close

image
image

The 4 questions for every dashboard : Not a linear process
  1. Who is the audience
    i. What do they know already?
    ii. How much detail to include or exclude?
    - Examples :
      a. Your team
        - Consider the visualizations for explanations
        - Simple chart types
        - Understand the data
      b. Your department
        - Add more explanations and terminology
        - Explain the data
      c. Senior management
        - Much higher level, more data from more sources
        - Not as much detail
        - Ability to drill down
        - Explanations on terminology and data
      d. Operational
        - Real-time data
        - Filters and interactions to drill down
      e. General Public
        - Carefully explain the data and what's important
        - Consider chart type and explanations
image
image
  2. Where will the dashboard be viewed? Size and Level of interaction
    i. Large or Small screen?
    ii. Print, Mobile, or desktop?
    - Examples :
      a. Desktop/Laptop
        - Greatest freedom
        - Large amount of space
        - Lots of interaction
        - Filters, parameters, actions
      b. Tablet/Mobile
        - Smaller maximum size
        - Consider the orientation
        - Scrolling v.s. single screen
      c. Print/TV Display
        - Can't interact
        - Everything needs to be on screen
        - Large fonts and labels
        - How close or far away does it need to be seen?
  3. Why are you sharing the data? What is the purpose?
    i. Is it to form?
      - Informative : Built for information
        - All about information
        - Present the data to easily find answers
        - Stil provide some interaction
        - More focused on a single question
    ii. Is it to explore?
      - Built for discovery
        - All about discovery
        - Present data to easily find answers
        - High levels of interaction
        - Filters, actions and highlights
    iii. Is it to explain?
      - Built for insight
        - All about insight
        - Presenting the answer of the user
        - Show the interesting thing we have found
        - Persuade or make the case of change something
    iv. Is it to measure a goal?
      - Key performance indicator (KPI) : Built for attention
        - All about getting someone's attention
        - Highlighting things of interest
        - No filtering, no interaction
        - Clearly obviously that something is right or wrong
  4. What information will you show?
    i. What question will it answer?
    ii. What do you want the audience to take away?
    - Questions
      - What elements will help?
      - What data do we need to display?
      - High or low level of detail?
      - Live or extracted data?
      - How many filters do we need?
      - What do we need to include?
      - What should we exclude?
      - What is the minimum viable information required?

==== Dashboard Structure ====

Create a basic dashboard
image

Size your dashboard
  1. Amount of Dashboard Space
    - How many elements
    - Layout options
  2. Location of the Dashboard

Tiled vs Float
image

  - Don't mix a tile and a float

How to use layout containers
  - Layout containers are useful for designing your dashboard.
  - Create layout container
    
Titles on your dashboard
imageimage

Challenge : Create a basic dashboard
imageimage

==== Dashboard Component ====

Apply filters across sheets
image
image
  1. How to make filters apply to some sheets, but not others.
  2. Where to place filters

Highlight data with annotations
image
  - Point annotation refers to axis, not data, so filters changing cannot impact its.

Additional information and help
  - Instrucitons on how to read the dashboard or where the data came from.
  - Help buttons or informational buttons as to not take up valuable space

Collapsible containers to show or hide elements
image

Challenge 
image

==== Dashboard Design Elements ====

Colors :easy to get wrong, too much color adds to confusion
image
image
image
  - Better : Blue for positve profit, Orange for negative profit

Remove chart junk
image

image
  1. Remove the map title
  2. Remove line chart title, axis title/header
  ...
image

Visually Group elements with borders
image

Challenge
image

Add interaction with actions :
imageimage
imageimage
imageimage

Filter Actions :
URL :
imageimage
imageimage
imageimage

Go to Sheet :
imageimage

Proportional brushing using set actions
imageimage
image

Asymmetric drill down using set actions
image
Challenge :
image

==== Using Stories to Make Dashboard ====

Data viz vs data stories
imageimage
image

Narrative structure
image
imageimage

Big Stories
image
image

Small Stories
image
image

==== One Dataset : Five Dashboards ====
imageimage
  - Look for Regional Variation
image

Build an informative dashboard
image
image

Build a data story : For changes
imageimage

Build a KPI dashboard
imageimage
image
image

Design for tablets and mobile
imageimage

Next 
image

arrow
arrow
    全站熱搜

    pixnetinpenang 發表在 痞客邦 留言(0) 人氣()