Dynamic Dashboards - Mike
Automatically Maintaining Targets for Selectors
- All attribute and metric selectors automatically target all grid graphs and panel stacks that are in the same panel or document section as the selector - You cannot change the targe of an attribute or metric selector except by moving controls to different document sections - Any new grid graph or panel stack is automatically added as the target of all attribute and metric selectors in the same panel or document section - Does not work for panel stack selectors (manually managed)
Depending on if you enable or disable the following options impact the behavior of selectors
- Allow multiple selections - Automatically update when there is no data for the current selection - Show option for total - Show option for All - Automatically apply selector changes
Process of creating a multipanel dashboard involves
- Consulting your target audience - Gathering related datasets - Devising the investigative workflow - Building the dashboard
MicroStrategy Out of Box visualizations
- Gauges, Thermometers and cylinders for single value visualizations - Time Series Slider - Interactive Stacked Graph - Heat Map - Interactive Bubble Graph
Selectors in a Document that you view Offline
- If selector uses filtering only the data for the current selections are included in the document - If selector uses Slicing all data is included in the document so the selectors are still functional
Selector to control another Selector
- Place the selector that you want controlled by another selector on a panel stack - Define the master selector to use the panel stack as its target. - To avoid no data returned scenarios enable the Automatically update when there is no data for the current selection check box for the "child" selector
DHTML Benefits
- Shorter initial load time - Incremental fetch of data - true zero footprint on the client - Minimal security issues - Smaller output files - Best mode for PDF display, printing, and archiving
Widgets data requirements
- Source dataset must contain one attribute in the rows and one metric in the columns - You must include a selector if you want users to interact with the widget
Design recommendations for graphs
- Use pie charts sparingly unless you are analyzing the parts of a whole use bar charts instead of pie charts - When screen space is precious, use thermometers instead of gauges and dials - round numbers on graph labels - Use caution with 3-D effects and distracting backgrounds on graphs - Downplay the non-data visuals and emphasize the most important data - Stick with a graph style that works - Label scales and graph sections for context and use tooltips when you need to provide more detail
Differences between filtering and slicing selectors
- With slicing you can enable the total to be displayed as an item in the selector. "Show option for Total" - With slicing you can specify that the selector automatically uses a default selection made by the user to retrn no data "Automatically update when there is no data for the current selection" - MicroStrategy Web slicing is performed on Intelligence Server in Interactive Mode. Flash Mode slicing is performed on client machine - Filtering is performed on the Intelligence Server in both Interactive and Flash modes
Number of Datasets Performance
1 to 5 - Optimum (recommended) 5 to 10 - Acceptable > 10 - Slow (not recommended)
Layer Dashboards in two ways
1. Analytic Panel Layers - layering grids and graphs on a single panel stack within a larger dashboard 2. dashboard panel layers - Use a large panel stack as the basis for a multilayered dashboard
High Level steps for inserting and defining panels
1. Insert a panel stack 2. (optional) Display the title bar which shows either the panels tack title or the panel name to identify the panel stack 3. Insert as many additional panels to hold layered data 4. Add object to each panel to display the data 5. (optional) Specify an order for the panels 6. Choose a panel to display when users initially view the document 7. Format the panel stack, panels, and title bar 8. Add a selector to enable users to switch between panels
What is a Scorecard
A scorecard is another type of document you can create with Report Services that focuses on KPI's. Like dashboards scorecards use graphical indicators to show trends, actuals against targets, or changes in scores Adhere to Six Sigma and the Balance Scorecard
Bar and Sparkline microcharts
Are based on the same metric so they are designed the same way.
Widgets
Are interactive flash only graphs that dynamically update when you select a new set of data to view Built in Design or Editable Mode To interact with them you must be in Flash mode
Graph Matrix Columns Requirments
At least one attribute - The first (top-most) attribute on the columns provides the column headers in the widget. These values are used to slice the data or group it by the attributes on the columns - Additional attributes on the columns produce separate line graphs within each area graph in the widget Note: if there are no attributes on the columns you must place at least three attributes on the rows. - At least one metric, to the right of or below the attribute on the columns - Additional metrics are depicted as forecast line graphs in each area region
Graph Matrix Rows Requirements
At least two attributes - The first (left most) attribute on the rows and the second, thirsd, etc provide the row headers in the widget - The last attribute on the rows provides the X-axis of the graphs. This attribute which is generally time based is used to drive the time series of the graphs
Row Data requirements for Bar or Sparkline Microcharts
At least two attributes - The first attribute on the rows and the second, third, etc determine the total number of rows displayed in the widget. The elements from those attributes become individual rows in the widget - The last (right most) attribute determines the X-axis values in the bar and sparkline microcharts in the widget
Column Data requirements for Bar or Sparkline Microcharts
At least two metrics - The first (left most) metric on the columns determines the size of the bars (the series) in a bar and the fluctuations in the line graphs in the sparkline - The second metric creates the horizontal reference lines that display in the sparkline. - If you place a third, fourth, fifth, and sixth metric on the columns the metrics display as individual columns on the widget If you have a seventh metric you can produce a bullet
Limit the Number of Fonts in a Dashboard
Avoid using multiple font types on a dashboard. Only use different font type for headings if necessary.
