Fabric Awards – Ep. 492
Mike and Tommy hand out their completely made-up Fabric Awards for 2025 — from “Most Likely to Be Disabled by IT” to “This Is Why We Can’t Have Nice Things.” Plus, Marco Russo confirms what they’ve been saying all along: notebooks win.
News & Announcements
Marco Russo on Notebooks and Fabric Performance
Mike highlights a January 2nd post from Marco Russo (SQLBI) that validates what Mike and Tommy have been experiencing firsthand. After running controlled tests, Marco concluded:
- Notebooks are the cheapest and fastest way to ingest data in Fabric — beating pipelines, Dataflows Gen2, and copy tasks on both CU cost and execution time
- Both PySpark and T-SQL notebooks work well; Marco prefers SQL notebooks given his background
- LLMs make you significantly more productive with PySpark — great for bridging knowledge gaps between SQL and Python
- Converting CSV to Parquet (not Delta tables, just the raw Parquet file) yields a substantial performance increase for downstream operations
Mike’s pro tip: use the free Copilot in Edge browser instead of the Fabric-embedded Copilot — zero CU cost, faster startup, and no “unable to process” failures. At $10/month for Microsoft 365 Copilot, unlimited queries beats burning 100-400 CUs per notebook Copilot interaction.
Tommy adds his approach: browser sidebars loaded with ChatGPT, Claude, Gemini, and Grok projects — each pre-loaded with Fabric documentation and context (his “Fabric Guru” project) so he doesn’t have to re-explain the stack every time.
Main Discussion: The Fabric Awards 🏆
Mike and Tommy invented their own awards ceremony — “Where the points mean nothing and the awards are fictitious” (Whose Line Is It Anyway style). Here are the 2025 Fabric Awards:
🏆 Most Likely to Be Disabled by IT
Mike’s pick: User-Defined Functions (UDFs) Functions can talk to external APIs, do arbitrary data transformations, and are required for translytical task flows. But most BI developers will never touch them — they’re a pro/developer feature. IT teams will see the security risk of arbitrary code execution and reach for the disable button.
Tommy’s pick: Fabric MCP The MCP is “made to be turned off by IT.” It’s a plugin that interacts with enterprise data, authentication feels like a black box, and most MCPs don’t have standardized security models. For large organizations, Tommy doesn’t expect Fabric MCP to be deployed by IT anytime soon. Mike pushes back — the term “MCP” covers a lot of ground (VS Code, Power BI Desktop, Fabric environment), and the local-machine MCPs are a different conversation from the enterprise Fabric MCP.
🏆 Best Feature Nobody Reads the Documentation For
Mike’s pick: Licensing Eight years running and still nobody reads the licensing docs. Which SKU turns on what? Can I use Copilot at this level? What’s the CU cost model? Everyone has opinions; nobody reads the actual documentation.
Mike’s second pick: Data Agents The documentation is paper-thin compared to what Microsoft used to ship. The old Q&A feature had four comprehensive articles on data preparation for natural language. Data agents? “Set up an agent, write a good prompt, connect to your model, deploy.” That’s it. No best practices, no prompt engineering guides, no deployment patterns.
Tommy’s pick: Mirroring and Shortcuts So easy to set up that nobody bothers reading the docs — just click the button and it’s connected. But when a shortcut needs updating and you open the configuration panel for the first time, you realize you have no idea what any of those properties do.
🏆 This Is Going to Matter Way More in 2026
Mike’s picks: Fabric Workloads + OneLake Catalog Workloads are gaining momentum with Microsoft running competitions for custom workload ideas. OneLake Catalog is an underrated governance and discoverability tool — view all tables, semantic models, lineage, add descriptions, use tagging and domains. Missing pieces: descriptions on lakehouse table and column names.
Tommy’s pick: Data Agents (with Copilot Studio) Not demos — real deployments. The vision: a Fabric data agent connected through Copilot Studio, accessible in Teams, Outlook, on your phone. Not just “show me a bar chart” but triggering actions based on data. If that’s all data agents do — answer questions about dashboards — they’ve missed the mark.
