Why Power Platform licensing is complex, part 1: products

It’s hard to avoid this question whenever people discuss the business scenarios for Microsoft Power Platform. The tools and experiences for creating apps are getting more & more streamlined, yet it feels like the complexity of rules on what licenses are needed for each specific scenario just keeps growing. With great powers for designing creative new solutions comes also great responsibility for understanding the commercial factors of implementing them. This means that the solution architecture needs to take the licensing aspects into consideration, even if many technical people aren’t very comfortable with the topic.

The main factors I consider to be sources of complexity in the licensing discussion around Power Platform, Dynamics 365 and Office 365 are these:

• Protection of intellectual property as well as development budgets across Microsoft product portfolio
• The old world of apps as data silos & the concept of multiplexing
• Light use scenarios for information workers vs. “Real Business Applications”
• The role of CDS (now: Dataverse) as the platform, not just a data source/target

I’ll cover each of these dimensions in a separate blog post, to avoid building up too many stress factors into a single post and scaring away people. None of these obstacles are insurmountable, instead I believe raising awareness of them will make it easier for everyone to answer the question of “why” and move on to the “how” part of planning the actual solutions and business cases for leveraging the full potential that lies within the Power Platform.

Let’s begin from the complexity source nr. 1: Microsoft’s product portfolio structure.

Protecting the products & investments

XRM, the predecessor of CDS and partially also Power Apps, was never sold as a platform. The business model was built around selling a suite of CRM applications, which you could then customize and extend to cover business processes not originally within the scope of the product shipped by Microsoft. This meant that a lot of money was left on the table because customers that weren’t looking for a new CRM system didn’t really have much reason to consider XRM as a platform their other business apps needs. From the perspective of Dynamics CRM as a product, there wasn’t much threat of competing apps within the MS ecosystem, but unfortunately it also meant that investments into the core platform weren’t likely to grow very fast if the Dynamics offering was to remain outside the mainstream product portfolio of Microsoft.

Now that Power Apps is offered specifically as a platform where your citizen developers, pro developers and ISVs alike can all come and build their apps, the upside to everyone is that the common infrastructure for data and application management is evolving crazy fast. The new problem that arises from this is the battle of the internal buckets where money should be flowing in. Unless all customers unanimously purchase both the platform (Power Apps premium licenses) and the first party applications (Dynamics 365 Enterprise Apps), someone could easily consider their bucket to have a leak caused by these new buckets. Cannibalization of revenue is a rational fear to have when there is overlap between product offerings within the organization.

The way how Power Platform has been positioned as the extensibility layer for both Office 365 and Dynamics 365 is undoubtedly one of its key strenghts from the business perspective on Microsoft level. PP is everywhere because so many customers already have it, underneath their existing information worker tools and CRM systems, whereas low-code competitors like OutSystems or Appian will need to find a way to get their foot in the door somehow. However, this mechanism of having “seeded” Power Apps and Power Automate features inside other products just delays the commercial discussion to a later point in time. “Why do we need to buy these new premium plans if we already have the Office/Dynamics features?”

Before Power Apps merged with XRM, what we know call “Canvas apps” were introduced to the markets as a technology that could easily work with the ubiquitous Office 365 services as data sources. The same with Microsoft Flow. I believe this really was more of a marketing positioning strategy rather than a deep architectural commitment from MS, since many experiences weren’t really that seamless once you got down to working with SharePoint, for example. Still, this approach made it possible for the community of citizen developers to start gathering around Power Apps, when power users all over the world discovered how they could go beyond Office 365 standard capabilities with these no-code tools – while still remaining within the existing corporate licensing of Office tools offered by IT.

Had Power Apps and Power Automate (Flow) remained there as just tools in the Office waffle that come with the subscription, we would have likely seen only a tiny fraction of the features that now are rolling out into Power Platform. After all, if you’re not paying any additional licensing fees for the products, then are they really products in themselves at all? “Freebies” like Sway or Bookings that we’ve seen appear in the cloud service nowadays known as Microsoft 365 are examples of what you could expect from technology that isn’t large enough to warrant a SKU of its own in the MS product portfolio. Not a very lucrative position for any ambitious app team in the long run.

There’s this dilemma that if you raise the barrier of entry too high for a product, the adoption curve is going to take a hit and you’ll miss out on the viral effect of “free” software. At the same time, if you can’t directly tie the usage rate into a visible stream of new money, it’s difficult to collect funding for product development investments beyond the initial hype and excitement of launching something new & cool. In the enterprise software business you can’t even monetize the users by exploiting their data and eyeballs for selling ads, because privacy and confidentiality tend to be bigger factors here than in the consumer market (where unfortunately everything is moving to “pay with your data” business model). Just look at the example of Google, who did build up their own App Maker product into the G Suite, then saw too low usage numbers for it and ended up killing the service and customer apps built on top of it. Now they’ve acquired the low-code application development platform provider AppSheet and are trying to gain a foothold in this growing market.

