At the virtual Microsoft Ignite 2020 event I had chance to join many fellow Finns in our FIgnite session and spend 5 minutes on highlighting a topic that I considered to be important in the MS ecosystem right now. It turned out to be an easy choice to make.
Project Oakdale (briefly known as Dataflex) is bringing the low-code/no-code functionality from Power Platform into the hands of pretty much all the information workers leveraging Microsoft 365 tools in their day-to-day job. This is not so much about any single technical feature as it is a change in mindset about who the app makers should be. So, I set out to explain this my 5 minute presentation titled How Power Platform Now Empowers All Teams Users.
https://www.youtube.com/watch?v=Dj_IH4yq1dE
This is all part of the larger Microsoft Teams as a platform story that I covered in my previous pre-Ignite blog post. It’s no surprise that MS has created much more content around this topic than you could have ever expected to see on Power Apps or Power Automate as independent technologies. It’s a sign of the size of the audience to which this “no cliffs” application platform message is now targeted at.
Will the inclusion of “CDS Lite” in the Microsoft Teams subscriptions prove to be a gateway drug that makes Common Data Service a household name in Microsoft 365 customer organizations? And if so, what will the actual brand name be for Project Oakdale once wre’re past the preview phase? Time will tell!
2020 became the year of #WFH (work from home) and for many organizations also the turning point when Microsoft Teams became the primary place where being “at work” happens. This is accelerating the evolution of Teams from being merely a communication tool that connects human beings into a foundational service layer for many types of business applications.
How the concept of Teams as a platform contrasts with Microsoft’s Power Platform suite of technology is something I’ve been thinking about a lot lately. In this post I’ll first reflect on the relatively short history of where Teams came from. I’ll then examine how the recent feature announcements are brining apps front & center in Teams. Finally, a few words on the possible future for Teams as part of Microsoft’s broader strategy.
The road that lead to Teams
Looking back ~10 years, the real-time communication & instant messaging tools from MS seemed to be going through an endless renaming cycle: from OCS to Lync to Skype for Business. The core feature set presented to the end user didn’t seem to evolve nearly as much as product branding did. On a broader level, the communication activities of information workers within an organization still typically took place within Outlook’s inbox, and different servers like SharePoint and Dynamics CRM all packed their own features for posting short messages to other users.
4 years ago, when the first images of what was then called “Skype Teams” started to leak out, we were already waiting for MS to create something a bit more ambitious than just another online meeting tool. Office Groups had began to emerge in various different places inside the MS Cloud, but they were primarily a technical construct with no sensible UX for everyday people to approach them. Even Dynamics CRM had it’s own solution that attempted to bring together the dicussion, calendars, notes, documents and team memberships from under an Office 365 Group associated with a record like account or opportunity:
I remember having many discussions with our CRM customers where I attempted to steer people away from deploying this Groups solution. Instead I wanted to encourage them to wait for something a bit more polished that I knew had to be on it’s way sooner or later.
At one point there was a clear & present danger of another “Yammer moment” taking place, as Microsoft was reportedly quite serious about their plans to acquire Slack. In retrospect it was a blessing for both parties that MS decided to keep investing in building their own product, instead of trying to retrofit an established service like Slack into their existing software offering.
I would argue that this “build over buy” strategy which Microsoft has since then followed across their business software stack has been a key success factor for BizApps in particular. It has enabled MS to move from merely chasing CRM competitors like Salesforce into redefining the business apps playing field with Power Platform. There’s a stark difference between acquiring companies and bundling them as “X Cloud” versus engineering your own software stack to act as a true platform.
Teams: the collaboration chapter
Initially the first version of the Microsoft Teams product that became generally available in Spring 2017 was pretty much focused on being three things:
Replacement for Skype for Business
Alternative to Slack
UI layer for Office 365 Groups
From a business applications perspective there wasn’t all that much you could do to hook Teams up with Dynamics 365, until Fall 2018 when the previews for the first integrated features were launched. In particular the integrated file sharing experience that Teams offered seemed almost like the Holy Grail for many CRM professionals, offering to fix the glaring hole in the SharePoint integration story that lacked any security model synchronization. The roadmap image below presents the plans from 2 years ago on how Teams and Dynamics 365 were going to be integrated:
The last item on the roadmap has still not been delivered, which is the visibility of Teams conversations inside the Dynamics 365 record form. Why this hasn’t been a higher priority for MS to implement seems to me like a sign of how Microsoft Teams is nowadays positioned as the primary UI for all information work. MS probably would prefer if everything always started from inside Teams. You pin record tabs into channels, you show previews of records inside teams discussions, you interact with records via bot interfaces and so on. As long as Teams is that big umbrella under which all work takes place.