Honorable mention: Fabric IQ Both agree it has promise, but it “didn’t play enough minutes” in 2025 to be eligible. Announced at Ignite, available shortly after, but the documentation is sparse and building ontologies requires too much manual clicking. They’re skeptical but intrigued — will check back mid-2026.
🏆 This Is Why We Can’t Have Nice Things
Mike’s pick: Workspace Tabs The colored tab system for switching between workspaces inside Fabric. Mike can’t stand it — the PowerBI experience puts workspaces at the bottom, Fabric puts them at the top, tabs are too small to read, and you can’t Ctrl+click a workspace to open a new browser tab. “Just let me Ctrl+click and open a new browser tab. That solves all my problems.”
Mike’s second gripe: the workspace pane width button. Instead of a draggable resize handle (like every other pane in Power BI Desktop), you get a button that toggles between two fixed widths. A miss.
Tommy’s pick: Git in Fabric The #1 question Tommy gets cold-contacted about on LinkedIn. Teams want to use Git with Fabric but don’t have the DevOps background — BI developers have historically only worried about backups, not deployment pipelines. It’s borrowed from the IT/software world, and the learning curve is steep for business intelligence teams. Mike agrees: at enterprise scale, you can’t rely on someone manually pushing reports from dev to test to prod on a Thursday afternoon.
🏆 Best Feature Requiring a Long Awkward Conversation
Tommy’s pick: Fabric Capacity Billing The model keeps changing, database costs are separate from Copilot costs, and nobody’s quite sure what they’re paying for.
Mike’s pick: Task Flow Diagrams Every demo, the task flow pane is open. Every demo, Mike closes it. When clients ask what it does: “It helps you group things and gives you a broad diagram of how things link together.” Eyes glaze over. Move on. The feature shipped buggy (renaming would accept after three keystrokes), it’s rigid, and it’s not getting new features. It’ll just… be there.
Bonus awkward conversations: Metric Sets (deprecated — Tommy lost a gentleman’s bet with Carly that metric sets would surpass report usage in five years) and Dashboards (the single-pane-of-glass feature nobody has used since 2018).
Episode Transcript
Full verbatim transcript — click any timestamp to jump to that moment:
0:13 [music] Good morning and welcome back to the Explicit Measures podcast with Tommy and Mike. Good morning, everyone. And Tommy, how are you doing? Good morning, Mike. , back into it, man. Back into the grind. Here we are. Our second episode of 2026. We’re doing a couple of fun episodes here. , this episode’s another one of these fun episodes. It’s called We’re going to call this one the
0:46 Fabric awards is what we’re going to call this episode. We’re just going to give out some silly awards for things that happened around fabric, things that we liked, things that we are going to tease about and pick on a little bit here. , but just some funny awards. This is what is it Tommy? There was a show that’s where the points mean nothing and the awards are fictitious. Whose line it is anyway? Yeah, that’s exactly right. Yeah, the game doesn’t matter and the points mean nothing. Yeah. So, this is literally Mike the fabric madeup awards. None of these awards exist. They’re not real. You cannot find them anywhere. We literally made them up as
1:19 An excuse to talk about more things about fabric. But [laughter] but really like if we were to hand out awards to 2025 and what Microsoft did, this is the awards that we would want. Yes, agreed. All right. All right. So, let’s jump into some awards here. I will move things around here a little bit on our note our noteboard. Before we get into our main topic around giving out some awards, I do want to [clears throat] I I found an article. I found a post and Tommy I get up in the morning I read through some things and I I see a lot of
1:52 People from the Microsoft community and there’s one individual that Tommy and I really respect. He’s a legend in the space and this is Marco Russo. So for the chat here, everyone knows who SQLBI is and Marco Russos are. I’m going to I’m going to say I feel like I’ve been vindicated here a little bit. , Marco is a staple in the fabric PowerBI space. He’s been heavily semantic model. That’s that is the primary place that he’s been spending his time and his his knowledge.
