by James Townsend
Microsoft has created a business software platform which not only includes complete commercial products such as accounting and CRM, but also development tools which may be used by professional software developers and power users. Microsoft is enpowering power users, sometimes called citizen developers, with tools that require less coding and more configuration through templates and wizards.
All these apps are connected through a shared data model called Dataverse (formerly Common Data Service). Some apps such as Dynamics 365 Sales store their data in Dataverse. Security is managed via Office 365 hosted in the Microsoft cloud. With Power Apps, you can build standalone apps that tap the power of Dataverse.
Dataverse and Power Apps are not the first tools to target citizen developers. Microsoft Access, for instance, was a product that allowed end users to create database applications and connect to SQL Server which was often the enterprise data repository for an organization. SharePoint was also geared toward end users creating sites, lists and other SharePoint objects with or without calling on IT professionals.
This post is not about how to use these tools, but rather how communication and policies are needed to govern information architecture. Without governance, business people may become confused on where to look for definitive data. What happens if there are three independent places where sales or customer data is stored, with no method for reconciling differences?
Here are some examples of areas that require governance:
- Security. Microsoft provides authentication capability and a rich security model which allows fine control over who has access and control over what. This does not mean that managing security is simple or easy. Giving a large number of people full admin rights increases the chance of serious errors and detracts from accountability. Security governance requires detailed processes for onboarding and offboarding employees, for instance.
- Data model changes. Although adding a field to a table is easy, governance calls for a process to manage changes in order to avoid duplicate or ambiguous fields, and to ensure that fields are added to the correct table or tables. Data model changes may also necessitate updates to documentation and training.
- Reports and business intelligence. While reports and business intelligence visualizations and dashboards do not modify data or change the data model, they must be reviewed to ensure that they are correctly showing what they represent. For instance, a sales commission report must be accurate to be valuable.
- Upgrades. Cloud computing has increased the pace of innovation and brought shorter upgrade cycles. These must be managed to ensure that they take place and fix any errors that may come with new versions.
- Resource management. Computing power and storage are now bought as needed. Successful governance will monitor resource needs and costs, archiving unneeded capacity and allowing for growth.
- Monitoring and auditing. You can measure the health of your apps and understand how they are being used through logs and reports. These findings may alert you before you run into a major disruption.