The lack of a deep 2-way integration does not therefore mean that investments aren’t being made into the products involved. It can simply be a reflection of the new vision that is being built, by aligning many existing services to form a whole that aims to be greater than the sum of its parts.
As an example, if you look at Microsoft’s task management story, you’ll see that features and data from across various apps like To Do, Planner and Outlook tasks / flagged emails are currently being collapsed into a central location that is the Tasks app for Teams. Tasks as a generic construct don’t necessarily need to be fully controlled by a single database, yet they very much need to be logically represented within “the hub for teamwork” that Teams is positioned as.
Going forward, when new apps appear into the MS cloud product portfolio and they need to offer task management features to users, the logical integration point to focus on would be Teams. For activity feed type of functionality the choice is even more clear for product development: choose to piggyback on Teams instead of inventing yet another stream of short messages.
Teams: the platform chapter
Moving beyond simply integrating Teams with products X, Y and Z, we’re now seeing the rise of a model where apps are built specifically to be used in Teams. This has of course been possible for a long time already, by developing custom web services and using the SDKs. Now there are many features coming up that will amplify the platform story around Teams on the no-code/low-code front specifically.
Microsoft Lists app has been the first to reach GA and offers an ultra low barrier for users to process data in a single table through a configurable, readymade UI. When accessed via Teams, the list data gains one more special dimension: discussions to be had regarding a list item. This is pretty much the same as the usage pattern offered for a Dynamics 365 record with the integration mentioned earlier.
Underneath the new covers of MS Lists is the technology familiar from SharePoint lists. If we were to only examine the UI layer, there is actually a remarkable similarity to a popular no-code service called Airtable. So much that the accusations of MS simply copying the visuals and core features from this competitor don’t seem entirely unjustified.
Comparing these two offerings gives us some perspective on what exactly is the market position these tools are aiming to conquer. Simple lists themselves are not a particularly unique feature, rather it’s the team collaboration capabilities and ease of data sharing that turns these tables into what you’d call an actual app. Incidentally, just this week Airtable announced they were building a full platform with apps offering JavaScript based extensibility, a marketplace for sharing apps, automations for executing business logic, and finally a sync service to transfer data across environments (“bases”).
Collaboration scenarios around semi-structured data like lists and Excel style tables can be seen as a gateway drug. They allow turning email or paper based manual processes into a quick first draft of what the digital process could be like. If there are indeed clear business benefits in automating the said process, the requirements for more complex app features will soon begin to emerge from the user base. Hence the collaboration platform should offer an obvious path to grow these pre-built app experiences into more advanced no-code/low-code apps.
Project Oakdale a.k.a bringing CDS to Teams
If Microsoft Lists is the equivalent of an Excel table within the Teams context, then Project Oakdale / “CDS Lite” could be though of as bringing SQL Server inside Teams. Now, obviously Microsoft has zero intent on actually replacing Excel nor SQL with features built into Teams. They only need to introduce those parts that make sense from a team collabocation perspective.
Microsoft Lists is a far cry from what a real Excel workbook can do, yet it can offer much more value in a collaboration scenarios that those lone .xlsx files ever could. Similarly, the version of CDS that will very soon be available for building Power Apps within Teams is nowhere near as powerful as the services powering enterprise CRM systems like Dynamics 365 (or the raw power offered by SQL). Still, the fact that it can be found from within every team and used by a much larger audience than what Power Apps citizen developer tools could hope to capture – those are the factors that can truly make CDS a mainstream service that most information workers in the Microsoft 365 cloud interact it.
The experience of defining the CDS data model in Project Oakdale will be very different from the path that Power Apps makers have gone through – let alone the XRM veterans. In fact, you could easily mistake the table design and row entry UX to be that of Microsoft Lists rather than CDS. This highlights a key aspect that not all Power Platform experts may yet have grasped: for MS this “CDS Lite” is not so much about deciding what premium features of the full Power Platform to give away for free to Teams subscibers – rather it’s about how to best simplify the enterprise CRM features of CDS into a new product that Teams users could adopt on their own.
This doesn’t mean that Microsoft Teams should be viewed only as a mechanism for MS to scale Power Platform to the masses, by “dumbing it down”. If the app platform story of Teams plays out like it ought to, there should also be clear benefits from it to enterprise business applications development.
Capabilities like messaging, notifications, task management, documents or group memberships are not something the Power Platform tools are very good at. For historical reasons there has been the need to build standalone features into XRM for these type of common requirements found in business application scenarios. For the future generations of apps being created, it’s easy to see the benefits of having these non-core capabilities offloaded onto a platform more suitable for managing them – meaning Teams.
It doesn’t really even matter if the feature set offered by Teams couldn’t cover all the deep business logic integration of native Dynamics 365 functionality. Ultimately it’s not about supporting the system-of-record legacy but rather encouraging the new low-code scenarios that will generate 100x more apps onto the platform.
