily understood as a tree map.

It is typically better to split this into two separate visualizations.

Having insufficiently labeled data

It should be possible to know the values being displayed on the dashboard without
the use of tooltips or other interactions.

Redundant visualizations

Confirm that the same business question with the same data is not being displayed
different ways in multiple visualizations, like having two views to illustrate ‘percentage
of customers by the state’ or ‘total sales by salesperson.
Misapplying geographical maps Displaying too much granular information

Do not use geographical maps for displaying aggregated data such as totals. Only relevant data should be displayed. A large undifferentiated table can confuse
the relevant data points.
Large contributions from the geographically smaller state can be missed.

Geographically larger states can be given undue weight by the business user
due to their size.
Overly formatted visualizations Being unable to see the visualization from the business user’s perspective

Control the use of borders, so they do not overwhelm the business user. It is not uncommon for developers to know the business area so well that they assume
Borders should not be used to separate sections, table rows, or define top panels certain points are obvious, so it is key to get a new, third-party perspective and validate
and side panels. that your message comes across to others.

Ask colleagues to review your design before release.

And if you are embedding analytics for your customers, always clarify their exact needs
and get feedback on the dashboards to make sure they are useful.

NOTE: Usage analytics can help determine which dashboards are being used, therefore
indicating which ones should be investigated further.
Dashboard Design
Worksheets & Checklists
#4 in a Series of 4

Peer Checklist:
Common Dashboard Pitfalls
Initial Planning Checklist

Stage Task Developer Stage Task Developer

Confirm metrics and dimensions required with Dashboard Design Sizing
business user
Confirm Data Plan Workspace
Efficient Structure
Sourcing
Field Naming Convention
Correct Aggregation
Filtering
Visualization Design -To be checked for each proposed visualization
Functional
Functional Visualization Options Developer
Visualization Options Developer Requirement
Requirement
Bump Chart
Bar Chart Ranking Comparisons
Other (Explain)
Column Chart Dual Axis Bar Graph
Compare Aggregated Year over Year
Values Tree Map Comparisons Other (Explain)
Pie or Donut Chart (Confirm appropriate) Heatmap
Relationships
Other (Explain) between dimensions Scatter Chart
Bar Chart or metrics
Other (Explain)
Column Chart Box and Whisker
Distribution of
Compare Statistical Tree Map Individual Data Points Other (Explain)
Functions Pie or Donut Chart (Confirm appropriate) Distribution of Binned Histogram
Geographic Map Values Other (Explain)
Other (Explain)
Line Graph
Time Series
Other (Explain)
Table (Confirm appropriate)
Individual Details
Other (Explain)
Stacked bar to 100%
Composition
Comparisons
Tree Map
Individual Visualization Review
Visualization Visualization
Functional Check Developer Functional Check Developer
Component Component
Fonts match the client standard Clarity Titles, metrics, dimensions in clear language
Font
Consistent with other fonts on the dashboard Tooltips give additional details, not used as
substitutes for labeling
Colors match the client standard Tooltips
Other standards regarding font, numbering,
Color Scheme Consistent with other colors on the dashboard, color scheme, clarity also applied to tooltips
particularly for the same dimension
All data points labeled
Clear what data point is referenced
Labels are unambiguous and avoid
Labeling abbreviations where possible
Total amounts in titles where relevant
User selectable or pertinent values highlighted
in the title
Axes only used where they are necessary to
Axes, Gridlines, and understand the visualizations
Borders
Gridlines and borders avoided where possible
Number convention intuitive and does not
require mental arithmetic
Numbering Currency and thousand symbols displayed
Conventions where appropriate
Small numbers presented as whole numbers,
larger shown with K, M as required
Assembling Final Dashboard
Dashboard Dashboard
Functional Check Developer Functional Check Developer
Component Component
Title correctly formatted Selections are presented in the bottom right
panel
Icons present and aligned Selections displayed for filters made on
Help Icon provides information the visualizations, rather than the docked
Selection dropdown filters, which will be readily visible
Header Header displays core metrics above
The header shows time periods (including to/ Selection sheet follows the standard format
from dates) where applicable
Selection sheet checked after any dashboard
Title. Metric, Date displayed in clear language
changes
Header consistent with Client standards
Filters Docked in one Panel on the right
Undocked filters are positioned near relevant
visualizations
Filter Titles sufficiently descriptive to a non-
Filters specialist
Filters follow the standard format (dropdown/
slider etc.)
Dimension member names visible even when
long
KPIs are at the top of the visualization area
Standard fonts, colors, and numbering
KPIs
conventions used
Appropriate use of font size and icons
Final Checklist