View Mode Limitations
Cannot create a new document Cannot edit an existing document Cannot manipulate any objects on the document as you can in any of the other display modes
Interactive Mode limitations
Cannot create a new document Cannot format the layout and positioning of objects or the entire document Cannot format widget Flash properties
Flash Mode Limitations
Cannot create a new document Cannot manipulate or format grid graphs except to sort and pivot objects on them If a grid graph uses a graph type that is not supported the graph is not displayed
Design Mode Limitations
Cannot view the results of the document without switching modes Cannot format all aspects of grid/graphs including metric values and attribute headers Cannot format widget Flash properties Cannot use selectors to flip through the panels in a panel stack or display different attribute elements or metrics in a grid/graph Cannot use Page-by to group data Cannot sort grid reports Cannot use the Fit to contents/window features Cannot show or hide rulers
Focus on the Data First
Choose right data for target audience and display the data in the best possible format. Keep dashboard simple. Too many graphical elements on a dashboard detract from the importance of the actual data
Bubble Grid data Columns requirements
Columns At least two metrics - The first (left-most) metric on the columns determines the size of the bubbles - The second metric determines the color of the bubbles - Any additional metrics display in tooltips when the user hovers the cursor over a bubble in Flash Mode. These metrics do not affect the size and color of the bubbles
Dynamic Content Dashboards
Consolidate many reports into a single dashboard through use of multiple layers or panels. Enable users to interactively change the context of the dashboard by controls Target users who want to go beyond a birds eye view of the business to identify actual causes for exceptions and alerts. Provide guided workflow where they can identify problems and take immediate action based on what they see within layers of single dashboard
Formatting features available to graphs and other objects
Gradients Drop shadow effects 3D borders
Operation modes for Microcharts Widget
Grid Vertical Scroll Ticker KPI List
Item Comparisons reports show ranking of individual items compared to one another
Horizontal bar chart is best for ranks
Interactive Mode Features
Intended for document analysts Optimized for Dashboard viewing View the results of the document Use selectors to flip through the panels in a panel stack or display different attribute elements or metrics in a grid/graph Format grid grapms Sort grid reports and pivot report objects on them Add totals Resize row and columns Create metrics based on report objects already on the grid report
Flash Mode Features
Intended for document analysts View the results of the document Use selectors to flip through the panels in a panel stack or display different attribute elements or metrics in a grid/graph Access and interact with features provided by Flash such as widgets Format widget Flash properties Sort grid reports and pivot report objects on them
View Mode Features
Intended for document analysts View the results of the document only as you might in a static PDF file Provides better performance than all other modes but the document is not interactive
Design Mode Features
Intended for document designers Create a new document Edit an existing document Quicker performance since document results are not displayed Add and remove dataset objects Insert controls such as grids, graphs, text boxes lines shapes panel stacks edit and format controls Format grid graph containers Pivot report objects on grid reports Use Grouping Pane
Editable Mode Features
Intended for document designers Create a new document Edit and existing document View the results of the document Add and remove dataset objects Insert controls such as grids/graphs, text boxes, lines, shapes, and panel stacks Edit and format controls Format grid graphs inclduing the formatting of metric values and attribute headers use selectors to flip through the panels in a general stack or display different attribute elements or metrics in a grid/graph use Grouping pane Use the Fit to contents/windows features Sort grid reports and pivot report objects on them Show or hide rulers
Panel Stack
Is a container for a group of panels. With panels in place you add objects (controls) such as grids, graphs, text boxes, etc on which display the data that a user sees on a given panel Selectors - radio buttons, drop down list enable users to display different panels in a stack Selectors are not part of the Panel Stack
Current Panel
Is the one that is currently displayed and it is also the one that is exported to PDF or Excel
Time Series Slider Widget
Lets users see two views of time-related data, microview or macroview. The Time Series Slider has a controller that permits users to set the range of data taht is visible in a separate area graph.
