How Power Platform/Apps
can help you harness
the power of Data 

What is Microsoft Power Platform? 

Microsoft Power Platform is an integrated suite of tools:  

• Power BI is a robust data analytics tool designed to help users visualise and share insights from their data through a range of different dashboards and reports. 
• Power Apps enables users to build custom applications tailored to their unique business needs without requiring extensive programming skills. Power Apps includes Power Pages, which enables the creation of secure, data-driven websites that extend Power Apps capabilities to external users. 
• Power Automate, formerly known as Microsoft Flow, enables users to automate repetitive tasks and workflows. 
• Power Virtual Agents provides a platform for creating intelligent chatbots that can interact with users in natural language. 

Dataverse is an important underpinning Power Platform technology, it’s a scalable data platform (despite being built on MS SQL, it offers much more than a database as it has data access, business logic, and presentational layers) that ensures seamless data integration and data management across various Microsoft business applications and third-party applications via connectors. It is an evolution of Microsoft’s common data model. 

Collectively, these tools enable organisations to collect data, generate insights, automate processes, and create intelligent applications. They are all ‘low code’ in the sense that they – in theory at least – minimise the need for writing extensive code by providing abstracted user interface options. 

Note; Co-Pilot is part of Power Platform, but I am not going to cover that here. 

Real world use cases, case studies, and challenges 

The Pivotl team are early adopters of the Power Platform in terms of build expertise and client delivery. I’ve long held the belief – even prior to the recent AI hype – that in the not-too-distant future the only coding efforts that will happen will be on the no code/Gen-AI solutions that abstract coding (oh, the irony!).

So, I’ve been eager that we hold Power App expertise in Pivotl to help our clients to navigate Microsoft’s no/low code hype in a safe, secure, appropriate and high-quality way – engineering work still requires engineering skills in my mind regardless of the type of creation tools we use.

We’ve worked on lots of Power Apps projects at Pivotl. A few recent examples include:

• Data research service (internal and external users) deployed to gov.uk for a major government department
• Data and AI tools to streamline vaccination processes for a digital health client
• A data driven service to tackle homelessness for an education provider.

Drawing on past experiences as well, I’ve compiled a list of some of the common challenges and lessons learned that could be of benefit to you if you’re embarking on your own Power App journey (I won’t cover Power BI etc., here because it’s well-trodden ground):

• Development nuances:
  
  
  • Business logic often requires custom, complex formulas that can be challenging to implement and maintain.
  
  
  • Power Apps often run very slowly making development frustrating. They usually perform better in a production environment, but there are still issues. This can potentially be resolved by optimising data retrieval, etc.,
  
  
  • There are limitations to UI customisation, and often the only workaround is to be creative with the designs provided.
  • Not everything that works for Power Apps works for Power Pages, e.g., business rules are different, and aspects of Power Pages rely on outdated technology.
    
• Complex licensing: Power Platform licensing is very complex and difficult to understand. You can use the Microsoft licensing calculator but even then, it depends on your current licensing, so you need to make sure you get support.
  
• Steep learning curve: The learning curve is steeper than internal teams realise, and low code still requires engineering knowledge. It’s important, early on, to consider whether you need a managed service wrapper or if not, how you will bring internal people, often non-engineers, to a level where they could build/run a Power App.
   
• Engineering practices still matter: It is possible to design, build and deploy a power app and to do it badly i.e., without writing tests or using different tiers of environment, which carries a lot of risk if it’s a relatively important service. 
  
• Dynamics knowledge helps: Power Apps and Dynamics are closely coupled and when you need to do something in Power Apps, like sending specific templated emails, be prepared to have to problem solve in legacy parts of the Dynamics system or to have to meddle with obscure permissions in 365/Azure. 
  
• Low code works well until it doesn’t: What I mean by that is that with low code you accept the trade-off that the abstraction of code gives you and that works fine until you encounter something that doesn’t quite work, like particularly complex form logic or validation, and you do then need to be prepared to go to code, but if you take that too far you erode the benefits of low code, so you need to consider if it was the right use case from the outset. 
  
• Standardise: Related to the above point, the best way to adopt Power Apps is to limit the use to standard, core functionality and to not introduce any change that means that you need to write and integrate complex code. This requires discipline and good stakeholder management from a product management perspective. 

On the positive side, Power Apps can enable rapid development if you elect to use the technology for the optimal use cases, and non-developers can learn how to create apps subject to certain constraints and with the right foundations. Power Apps provide seamless integration with Microsoft services, e.g., Office 365, Dynamics, and Azure, which is one of the reasons CIO’s try and maximise the value in their Microsoft subscription.

Power Platform consists of a potentially transformative set of tools that can enhance your organisation’s data strategy. Through integrated data collection, unified data management, advanced analytics, and process automation, the Power Platform can help to unlock the power of data in your organisation and is a good ‘gap filler’ where you need to replace data siloes (spreadsheets, access databases etc.,) or legacy applications that are difficult to integrate, change or maintain. 

Organisations should, however, be appreciative of the constraints of the Power Platform and use them sparingly for use cases that suit its intended purpose; however, I would expect the platform to continue to evolve and become more flexible as AI code generation capabilities continue to improve and the underlying technology matures, so I would recommend careful, sensible adoption as part of a wider strategy whilst being careful to avoid supporting inappropriate use cases and seeing Power Apps as a silver bullet.