Task Developer

Dashboards tell a coherent story

No redundant visualizations
The business user gets the information they require without extensive
knowledge of the dashboard, data modeling, or Sisense
Notations are in plain English
There are no typos, truncations or unexplained variations from the
house style
Efficient use of space
Only abbreviations where absolutely necessary

Visualizations aligned

No unnecessary borders between the visualizations

Where values may be unlabeled due to space constraints, this
limitation is explicitly recognized
The dashboard is aesthetically pleasing
Peer confirms visualizations are extremely clear to the non-specialist
business user without mental effort
Peer confirms user can understand the values displayed by the static
dashboard without having to interact with it further
Display in published version is checked for issues
Peer Checklists
Individual Visualization Review
Dashboard Dashboard
Functional Check Peer Functional Check Peer
Component Component
Currency and thousand symbols displayed
Fonts match the Client standard Numbering where appropriate
Font
Conventions Small numbers presented as whole numbers,
Consistent with other fonts on the dashboard
larger shown with K, M as required
Colors match the Client standard Clarity Titles, metrics, dimensions in clear language
Color Scheme Consistent with other colors on the dashboard, Tooltips give additional details, not used as
particularly for the same dimension substitutes for labeling
All data points labeled Tooltips
Other standards regarding font, numbering,
color scheme, clarity also applied to tooltips
Clear what data point is referenced
Labels are unambiguous and avoid
Labeling abbreviations where possible
Total amounts in titles where relevant
User selectable or pertinent values highlighted
in the title
Axes only used where they are necessary to
Axes, Gridlines, and understand the visualizations
Borders
Gridlines and borders avoided where possible
Numbering Number convention intuitive and does not
Conventions require mental arithmetic
Final Dashboard Structure
Dashboard Dashboard
Functional Check Developer Functional Check Developer
Component Component
Selections are presented in the bottom right
Title correctly formatted
panel
Icons present and aligned Selections displayed for filters made on
the visualizations, rather than the docked
Help icon provides information dropdown filters, which will be readily visible
Selection
Header Header displays core metrics above
The header shows time periods (including to/ Selection sheet follows the standard format
from dates) where applicable Selection sheet checked after any dashboard
Title. Metric, Date displayed in clear language changes

Header consistent with Client standards

Filters Docked in one Panel on the right
Undocked filters are positioned near relevant
visualizations
Filter Titles sufficiently descriptive to a non-
Filters specialist
Filters follow the standard format (dropdown/
slider etc.)
Dimension member names visible even when
long
KPIs are at the top of the visualization area

KPIs Standard fonts, colors, and numbering

conventions used
Appropriate use of font size and icons
Final Checklist

Task Developer

Dashboards tell a coherent story

No redundant visualizations
The business user gets the information they require without extensive
knowledge of the dashboard, data modeling, or Sisense
Notations are in plain English
There are no typos, truncations or unexplained variations from the
house style
Efficient use of space

Only abbreviations where absolutely necessary

Visualizations aligned

No unnecessary borders between the visualizations

Where values may be unlabeled due to space constraints, this
limitation is explicitly recognized
The dashboard is aesthetically pleasing

Clear to the non-specialist business user without mental effort

The user can understand the values displayed by the static dashboard
without having to interact with it further
Display in published version is checked for issues