Teams is the new Windows?
The concept of an operating system is something many of us may relate back to the origins of the Personal Computer era, even if OS’s of course have existed far longer than the IBM PC. Windows was the first runaway success in the OS space when it comes to both awareness and commercial results, shaping the fate of Microsoft for roughly three decades. Then along came the era of mobile computing and Android & iOS took over in the number of devices running them. MS could no longer hope to regain that position so they decided to take over a different layer in the computing space: the (business) app cloud.
Azure has been called “the world’s computer” and this does offer some perspective on how the computing concept has evolved since the PC days. Still, Azure is not something most people will ever interact with directly. To remain relevant in the decades to come, MS needs to have presence in the minds of the end users, too. Now that Windows has become merely an optional part of the modern computing stack, it would be pretty darn critical to gain a strong enough foothold on a level that’s above the traditional OS but still below the individual apps. A platform that spans across all the devices people in the business world are using.
Teams is now the closest thing that Microsoft has at its disposal to transform into an OS style fabric that connects a significant share of information workers globally. Nothing like the glory days of Windows, of course, but we should expect to see very conscious steps from MS to further the goal of Teams becoming more OS like. The place where the user interacts with a multitude of apps, share their work context with those apps, a “service bus” for the various apps to exchange data with one another, and finally a unified communications channel for notifications and messaging.
It’s still a long, long way to go for this type of shift to happen where the collaboration tools become the true center of gravity for the multitude of other apps and services that people use today in their #WFH offices. Personally, I can’t live with the limitations on multitasking that the MS Teams content embeds enforce upon the user and much prefer a collection of separate browser tabs to freely switch between. Nevertheless, it’s not an entirely crazy though that the resulting congnitive load for the user isn’t everyone’s ideal way of working. Which means organizations need to be looking for ways to optimize the employee experience via common information work hubs like Teams.
Dion Hinchcliffe has written an excellent analysis on Microsoft’s platform strategy with Teams, where he talks about seeing Teams as the operating system for work. While it may indeed be difficult to get the current owners of the collaboration tools in customer organizations to accept the business app side of Teams onto their plate (especially now with the #WFH boom and its unexpected requirements), the perspectives may change when the time is right from both the technical capabilities side as well as the organizational targets. In the same way as the MS CRM foundation evolved into a key element of the broader low-code application platform known today as Power Platform, the barriers between collaboration tools and business apps should not be perceived to be carved in stone.
There are a lot of benefits in taking long enough holidays during the summer, like many people here in the Nordics do. Going offline for a while is a great way to reset your brain and force-close all those open browser tabs of the mind, to free up memory capacity for properly engaging in brand new tasks.
If you were lucky enough to set your vacation time between July 21st and August 11th, you managed to dodge an unfortunate episode in software product launches from Redmond. Here are the key dates:
July 21st (first day of MS Inspire conference): a new brand name “Dataflex” is announced, with Common Data Service renamed as “Dataflex Pro” and a new subscription tier “Dataflex” promised for all Microsoft Teams customers. (My blog post about it is still here.)
July 29th: the name “Dataflex for Teams” is used for the first time to reference the previous “Dataflex” tier that’s bunlded with Teams, in an effort to make this product connection more clear.
August 11th: all references to “Dataflex” in all Microsoft websites are taken down. The announcement blog posts are brought back online with the terms “Dataflex” and “Dataflex for Teams” replaced with “Project Oakdale”, and “Dataflex Pro” now reverted back to “Common Data Service”.
What happened? The answer is: litigation from the owner of the trademark “Dataflex”.
Here's why Microsoft has felt the pressure to remove the word "Dataflex" from its websites, as (rightfully) claimed as a trademark infringement by Data Access Corporation: https://t.co/s2RXCPTl5Ypic.twitter.com/F0fufjBTPo
— Jukka Niiranen mstdn.social/@jukkan (@jukkan) August 11, 2020
There already was an established software product with the name “Dataflex” out there – in fact it had been around for 4 decades already. Immediately upon the launch day of July 21st several community members were quick to point this out, as indeed it was in practice impossible to miss this brand name overlap when using a search engine. Why Microsoft decided to proceed with their plans on renaming CDS into something like this is something we’ll never know for sure. Some had initially speculated that perhaps there was a preliminary agreement in place with Data Access Worldwide to transfer the ownership of the trademark, but based on the immediate legal action that the company understandably took to defend their rights, nothing like this was likely ever discussed. Our conclusion must therefore be that Microsoft knowingly and intentionally attempted to change the name of its existing technology (Common Data Service) into something that was already in use within the same domain by another software company (Dataflex) without considering the consequences.
