Delivering Power Platform Projects: Truths from the Field

I have successfully delivered a number of Power Platform Projects for nearly a year now. Even though I resigned as a Microsoft MVP Awardee after nearly ten years due to the behaviour of some of my ex-fellow MVPs, I still work hands-on with Microsoft products as well as providing more strategic thought leadership to business leaders. The current blog forms a summary of what’s good, and what’s not so good, about working with Power Platform projects.

What’s good about Microsoft Power Platform Projects?

With technologies such as Microsoft Power Automate, business users can get quite far pretty quickly with simple things. I have mentored organisations through a number of interesting use cases with it, such as building a Health and Safety Power App where people can record any potential breaches and execute alerts.

What’s not so good about Microsoft Power Platform Projects?

Licensing.

The Microsoft Power Platform licensing is confusing. Within the MVP Program, and now outside of it, it is really hard to get a straight answer about licensing from Microsoft and I personally got different answers from different people, and then nothing at all. It is hard to get a definitive answer.

Perception

The perception from the Microsoft advertising is that the technology is super easy and point and click. However, you still need to think properly about data. What’s the data structure like? Is it clean data? What needs to be done to get it into a good shape? These questions are the hard questions, and the Power Platform sidesteps it to an extent with the Common Data Model. However, if you are outside of that, then you need to understand that data and technology are two different things. This is not necessarily just a Microsoft issue but it can be hard to undo perceptions until people try it and get very disappointed very quickly. This builds mistrust.

Just because technology is easy to use, it does not follow that the data is easy to understand. Don’t be fooled by easy-to-use technology; data can still be hard.

Compliant or Complaint?

Multiplexing refers to the use of hardware or software that a customer uses to pool connections, reroute information, or reduce the number of users that directly access or use the PowerApps and Flow service.

On the face of it, it looks simple to licence Power Automate, for example. Power Apps does not necessarily look like a lot of ‘moving parts’ to the business users and a low-code/no-code user, who may not really understand that they are using APIs, data sources and a bunch of connectors. Once you introduce Power Apps into the picture, there is potential risk of multiplexing violation. Does each Power Apps user need a license too?

Power Apps is a solution that is designed to access and move data around. This could introduce the potential for someone to inadvertently fall into the multiplexing trap without realizing it.

Takeaways

If I recommend Power Automate again, it is with caveats that the licensing is hard to get right. The licensing is hard to get right and it is hard to get a right answer, and it is too easy to break. I don’t want to put people in that position of potentially being non-compliant and I don’t feel confident that I get good answers. I am likely to suggest it, but with the strong recommendation that the organisation needs to consult their software and licence vendor. I helped check licensing out for one organisation, who were so put off by the lack of answers that they switched to another technology. It was a time suck for them and for me.

I would recommend Logic Apps. Super easy to build, and it costs pennies rather than dollars to run. Much cheaper and worth the investment of time to learn. Instead of Power Automate, look at Logic Apps instead. Logic Apps are hosted in Azure and have a code view rather than a ‘business user coder’ view. They have more connectors and have better monitoring. Plus, they are way cheaper and the billing is easier to understand. It’s a lot cheaper. It also side-steps a lot of the Power Automate confusion, which starts to kick in when you examine it more closely in different scenarios.

SharePoint List are a hellish data source. Any time I have come across them. I have had to rip them out and replace with another solution using another interface with Azure SQL Database and Power BI solution. They are flexible and user-oriented but they do not produce good data behind the scenes. Again, putting the data into Azure SQL Database and then Power BI allows for more Business Intelligence scenarios to be implemented, such as time series.

Summary

I’m available to hire if anyone needs any help with Power BI or the underlying Azure and data architecture which supports these systems. Please feel free to contact me.

Leave a Reply