Flash Benefits
More interactivity - Immediate interactive response - Advanced visualizations and animations - Disconnected operation (can use MicroStrategy office to run documents saved in Flash Mode in Word, PowerPoint, Excel, and Outlook) - Unlimited visualization extensibility
Document preparation is impacted by
Number and size of datasets Type of joins between datasets Number and complexity of the grid/graphs on the document Number of selectors and their targets Number of grouping objects Number of panels (including nested) Number of text boxes Number of thresholds
Pay Attention to Object Placement
Objects in top left or center of dashboard receive the most attention. Object placement is largely influenced by how a person reads (left to right, top to bottom) Larger objects can attract attention no matter where they are placed
Data Cloud Widget Row Requirements
One Attribute - This attribute on the rows determines the elements that display in the widget - If you include more attributes to the right of the first (left most) attribute, elements from all of the attributes are combined and displayed in the widget
Data Cloud Column Requirements
One metric - The metric determines the size of the font of the attribute elements - Any additional metrics display in tooltips when the user hovers the cursor over an element in Flash mode
Microcharts Widget
One or more compact charts integrated into a grid of data that enables analysts to quickly identify trends.
Datasets Sizes Depending on Data
Optimum (recommended) Size Bytes < 250 KB Cells < 6500 Acceptable Size Bytes Between 250Kb and 1.5 MB Cells Between 6500 and 25000 Slow (not recommended) Size > 1.5 MB Cells > 25,000
Dashboard Categories
Performance Summary Dashboards Metrics Dashboards Dynamic Content Dashboards Dynamic Visualization Dashboards
Editable Mode Limitations
Performance reflects the fact that you see all document results as you work Cannot format widget Flash properties
Component Comparison reports show percent to total
Pie charts can be effective. Keep pie slices to 6 or less Vertical Bar Charts when you want to compare components of more that one total.
Keep Simple Navigation in Mind
Place navigation buttons and links in consistent locations
Performance Summary Dashboards
Provide at a glance high level views of business. Present several interrelated sets of data in tables and graphs on a single screen. Target audience is executives, high level managers who need to see company performance in a summary view
Bubble Grid data Row requirements
Rows Only two attributes, consolidations, or custom groups - The first (left most) attribute on the rows provides the X-axis The attribute can be time-based - The second attribute on the rows provides the Y-axis
Selector Components
Selector Style - Determines how selector displays and whether it enables multiple selections Action type - Determines whether selector displays elements, metrics, or panels Source - Defined only when the selector displays elements The source is the attribute, custom group or consolidation whose elements display in the selector targets - Determines which grids, graphs, or panel stacks the selector affects
Show totals in a selector
Selectors that contain attribute, consolidation or custom group elements as selector items can also include an option to display totals. The total is calculated for all of the selector items
Keep Related Data on a Single Screen
Show a dashboards data on a single screen users can better assimilate the information displayed Avoid dashboards that scroll Layer dashboards or link to source reports
Bubble Grid Widget
Shows trends or anomalies in data relative to the total contribution of accompanying data Sizes and color are controlled by two distinct metrics Each bubble is plotted at the intersection of two different attribute elements. Beneficial when used to perform business ration analysis ex numbers of customers in a region vs. the revenue generated per customer
Frequency Distribution Comparisons show how many items fall into a series of progressive numerical ranges
Step-column charts (histograms) Line Charts (histographs) Use two scales and display as a bell shaped curve
Define Dashboard
Strives to make data consumable for everyone. Through use of visual presentations (graphs) any business user can benefit from retrieving information from a dashboard. Dashboards can provide enterprise wide transparency through the use of presenting multiple data sources in one location. Dashboards offer a unified view of the business to users from different areas within the business. They help organizations identify and strive to meet common goals.
Design with Target Audience in mind
The higher up your users are in the org chart the more they want to see data displayed visually.
Flash Specific Formatting Features
Widget formatting Transition animations Change selection on mouse over (for selectors)
Date Selection widget
calendar selector.