The end results is almost like a story taken from The Onion. Even if we are living in 2020 right now and the world appears to go mad every few days, it’s still an utterly bizarre episode in software business. I can’t recall seeing anything killed this fast after it’s announcement, which in the case of Microsoft Dataflex was only 21 days. Neither can I think of an example where something with such a clear potential for software product naming conflict would have been attempted. It is indeed similar to as if MS would have launched “PlayStation 6”.
It’s of course not unusual for Microsoft to pull a product name between its announcement and the planned GA. Remember Dynamics 365 Business Edition, for example? Just because we see something presented on slide decks aimed at MS Partners as guidance for building their business around it, that doesn’t guarantee it will ever see the light of day. Shift happens, plans change and a lot of the partners have already become accustomed to this risk as simply the cost of doing business in this ecosystem. This time the product itself is not pulled, though. The need for coming up with a new brand will most likely delay the commercial roll-out of CDS features for Microsoft Teams based apps, but it’s important to keep in mind that this is all still happening. CDS will most likely become a mainstream data platform to replace use cases where SharePoint Lists have previously been used.
SkyDrive is another example of a Microsoft product brand that suffered from the conflict with an existing trademark and had to be renamed OneDrive as a result. BSkyB as a media broadcaster and digital services provider wasn’t entirely in the same business as Microsoft with it’s personal cloud storage that was following in the footsteps of Dropbox, yet the end result of that case showed that at least between large enterprises such near misses with strong brands can become costly. Even features of a product like the “Metro UI” with Windows 8’s modern user experience had to be avoided due to objections from German reltailer Metro AG. Both of these are examples where it wasn’t obvious to an outsider if a trademark was actually violated, though. With Dataflex the evidence was in plain sight for everyone.
Lucky for us, MS had not yet made any public preview or a commercial subscription of the new Dataflex branded services available by the time they had to call the whole thing off. It was only the early adopters in the partner field and the passionate community members a.k.a. #PowerAddicts that lost some of their work hours due to the branding incident. Yeah, I had to go back and edit a number of web pages, rewrite presentation decks, revert our service documents etc. to recover from the erraneous information that had been published on July 21st. The majority of people out there in the MS ecosystem probably didn’t yet even get a chance to understand what Dataflex meant or how it was different from CDS, so they will have dodged the bullet when it comes to loss in productivity.
There is, however, a much greater loss here than the “Find and Replace” activities needed for changing product names. It is the loss of credibility that Microsoft’s Business Applications specifically suffered from this very unfortunate product naming incident. The worst thing is that there has not been any explanation presented to the genereal audience on what’s going on, rather we’ve just seen the word “Dataflex” disappear and get replaced with “Project Oakdale” or CDS. Microsoft at this point is facing legal action from the owner of the Dataflex trademark, which we all know from the many Hollywood movies we’ve watched that “anything they say can and will be used againts them in a court of law”. It is at moments like these where it would be crucial for companies to have a voice, to be able to reach out to their most important audience (customers) and explain what happened & what are they going to do about it.
I can tell you that it would honestly be a full time job keeping up with A) product naming and B) product licensing of Microsoft Business Applications. It’s not full time for me (luckily), but since I also do it as a hobby / community activity, I can say I’m pretty well educated on what the “right” answers in this area would be at any given time. For anyone who actually IS covering the actions of Microsoft full time in a professional manner, Business Applications is an area they usually openly admit to struggle in.
Power/Dataflex naming best guess as of today:
Dataflex -> Project Oakdale
Common Data Service -> Dataflex Pro -> back to Common Data Service
At the end of the day, what matters to each and every one of us the most in this MS ecosystem is how the end customers react to what they see & hear. While the marketing machine at Redmond surely is capable of capturing the attention of IT professionals worldwide with their news and campaigns, what it does not seem to be very good at is explaining “what does this really mean to you”. As the cloud platforms grow and become an ever larger pool of individual apps that form a spider web of integrations underneath the user interface, the need for someone to cut through the marketing speak and explain the customers what’s really happening here seems to be growing exponentially. What we all could do with the Power Platform is simply magical, yet the understanding of A) through which products, B) based on what technology, with C) what levels of effort and D) for what license cost – this level of understanding is lacking pretty much anywhere I look. I’m not talking about the “how to” instructions of building apps but rather the higher level discussion of where business decisions get made. It is here where the damage from confusing messaging is, in my opinion, holding this ecosystem back.
To close things off for now, as we wait for the new product names that Microsoft has said they will be giving to the technology briefly known as Dataflex, let’s recap where we are right now. “Project Oakdale” is the codename that is for the time being used to refer to the “Lite” version of CDS that will be made available to all Microsoft Teams customers at no charge. Common Data Service is what we will continue to use as the term (for now) when talking about the full capabilities of data management, entities, solutions, Model-driven apps and many other advanced features in Power Platform. This post on the Power Apps blog (originally from July 29th) is still valid in every other aspect, apart from that one “D” term that got removed on August 11th.