While this explains why Microsoft couldn’t just keep the doors open for ever more advanced Power Apps development on top of exisiting Office 365 subscriptions, there’s also the other direction to consider. At some point these apps become so advanced that they’ll start to challenge the Dynamics 365 applications formerly called “CRM” or “Customer Engagement”. These commercial apps from MS are nowadays just called “model-driven apps in Dynamics 365”, which further underlines the role of Power Apps Model-driven Apps as the infrastructure that makes these apps tick. They are preconfigured instances of applications on top of CDS, but they are not the same as the platform itself. Therefore, starting from October 2019, buying a Dynamics 365 Enterprise App no longer gives you an Access All Areas pass to using apps in just any vanilla CDS environment that hosts a custom app for process X. You’ll need to be wearing the lanyard that specifically says “Power Apps” if there’s not Dynamics 365 applications installed in the environment you want to enter.

In short, buying the Dynamics 365 product doesn’t give you the Power Apps product, just a “seeded plan” to work within the boundaries defined by the application context. The same holds true the other way around: buying a Power Apps Per User license doesn’t grant you the rights to use everything that’s running on the platform within the tenant. Similar to how ISVs (independent software vendors) would charge you a fee to use their application features via their own licensing mechanism, Microsoft is setting up mechansims that will control who can access which App Module.

Regardless of how that might sound like initially, I believe it could actually be a good thing in the long run. The reason being that the way how Microsoft has previously attempted to limit access to the schema and data of CDS by drawing lines on restricted entities available only to Dynamics 365 license holders (aside from data read operations) isn’t really going to be a sustainable model. In his third chat with MVP Steve Mordue, the Corporate Vice President of what MS internally refers to “Citizen Application Platform”, Charles Lamanna, has stated that an alternative licensing model is being prepared right now:

“So there is something in the works that we’re working towards and I would say at a high level, restricted entities as a concept are largely antithetical to our common data service, common data model and vision. And they were just like the least bad option to go make sure that we appropriately can license Dynamics apps. So we are working feverously on many proposals to get out of that restricted entity business, but still have a model which more appropriately captures and protects the value of the Dynamics apps without introducing restricted entities.”

I’ll dive deeper into the whole Common Data Model (CDM) discussion in the next part of this blog series. In general, what I believe to have been a big barrier for the licensing options available to MS work with is the amount of effort needed in making the application/platform separation initiative a technical reality. It has surely been also holding back many teams in terms of how a specific Dynamics 365 App like Sales can deliver features that differentiate it from the sea of business applications that are to be built on Power Platform, by citizens, by customer development teams, by ISVs. A level playing field is needed for the future solution ecosystem to bloom, yet in the short term it requires work behind the scenes that doesn’t surface as application features directly.

The fact that such investments and also compromises have been made during the past couple of years is a clear signal of how Microsoft believes the demand for its different services will evolve & how they will be competitive because of their dedicated low-code application platform product offering. Customers will continue collaborate via Office 365 productivity tools, likewise they’ll keep on adopting readymade business applications from the Dynamics 365 family of Apps for scenarios like omnichannel customer service. There in between lies the opportunity for 450 million new low-code apps to be built within the next 5 years, as Charles Lamanna states in the recent CNBC article “Next frontier in Microsoft, Google, Amazon cloud battle is over a world without code”.

It is not a market that the Office products could easily rise to cover, nor is this bottom-up innovation a natural fit for the Dynamics way of delivering top-down enterprise systems. A dedicated product offering is needed here, which is why the Power Platform exists. Since Microsoft has chosen not to follow the “buy” path of Salesforce, Google and other competitors, but has rather adopted a “build” strategy to create their low-code application development platform from in-house technology, many of the elements in this platform have already been used somewhere else, and as a result also commercially licensed for scenarios that predate this new low-code era. This is the reason we’ve seen so frequent changes in the licensing model across Power Apps and Dynamics 365. It is a source of licensing complexity that is difficult to avoid when the different apps don’t live in their dedicated silos but rather share so much of the common platform capabilities.

Microsoft talks a lot about having three clouds: Office 365, Dynamics 365 and Azure. Where Power Platform fits in is right smack in the middle, and with plenty of overlap in relation to the existing clouds. It’s a new “aPaaS” layer added between the SaaS apps of Office & Dynamics and PaaS/IaaS offered by Azure. As a result, the relative position of existing products needed to shift a bit, which created a before/after dimension into both how Microsoft envisions each cloud to be used and as well as how they commercially offer the services to be licensed.

Other parts in this article series can be found here:

• Why Power Platform licensing is complex, part 2: multiplexing
• Why Power Platform licensing is complex, part 3: light use scenarios
• Why Power Platform licensing is complex, part 4: the role of CDS