I heavily requested feature from customers is the ability to drill from Region>Country>State or some variation of this workflow where end users, desire a visual representation of geography at multiple levels. This article explores some of the challenges and opportunities for displaying geographic data within SAP Dashboards across multiple geographic levels.
Disclaimer: Regardless of how you want to approach geo-visualization, it could require some third party data and licensed software.
Common administrative areas (Countries, States, Zips, Etc)
There are over 230 countries in the world, Over 3500 States/Provinces, and hundreds of thousands of postal codes globally. Some of this data is organized and hosted in a public domain, and some requires a a little work or money to obtain. My team is perpetually working to obtain public domain or open source data and making it available to customers.This is one of many problems I get to work on making geo-visualization more readily available.
Your custom regions
Typically, custom regions are crated by merging or manipulating common administrative areas. For organizations that actively manage and use maps and geography as a decision making tool you can easily obtain and re-use this GIS data. For BI practitioners who don’t have access to GIS data, but do have raw data from a data warehouse that describes geographical hierarchies, we built a free plugin that auto-generates custom regions for you.Tutorial for creating Custom Regions
Geo Data Editing and Format
All GIS and location analytics technologies should have the ability to consume various data formats for boundary data. One of the more popular formats, ESRI Shapefiles, is a commonly used format exporting and transmitting geospatial data. Other formats like KML and GeoJSON are widely adopted web service structures for transacting geospatial data. Regardless of the initial data format, geo-data that you obtain can be easily translated, edited, and transformed into the appropriate format accepted by your mapping solution via GIS tools like ESRI ArcGIS or open-source alternatives like Quantum GIS.
Your Business Data
Obtaining your business data is generally the simple piece of the puzzle for creating drillable location analytics. The challenge typically with dashboards and BI apps is the volume of data required to display and drill all possible geographies. For example, if you need zip code level analysis in the US, there are potentially 44,000 records without any additional dimensions. This is a fairly manageable volume of data, but as you add additional dimensions and runtime filtering capabilities you need to carefully and thoughtfully plan how your end users will interface with this location data. For example, in our implementation for state-zip analysis, users can drill one state at a time, allowing us to swap only thousands of records. View Zipcode Drilldown Example / Template
Combining Geo Data with Business Data
Location Intelligence can suffer the same pitfalls as any business intelligence dashboard or report. Data quality and availability are important to ensure the final step is a success. There are several approaches for combining business and geographic data together with SAP BusinessObjects. For tools like Explorer or Lumira, some geodata is packaged with the software for mapping common administrative areas, making it extremely simple to display and drill simple geographic hierarchies. Dashboards provide a limited set of maps that quickly paint you into a corner. Through the SDK, SAP Dashboards also provides the greatest flexibility to create custom drill paths through regions or granular administrative areas like zipcode or county. This is where third party technologies come in like CMaps Analytics (formerly GMaps Plugin) and ESRI integrations via APOS that are specifically designed to combine business data and geo-data into interactive mapping applications.
Where to Get Started
Location Intelligence starts and ends with your end users and the questions that they need to answer. When working with clients here are the first questions I ask to better understand:
2. What kind of decisions or results are you driving from using geo-visualization?
Examples: Where are the best / worst performing sales territories, where are assets being under-utilized, where are assets in relation to proximity?
3. How many layers and geographic levels are required to come to a decision?
Every dashboard will have different requirements, ranging from a single layer of information, a geographic, drillable hierarchy, or multiple layers of un-related data visually joined solely by geography (customers, regions/territories, assets, ec).
From this information you can determine if the goals for your dashboard and location analytics are realistic for a single dashboard / application or if you need to create multiple dashboard pages. All of the dashboard best practices you have learned will still apply when building location intelligence.
I would love to hear about your requirements and needs for mapping. Feel free to comment here, or contact me directly and I can post more examples..
About the Author
Ryan Goodman is Centigon Solutions, founder, global speaker, and BI Apps / location intelligence expert. In addition to leading business and product strategy, Ryan takes a hands-on approach to location analytics during his Tech-Tuesday broadcast, where new business requirements and technical challenges are shared and solved on a live webcast.