It’ll all work out in the end, I’m sure. This time next year there will probably be a whole number of other product names that we’ll be using when discussing the latest turn of events in the crazy & exciting world of Microsoft Business Applications. Dataflex and Oakdale will live on as one of those trick questions in the pub quizzes run by the community. No one died as a result of this episode.
In a way, us #PowerAddicts simply got a taste of what the BizApps whirwind must at times feel like for the end customer who isn’t involved with this technology of ours, all day, every day.
Microsoft dropped a big bomb this week with their Dataflex announcement. I don’t think I’ve ever had as many notifications on my social media apps as there were after I posted my two Dataflex articles on this blog and our company blog.
Knowing that it wouldn’t be a simple topic for outsiders to grasp, with the many different dimensions that Microsoft’s various product teams would use in their own messaging, I wanted to make sure there was at least one article out there that would explain what Dataflex means for Teams, Power Apps and Dynamics 365 in the big picture of business applications. It was great to see that this explanation I came up with was also adopted elsewhere in the mainstream tech media:
Here are some of the topics that the community has raised up since the July 21st announcement, based on the still fairly limited amount of public information that is out there on the coming Dataflex and the rebranded Dataflex Pro (formerly CDS).
Yeah, about that name…
For the small minority of techies who have actually heard of the Common Data Service (or XRM), the need for inventing a new name for their beloved service that remains the same (i.e. Dataflex Pro) wasn’t very obvious. For the larger crowd that works with Office tools and Microsoft Teams, Dataflex is a brand new thing, which understandably has made MS think whether a brand new brand would also be useful at this point.
Unfortunately the name “Dataflex” isn’t so unique that there wouldn’t be some clashing with non-MS products out there. In this case, there has been an existing trademark within pretty much the same application development domains since the year 1981 already.
What do you think about the following post from the owners of the #Dataflex trademark? https://t.co/WAPd9UA6Ut some people say the new #CDS name is ”Microsoft Dataflex” if I started using ”Microsoft nz365guy” I am sure I would get a letter from a lawyer. What do you think?
21 days after launching their Dataflex product name, Microsoft had to cancel their original plans due to litigation from the lawful owner of the Dataflex trademark (Data Access Worldwide). Read this blog post for more details: There never was a “Microsoft Dataflex”.
Why not call it “Power Dataflex” or “Power [anything]” then? Wouldn’t that have been better in line with the whole Power Platform story, as well as reduce the chances of a legal dispute? Probably so, but it’s important to understand that this Power thing actually isn’t the only purpose that Microsoft has in mind for the platform:
#MicrosoftDataflex it is – that is key because it applies to all of the Microsoft Cloud (that’s why no Power Dataflex :)).
The entry level offering of Dataflex for all Teams users is an important milestone for XRM/CDS/MDF (not sure I want to adopt “MDF” as the acronym just yet, actually). Don’t be surprised if the same technology will pop up in more & more places within the MS Cloud in the coming months and years.
Licensing details: TBC = “To Be Confusing”
The specifics of what is and isn’t included with Dataflex didn’t get released yet, as this new service is still pending for the public preview to start sometime in August. At the announcement day it has clearly been more important for MS to higlighlight what you can do with the non-Pro Dataflex vs. what you can’t. Even after the eventual GA, we can expect to see Microsoft use just the term “Dataflex” when referring to Pro capabilities like API access and custom code. When raising this issue of one additional source of Power Platform licensing complexity, Charles Lamanna confirmed that this is indeed intentional:
PowerApps with no spaces is just bad… I am hanging my head right now. Definitely unintentional 🙂
On Dataflex – Dataflex Pro is a licensing option (like Power BI Pro) and not actually a different thing.
This comparison to Power BI actually makes a lot of sense. In fact, much of what has happened with the platform side of MS Business Applications has been taken from the Power BI playbook used in conquering the Nr. 1 spot in business intelligence products within a relatively short period of time. So, here’s how the product levels are positioned over there:
Power BI: free for anyone to use, also for publish to web
Power BI Pro: user specific license required for sharing reports across the organization (non-public)
Power BI Premium: capacity based offering that allows enterprise customers to remove the per-user element from the licensing equation
What Dataflex and Dataflex Pro are going to offer is quite similar to this structure. No, I don’t quite expect the completely free edition of Teams to offer Dataflex features anytime soon, but the bundling into Office 365 & Microsoft 365 effectively makes it free for a huge pool of organizations to start using. When you then need to build some apps that go beyond the scenarios of team specific use cases and want to enforce some organization wide policies on it, Dataflex Pro may quickly become a requirement.