2:26 That’s where he came from. That’s how he started all that. He did MDX and then went into SQL now into the semantic models. There’s a there’s a whole bunch of stuff that Marco’s team knows how to do. Very wise individual. Well, Tommy, here’s a post that I want to review with you. Marco just recently put out here on Twitter recently and the and the the Twitter post here is actually in the chat as well in case you want to go look at it as well. Marco talks about it’s a this is my first post of the year and then it’s about fabric is what Marco says. This is January 2nd is when he pushed this out. He and let me just read a couple
2:59 Excerpts from this to you Tommy because I feel like at some level I feel a bit vindicated through his comments here. What I thought my what I thought was a gut feeling has been confirmed by validating multiple tests. I love this when people write tests about things and do and do stuff. Especially the things that Marco Russo team does. Like they’re very good about like doing specific tests cleanly and and clearly. I would love to see the results of his test in this as well. Clearly what follows is valid for me and may not be valid for other scenarios. He’s some caveats here, but I’m pretty
3:32 Confident with my conclusion. Okay, summarizing a couple things here. Using notebooks is the cheapest and fastest way to ingest data. Pipelines, dataf flows gen 2, and copied tasks are more expensive in cus and typically a slower execution time. I’d even go maybe so far as to say I would even prioritize those items. I would say notebooks by far has been the fastest thing that I’ve been using. Then there is the next fastest thing would probably be pipelines. I think it’s the next fastest and I
4:08 Don’t know about speed or cost but I think costwise dataf flows gen 2 is the next most efficient cus and fastest. Yeah. And copy copy jobs or copy tasks I’ve had mixed results with. It seems like it’s worked. It’s been somewhat helpful. It makes it easy, but it still seems like one of the more expensive items. seems on par with with the dataf flows gen two experience and that’s what I was experiencing and I it’s having other people confirm what I already have heard or been testing myself feels nice you
4:42 Know it’s funny Mike because this is not just a fabric thing like correct me if I’m wrong but notebooks are have been the de facto and the preferred choice for data engineers for years let me just give some context around this one. So, so I I’ve been doing a lot of data engineering. I stepped into the world of data engineering when Spark and Lakeous and data bricks really were coming onto this onto the place the playground basically. There’s a new kit on the playground and it was data bricks,
5:15 Right? That was the space in which I was really interested in. Really enjoyed it. I thought it was super fast. I really like this idea of like separating the storage from the compute. Microsoft had a number of visions of how this would work. , and arguably like Synapse was doing some of that. They had a Spark engine like so Microsoft was doing some things. There’s massive parallel processing with their, their SQL data warehouse. It was taking a lot of similar notes from what Spark was doing, but it was always too expensive. To me, the
5:49 Efficient piece of this was Spark and Lakehouse storage. So, I guess I would say Tommy, maybe [clears throat] I came in at a weird time, but it feels to me like the technology shifted to Lake Houses and Spark became highly efficient and very commoditized for regular people to use it and make it easier for us to to leverage these type of technologies. It just feels like a weird Yeah. No, but but again these aren’t nec This is the crazy thing like after really two years now of all these
6:22 Innovations with fabric. This is not anything new, right? This is something that this has been the preferred way. And like he’s not Margot’s not saying anything he shouldn’t be saying anything groundbreaking in terms of like whoa I can’t believe that’s finally the where we’ve come to the conclusion. We’ve known this pre fabric, right? Yes, I would agree. Yeah, and I feel like you’re catching up a little bit. Maybe I think that’s where just the validation is like yes, we can try all these things and copy jobs and get data flows better and have a real time, but at the end of
6:56 The day, right now, we have something that has not been beat and that is notebooks. And I I’ve seen that very quickly, Mike, when I first got started where it was in a an immediate even at my just beginner stage and doing basic things on how fast and efficient it would be. did I enjoy it the most? I was getting used to it. But in terms of going from data flows and going from power query and going from that world no there this is nothing if you’ve already been in this
