Inscrivez-vous ou connectez-vous pour rejoindre votre communauté professionnelle.
5 best practices for creating effective dashboards:
1. Choose metrics based on why they matter: Choosing what metrics to include in the dashboard is critical. Above all, they must be metrics that matter and that are relevant to the job at hand. But that doesn’t mean every metric should be included – far from it. You should be highly selective in determining which metrics earn a spot on your dashboard.
2. Keep it visual: Dashboards are meant to be fast and easy to read. Report and text-based tables are not fast or easy to read. This is a case where a picture really is worth a thousand words. Because the human brain processes a single number, visualization or a picture as single “chunks” of information, a report or data table filled with numbers requires the brain to store and remember multiple chunks while visualizations or pictures require single chunks. So the process of comprehension and insight is dramatically faster with visualization.
3. Make it interactive: Your dashboard will put everyone who sees it on the same page. But once on the same page, viewers each will have their own questions and areas where they want to know more. Your dashboard needs to allow viewers to customize it so that they get the information they need.
4. Make it simple to access and use: Making your dashboards easily accessible is critical. Web distribution is ideal – especially if your dashboards can constantly pull current data and can adhere to IT protocols and security standards. Products that provide web-based sharing and collaboration of data analytics, dashboards and visualization, are available and are easy enough for even someone with limited technical abilities to support.
5. Keep it current or don’t bother: Make sure that the data underlying your dashboard is current and that your selected metrics reflect current business challenges.
6 Mistakes to avoid:
1. Starting off with too much complexity.
2. Using metrics no one understands.
3. Failing to match metrics to the goal.
4. Using ineffective, poorly designed graphs and charts.
5. Underestimating the time or resources to create and maintain the dashboard
6. Cluttering the dashboard with unimportant graphics and unintelligible widgets.
Keep it simple at face valuel, while layer your dashboard to allow dynamic changes. Allow the user to explore and learn things. Make sure to include information boxes, never leaving ambiguity in the overall presentation of your dashboard.
A simple reference point would be time... you can use time as a base reference while changing different aspects of the same point.
People love visually pleasing work. Don't try to include too much wordy information, let your infographics do the talking.
Dashboard should be built like End User should use its features completely and easily understandable. Should be User friendly and by using proper Metrics.
The hallmark of a data-driven dashboard is the ability to see and understand data at the speed of thought. Well-planned dashboards will allow both business leaders and knowledge workers alike to ask and answer questions in real-time, turn insight into action and inspire true innovation.
Too often mistakes are made in the actual creation of dashboards, many times at the expense of efficiency, accuracy and hours of lost time.
you create the dashboards that are just right for your work, your team, and your company
1. choose metrics
2. keep it visualize
3. interactive
4.easy to understand
5.make it simple
1. Always bring in the fields that are needed, while accessing the data from various datasources, always bring in the columns that you need thereby reducing the size of QVD's
2. Date Formatting: If the date format coming form source is datetime and we dont require the timestamp, it is recommended to remove the time stamps to reduce the overall space.
3. DataModel Design: While designing datamodel, we should try to create a denormailzed star schema. The advantage is less RAM consumption, less data load time, better understanding of the data model attributes avoiding null values. Some of the more advantages are link tables, master calendars or canonical calendars
4. Avoid using resource heavy expressions: Always try to avoid using nested IF. Instead build the expression based on set analysis. Make use of Pick function where every possible to replace IF.
Function like InervalMatch should be avoided
5. Standard layout: Always try to follow a standard layout while designing the dashboards. KPI’s can be placed on top and any outliers should be marked. The filters that are referred often should be placed in listboxes whereas the other filters can be in the form of multibox.
Make judicious use of containers and try to limit the data in the charts to top vales.
Don’t make dashboards which are filters heavy. The dashboard layout should be as simple as possible so that users do not need to make multiple selections to arrive at a metric.
Expressions that are used multiple times should be placed in variables and try to use parameterized variables for YTD,QTD, MTD calcs.
1. Use the right type of chart
2. Even distribution of all the dashboard element.
3. good layout choice
4.Providing the context
The dashboard should be
1. Easy to Understand.
2. Less Complex.
3. User Interactive.
4. Proper showcase of the metrics.
5. Properly used KPI's.
1. Simplicity
2. Readability
3. Focus
in a simple state. keep it simple with direct KPI, and clean visualization
1. Make it more lively with different formats/types in the dashboard. 2. Decide on what kind of KPIs the business need. 3. Always set a default filter on the dashboard so that it wont retrieve the entire data every time it runs.4. Try to populate many records on the single click so that user don't need to do lot of clicks.5. Give all the options to the user like Download, export options of different types and formats.