What may initially sound like not a very profitable business model (give stuff away for free) can result in a lot of positive effects on the company wide revenue streams of Microsoft if played right. After all, we’ve already seen the plumbing from under the Dynamics products to being reused as a platform for DIY apps (Power Apps), so this strategy of exposing as many users to Dataflex as possible is a clear continuation of that path.
It's the same question as "why buy Dynamics 365 Enterprise apps when I can build my own on Power Apps". Ultimately it's the scale of Teams that will bring in revenue, even if MS is (again) cannibalizing their previous offering to some extent with #MicrosoftDataflex non-Pro.
— Jukka Niiranen mstdn.social/@jukkan (@jukkan) July 23, 2020
Following the Power BI model, the next logical step after Dataflex and Dataflex Pro would now be to offer a Dataflex Premium service for enterprise customers who are more willing to pay for the capacity rather than for each and every seat. Whether MS ever launch this or not, the question of capacity consumption in terms of storage and API calls is bound to become a big source of confusion with the “free” vs. paid plans for Power Apps. Luckily the team behind Power Platform Center of Excellence Starter Kit is already working on making some of these governance tools compatible with the non-premium resources:
We're still working on the final details, but the plan is for at least the Admin (Core) components to work in Teams with Dataflex & no need for a premium license. With the idea being you have a Teams for your Power Platform admins where they can chat, share files and use the kit
The debate on why you should not use SharePoint lists as your app’s data source vs. why they area a perectly valid option has been going on for probably longer than Power Apps has existed. The inclusion of Dataflex into the core services of O365/M365 subscriptions has now prompted people to claim that app development inside SharePoint should be avoided altogether.
There are plenty of greant points in the blog post by Andrew Welch on the impact that Dataflex will have on the app strategy for organizations. Certainly the old reality of avoiding CDS due to licensing costs will need to be replaced with a more modern view into the low-code application development landscape in Microsoft 365. SharePoint will remain as the service most well known in the Modern Workplace technology category for sure, but questioning whether it’s the right tool for managing structured data should now become a mandatory step in all app plannign discussions.
One thing that’s clearly stealing the thunder from the Dataflex announcement and making the M365 folks even more confused is the GA announcement of Microsoft Lists that took place at the very same time. Yeah, this is just classic MSFT – having multiple products with overlapping feature sets and competing marketing messages.
Teams: your business app platform?
No matter how cool the apps that we build for customers (and ourselves) may be, it’s important to keep in mind that the majority of the average information worker’s time is spent inside applications that are exactly the same for everyone – like Microsoft Teams. This recent news about Slack first claiming Teams is not even a competitor to them, then moving to filing a lawsuit against Microsoft for bundling Teams with O365/M365 subscriptions, tells a lot about the game being played:
After claiming Microsoft wasn't a competitor, Slack is now filing a competition complaint against Microsoft in the EU, claiming it is illegally tying Teams to Office 365: https://t.co/tRdoqCj6yz
It’s a fight for the future of work, which will largerly be remote work. The digital transformation of business processes will require many, many apps that are tailored for a specific task inside a particular organization. Nevertheless, it’s the umbrella above them, meaning the platform for teamwork, that will drive a lot of the technology choices and licensing dollars into the online meeting and remote work ecosystem that provides the different business apps the common context.
If there would indeed be a significant uptake on the modern task based “mini” apps that do one thing well and we’d see organizations abandond their earlier enterprise system monoliths, then the question is where would all these apps logically land in? I haven’t really seen the enterprise app store concept being heavily utilized out there in the real world yet. Nor has there been much hope for Windows to enjoy a similar success with users installing apps from Microsoft Store, like they do all the time with iOS and Android.
It’s been already a decade since the app store concept was first launched for Microsoft Business Applications. AppSource is a great channel for distributing Dynamics 365 and Power Apps solutions in theory, but in practice it’s a lousy marketplace for most ISVs to do business in:
The #MSBizApps store that's been in the making since 2010 and still hasn't been able to deliver results: Microsoft AppSource. If even @stevemordue can't get customers via this channel, it's hard to imagine any other #MSpartner can make a buck from it either. https://t.co/13QzpELulM
— Jukka Niiranen mstdn.social/@jukkan (@jukkan) June 13, 2020
I’m not yet going to claim “this time it’s different” on the Teams app store and the Dataflex powered apps that can soon be published there by both 3rd party vendors and internal citizen developers. The idea however does sound like something that might scale better. Instead of deploying add-ons or templates into the single system of record like CRM, the new Teams based deployment model could allow a smaller group of users to install apps into an environment with a lot more narrow impact on organization-wide processes and data models. This could well be a more logical approach for Power Apps based solutions to find their way into more and more tenants.
