9.2.3 Business Intelligence tools


TABLE OF CONTENTS
Offline solutions
If you need to link information from different databases or datasets, a good solution is to use the offline version of a business intelligence tool like Power BI, Tableau or QLIK, which makes this process relatively straightforward. If you do not wish to use a business intelligence tool and are at ease with Excel, you can use functions such as vlookup() or the combination index(match()) to link two datasets together. These solutions only work if you have set up a unique ID in your datasets!
Microsoft Power BI is a powerful analytical platform that provides the user with tools for analyzing, visualizing, and sharing data. To go further please refer to the Power BI User Guide that presents the different ways of:
- connecting to various sources
- cleaning the data structure, and
- creating the report for further publishing to the web and sharing with colleagues.
You can refer to Acaps beginner tutorials for Power BI and Tableau that are guided exercises on datasets.
If you’re interested in learning more about Power BI, we also encourage you to discover the Power BI workshop that was conducted by CartONG during the last edition of the GeOnG Forum, and which is available on our Youtube Channel!
Online solution
Another common choice is to set up an online dashboard that connects in real time to the Mobile Data Collection website/platform through an API (Application Programming Interface), in order to present the data from a in a more enticing fashion – this makes sense if your MDC tool does not have sufficient analysis features for your needs, and if you wish to have an online analysis tool that can easily be shared with other stakeholders.
You can look at this video of Humanitarian Data Solutions on how to Load KoBoToolbox data to Power BI via API connection.
The most common examples of this are Power BI (see example below) and Tableau. You can set up the analysis on a desktop tool and then publish it online if you like. This is particularly relevant for simple surveys or when you have different data collection instances concerning the same entities (households or infrastructure for example). If you have more complex surveys (with different languages, with advanced mapping needs etc), or if you would like a very specific design, this kind of tools can be quite time-consuming to set up.
Another possible solution is to set up a dashboard from scratch using libraries that are available online (D3 for example). This solution allows you to have complete control over the content and graphic design of your dashboard. However, this requires contracting a webdeveloper if not available internally, with consequent impact on the budget for the project.