Waterfall Widget
highlights the increments and decrements of the values of metrics over time. Can be used for "what if" analysis
Panels
A predefined portion of the doucment canvas that contains any content such as grids, graphs, text boxes, images and so forth. Can be layered to form a Panel Stack Can help guide users through a logical workflow in a dashbaord
Interactive Bubble Graph
Allows users to view a trend of three different metrics for a set of attribute elements One bubble represents a single attribute element. The bubble's position on the X axis represents the value for the first metric. its position on the Y axis represents the value for the second metric. The size of the bubble represents the value of the third metric
Selector Formatting Options
Background color Border Drop shadow font gradient color tooltip Make all items the same width
Microcharts Types
Bar chart - plot a metric with respect to time suing a single bar displaying a metrics current value and historical data to visualize the shape of the trend Sparklines - plot a metric with respect to time using a line graph displaying a metrics current value and historical data to visualize the shape of the trend. Consists of Line graph and a horizontal line which provides a reference between actual and reference values Bullet chart - compare the value of one metric against other metrics typically representing a target value. Ex compare YTD metric value to annual target or forecast Consists of Horizontal performance measure bar (actual metric) Vertical reference line (target metric) and colored reference bands that indicate a numeric range in which the metrics values exist
Flash Widgets available in Interactive mode and flash mode
Bubble Grid Data Cloud Date Selection Fish Eye Selector Funnel Gauge Graph Matrix Heat Map Interactive Bubble Graph Media Microcharts RSS Reader Waterfall
Widgets that can act as selectors
Bubble Grid Data Cloud Graph Matrix Heat Map Interactive Bubble Graph Interactive Stacked Graph Microcharts Time Series Slider Waterfall Weighted List
Selector Styles
Button Bar Check Boxes Drop-down Link Bar Listbox Radio Buttons Slider (used for time series analysis)
Multiple Selections
Button bar Check Boxes Link bar Listbox Slider
Group Like Items or Comparison Items Together
By placing like items close, users focus attention on identifying trends or connections between data Same concept applies to data elements that you want users to compare against each other
Flash Widgets in Flash Mode only
Cylindar Interactive Stacked Graph Thermometer Time Series Slider Weighted List Viewer
Use White Space or thin Grey Lines to Separate Groups of Data
Delineate separate groups of data by making sure that there is sufficient white space between the groups
Microchart display options
Display a bar or sparkline (or both) Display only a bullet Display all three microcharts (sparkline, bar, and bullet)
Heat Map widget (tree map)
Display a combination of colored rectangles each representing an attribute element The size of each rectangle represents its relative weight. The color represents the relative change in the value
When Possible, Limit the number of Metrics
Display between 3 - 7 Metrics have the most visual impact
Data Cloud Widget
Displays attribute elements in various sizes to depict the differences in metric values between the elements.
Graph Matrix Widget
Displays multiple series of data in a combined graph and line chart. Intended to highlight Actuals vs Forecast Dataset is sliced into a matrix of graphs to facilitate quick comparisons across multiple dimensions
Correlation Comparisons show whether the relationship between 2 variables fails to follow the pattern you would normally expect
Dot Charts Bubble charts Dots or bubbles represent each of the different data points Bubble charts enable you to use the bubbles size to show one metrics value and its position on the chart to show another metrics value
One Selection at a time Selector Style
Drop-down Radio buttons
Dashboard Characteristics
Emphasis on graphics Emphasis on KPI's Use of a single screen Emphasis on summary-level data and exception reporting Emphasis on customization
Interactive Stacked Graph
Enables users to see a contribution of an individual metric series to the total. A built in widget controls the displayed series.
After Enabling a widget to act as a selector
Flip to flash mode and specify remaining properties ex graph to behave as a selector
Focus on Important Indicators like Trends, Variances, and Performance Stats
Focus on KPIs that matter most to targe audience KPI's will typically be quantitative measures, trends over time, variances or comparisons, and performance states. May include non quantitative data such as Top 10 lists of customers or project goals
Metrics Dashboards
Focus on key metrics and their status compared to the company's goals. Exception reporting by using graphical indicators to show where metrics exceed meet or fall below targets Target Business analysis or operational managers
General Dashboard Design Recommendations
Focus on the Data First Design with target Audience in mind Keep Related Data on a Single Screen Pay Attention to Object Placement When Possible Limit the number of Metrics Focus on Important Indicators like Trends, Variances, and Performance Stats Group Like Items or Comparison Items Together Use White Space or Thin Grey Lines to Separate Groups of Data Keep Simple Navigation in Mind Limit the Number of Fonts in a Dashboard
Graph Formatting
Transparency effect Curved lines Bevel effects data tooltips
Analytic Based Selectors
Use a grid or graph to control another grid graphs you synchronize data across the grids and graphs
Design Recommendations for Color
Use color carefully and sparingly - Consider the corporate color scheme - Use color intensities versus color hues (different colors) - Use earth toned colors (greys browns oranges, blues, green) - Use the same color for a single measure that displays in multiple places on a dashboard - Use contrasting colors to emable quick comparisons - Use colors to show different levels of alerts - Consider the dashboard in Hard Copy - Use sharp contrasts between font and background fill colors - Consider color blind users
Dynamic Visualization Dashboards
Use dynamic graphing techniques and advanced visualizations (flash technology) to convey more meaning in less space. Users can view more data represented over time spans Include flash specific features like transitions, interactive widgets and sophisticated formatting
Flash Fish Eye Selector widget
Used to scroll through a long list of elements. Look like lists in other non flash views
Time Series Comparisons
Vertical bar charts or line charts If you have more than 8 points of data you should use a line chart Surface area charts for time series comparisons can be used but limit segments to 5 or less. Place the most important segment against the straight base line
DHTML modes
View Mode Interactive Mode Editable Mode
Microcharts in flash mode
View specific graph series values (hover with cursor) Move column of data and microcharts by clicking header and dragging Sort a column of data and microcharts by clicking header to switch between asc and desc (cant sore microcharts)