It’s that time of the year again when big announcements have been lined up for Microsoft Inspire, the (virtual) conference for the MSFT partner ecosystem. Last year the Inspire announcements gave us a wealth of changes for the way how the Power Apps, Power Automate and Dynamics 365 licensing model worked, but all the product names remained the same (as far as I recall, at least). This year from Inspire 2020 the Business Applications part of Microsoft’s cloud is getting a bit of both. Say hello to Microsoft Dataflex!
I have written an introductory post over on Forward Forever blog, “What is Microsoft Dataflex and who is it aimed at”, which you can check out to understand the impact from a customer organization perspective. Here on my personal blog I’m exploring the topic when viewed through the eyes of a long-time XRM professional.
Abort mission!
Microsoft will NOT use the name “Dataflex” to refer to the services described here, due to a blunder they’ve made with trademarks. Read this post for more details: There never was a Microsoft Dataflex.
Dataflex Pro is the new CDS
Let’s face it: “Common Data Service” wasn’t exactly the finest product name invented at Redmond. Originally it was introduced as Common Data Model, then later this name was switched to reference only the open source schema of CDM whereas the application platform and data management features became a service called CDS. In March 2018 the v1.0 of CDS was effectively put to rest as the far more established platform behind the Dynamics CRM / Dynamics 365 Customer Engagement apps, known as XRM, was rebranded to be “the new” Common Data Service (without the “new” or “v2.0” labels, of course, because why make things precise complicated). Actually it was called “Common Data Service for Apps” (CDS-T), since there was originally supposed to be a “Common Data Service for Analytics” (CDS-A) alongside the transactional platform. There never was, as CDS-A was rebranded as “Power BI dataflows” before GA, so we only had one Common Data Service left to talk about. Whew!
After these adventures in the great platform shuffle of 2018, we did have some product name related announcement in 2019 as well, like the deprecation of Customer Engagement in the cloud, rebranding Microsoft Flow into Power Automate (where you still create Flows), and finally the “space program” for making PowerApps be spelled as Power Apps instead. In 2020, it had been eerily quiet in the product brand front, so I guess everyone was already expecting to see something once FY21 kicked off. As of today, July 21st, CDS is no more and the Dataflex chapter has begun.
“Microsoft Dataflex” is not perhaps the most exciting name ever, but it’s certainly a better choice than Common Data Service. In part 4 of my series on Power Platform licensing complexity, I wrote that CDS should have rather been called “Power Platform”, due to its central role as the glue that ties the various MS low-code tools into a coherent platform. It’s a very unCommon service and it does a lot more than just host application data. Now, Dataflex on the other hand plays with the words “flexible” and “-plex”. The latter, when used as a noun combining form is described in Wiktionary as having the following origin:
From the Latin past tense of plectere (“to weave, braid, twine, entwine”).
Yes! That’s exactly what this technology does! It’s a platform that allows you to weave a set of processes, interfaces and data together, into an intertwined whole that is greater than the sum of its parts. It offers every developer flexible, no-code/low-code tools for achieving this. That’s how at least I intend to explain the meaning of Dataflex from now on, unless MS product marketing comes up with something even more compelling.
Oh, there is of course one tiny detail to keep in mind: it’s actually Dataflex PRO. What Microsoft announced today at Inspire is a new service called Dataflex that is a subset of what CDS always used to offer. So, everyone who’s using CDS today is already on the Dataflex Pro version, whereas we’re about to get a preview of the non-Pro Dataflex in the coming weeks.
Application platform for the masses: Dataflex (with Teams)
If you’re coming from a Dynamics 365 background, the introduction of a “lite” version alongside the CDS/XRM platform capabilities might not sound so exciting at first. After all, from a pure functionality prespective Microsoft is taking away capabilities from the existing platform, to establish the baseline offering of Dataflex that customers can then upgrade to Dataflex Pro when needed. Some might dismiss it as just another licensing hurdle to worry about when trying to develop apps.
From a commercial perspective, Dataflex is possibly the biggest thing that has ever happened to the platform formerly know as CDS/XRM. Biggest in terms of the number of potential app makers that will now have the possibility to build Power Apps on top of a true relational database instead of the dreaded SharePoint Lists. This is obviously the major immediate benefit from the announcement of Dataflex non-Pro, since in the previous licensing model the lines were drawn in a way that discouraged the use of CDS (or Azure SQL) databases for any Power Apps designed for light use scenarios – unless you had already acquired the full Power Platform license for your entire organization. With Dataflex usage rights bundled into Microsoft Teams at no extra cost, any user (internal or guest) with the necessary Office 365 / Microsoft 365 base license can now access Power Apps that leverage the true powers of a modern low-code application platform.
You won’t get everything that CDS used to offer with just a Teams license now, of course. For starters, the only place from where you can access these app built on the non-Pro version of Dataflex is from within Microsoft Teams. None of the existing “players” for Power Apps will be supported, neither is embedding the apps to other UIs like SharePoint pages. Data types, security model, business logic, integration points, ALM… All of these will be far more limited on the “free” Dataflex than with the “real” application platform that is Dataflex Pro. For small apps aimed to be used by small-ish teams, though, I bet the Dataflex version will be good enough for a large percentage of use cases. It certainly is more advanced than SharePoint Lists, although the restrictions on app access points is more limited with the Teams based approach.
Limitations can sometimes be a good thing. Dataflex is a simplified version of what CDS was, not just in terms of features but also the maker & admin experience. Do not underestimate how important this aspect can be in getting the citizen developers to make the right choice. Experienced CDS/XRM professionals may have been reluctant to move over to the new Maker Portal that is still missing features from the classic Solution Explorer. The new generation of app makers who have never even heard of these acronyms will most likely be happier within the Teams embedded app creation experience and give a higher NPS score to Power Apps as a result. Less is more, as they say.
There’s going to be one Dataflex environment per Team that you can create, and the users will then map to the members of that Team, with a simplified user security role structure of Owner-Member-Guest (a.k.a. OMG). No business units or custom security role privileges to worry about, nor any complex data types like activities, polymorphic lookups, currencies etc. Yes, I know, most of the apps you’ve built already probably depend on those, but remember that this is not aimed at you! Dataflex Pro will have all of it and the non-Pro environments can be promoted to full capabilites, at which point they will require the same licenses as CDS does today.
This is still a Big Deal. With the “free” Dataflex included in Teams, Model-driven apps can soon be created by an equally large crowd as Canvas apps before. Eventually it might no longer be an awkward moment when you need to talk about “that Power Apps app type that looks like Dynamics, not like all the other apps you have”, since every app maker can generate those beautiful, structured UIs on top of the database tables they’ve added into Dataflex. Since Model-driven apps are so metadata driven, I’d imagine this route would actually be a lot easier for many power users of Excel that haven’t necessarily been comfortable with building pixel-pefect mobile UIs in Canvas apps before.
Despite of everyone having the possibility to store app data inside Dataflex, it’s not by any means imperative that a single database is established as the place for all business data. That is the traditional CRM approach of doing things, but with the modern tools like Power Apps and Power Automate the source and target of data operations could be a number of different systems if needed, thanks to Connectors. However, I’ve understood that Premium Connectors will still not be included with the Teams license, so the options for data connectivity will be more limited here. This may again be perfectly fine for the simple apps that the non-Pro Dataflex is aimed to serve.
One thing the free access to Dataflex should make ubiquitous is the usage of solutions for app packaging and ALM. As the solution framework is a concept from the XRM era, replacing the earlier standalone .zip packaging of apps and Flows required that everyone would get access to this layer in the underlying platform. From an administration perspective it should be a clear step forward to standardize all the work to be done with Dataflex tools. There is also the promise of further deployment options being introduced with the Teams app store and its “single click” app installation, so this is an area for everyone to keep an eye on.
The business impact of Dataflex
The launch of non-Pro Dataflex can be expected to generate a wealth of questions from those professionals who’ve been working with CDS based solutions. What is & isn’t supported in Dataflex, where do the license and capacity limits actually get drawn, how are they enforced, where can apps be created for different Dataflex environment types, what admin controls are available on the tenant level, and so on. It’ll be confusing initially, but in the end I bet it’s gonna be woth it for all parties in this ecosystem.
One year ago I wrote a blog post examining the possible growth directions for Power Platform. This launch of Dataflex + Teams now broadens the scope of the platform both on the Citizen Developer dimension (number of potential app makers) but also in the “Other MS products” axis. Earlier I’ve enjoyed making the joke that one day Microsoft would build the next version of SharePoint on top of CDS. Well, that hasn’t quite happened yet (and most likely never will), but we did see a major step now in Power Platform catching up with the mainstream productivity apps in Office. After all, every time you create a new Teams team, a new SharePoint site is provisioned for it. We’re now at a point where technically each team could also have a Dataflex environment provisioned at the same time. Knowing this, what’s stopping future Teams features from being built on top of CDS Dataflex?
Because every Team will have Power Apps with a robust platform behind it, this also means you'll be able to package up an entire solution — apps, flows, bots, tables, etc — and publish it to the Teams app store. Users can find and install in a single click.
As we can see from Ryan’s example, there are interesting synergies between Power Apps and Teams that are made possible by having Dataflex “on by default”. From a partner ecosystem perspective, over 500.000 organizations around the world who use Microsoft Teams just became a potential target for no-code solutions that could be dropped inside various teams. I’m saying no-code here instead of low-code because presumably the non-Pro Dataflex environment wouldn’t support custom code. Anyway, there’s a lot to investigate here with the Dataflex public preview coming in August and GA planned for September timeframe, so I will definitely be returning to cover the topic more closely once detailed feature lists are publicly available.