7:31 World nothing groundbreaking here however it does need to be said and Marco feels the need to say it too where it’s like yes we’re trying all these things in fabric yes but don’t go away from the thing that’s been tried and true I still think he’s not saying anything about like where do you land it on the semantic Again, I I just like another set of ears or another set of technical people really reviewing this and giving me another the thumbs up of saying like look, notebooks are the way to go. , and other observations I guess that Marco provides here. ,
8:05 He said both PIS spark and TSQL notebooks work well. Again, Marco prefers working in the SQL notebooks. Again, that makes sense because of his background. I’m okay in the Python notebooks. , one of the things he does point out here that I think is very useful is when you use large language models, they definitely can help you write Python and language and build things, but you you you can’t trust them blindly all the time. You do need to make sure what you’re doing, but they do give you a lot of assistance around what the code is doing, and you can you can test it. I think it’s fairly easy to test these things. , he says large language
8:38 Models 100% definitely makes you more productive, especially for pi spark. So things you don’t know, gaps in knowledge. You may understand SQL, but you don’t know how to do the same thing in PIS Spark or Python Spark. Use large language models. I would agree. This is a a sentiment I’ve seen really well. I use a lot of the time I use the free co-pilot in my browser more than I use Copilot that comes out of the box from fabric because if I use the copilot in my browser, it doesn’t cost me anything. I get zero CU usage if I use the one that comes out of my Edge browser as opposed to using the copilot that’s
9:10 Actually in fabric. So, that’s one another trick I’ll give you here as well. It doesn’t need to spin up, which is great. , I don’t have to add extra code. And isn’t that crazy, Mike? Yeah, I use the co-pilot in Edge and right away it’ll just start generating code. The co-pilot in notebooks takes for takes a little time and sometimes it just comes back with sorry, gibberish, unable to process that like that was awesome. the the initial versions of C-Pilot in the the notebooks very
9:47 Lackluster. I think it’s gotten a lot better since then. So, I think it definitely hasn’t been improved. It seems to give me better more answers now, but still, why would I spend 400 or 100 CUS asking it a question when I can use the Edge browser and get the same CU usage for free, right? And also my Edge browser, I pay for the C-pilot, Microsoft 365 Copilot. So I have the work and then personal version of Copilot in my Edge browser. I just use that. Like again, that’s $10 flat cost. I can use it as much as I want without
10:19 Actually charging me any CU usage right now. It feels like a better deal. And honestly, Mike though too, well, we’ve talked about this with with AI and these AI tools where it’s going to get to the point as soon as you use it, if it doesn’t do what you want, people will move on. What’s going to cause you to in a sense get away from that habit? Because I do the same thing where the the speed, the just efficiency of it. Yeah, I’m use I have all the Do you use Edge sidebars? So, like you can actually add like apps to your Edge side
10:53 Panel. That’s an interesting question, Tommy. I have not used it. I have noticed every so often if I click a link in Teams, Teams will open the web page in Edge and immediately open up the Teams sidebar for me. interesting. I like that a little bit sometimes. But other than like Teams and Copa, there’s not really many other sidebars that I specifically use inside Edge. I I don’t use it very much. Do you have other sidebars, Tommy, that that work for you? I Yeah. And you don’t have to install anything. It’s just literally a bookmarking site. So I’ll I have open chatbt claude Gemini Grock and
11:28 Obviously the the one the co-pilot is a little more integrated because obviously it’s a Microsoft product. Yeah. But a lot of times I’ll just be especially if I’m dealing with a pretty complex thing. I’ll open one sidebar go hey perplexity can you research this because I don’t know how that works in Spark. and then go to open AI. So I have that nice little assistant which whatever in a sense tool I want and Mike I’ll add on here too since it’s really just like almost the mobile view
12:00 I just go to my project because I have a project in all my AI tooling like for chatbt and that’s for the fabric guru and has fed been fed all this information about fabric and notebooks and what I do. Yeah. So I don’t have to provide too much context where it’s just like hey I we’re doing more pi spark yada yada it already understands the framework of utilizing pi spark and fabric and we are on our way. I like it. I like it. I’m I’m with you Tommy. I I maybe I need to take another
12:32 Look at the sidebar things. There’s I do enjoy having the the chat icon there. I think it it makes sense or having the co-pilot available. I probably should be adding a couple more agents and things there to make it a bit more effective. , and I I don’t know where it fits in this spectrum of things, Tommy, but having more of like a knowledge base, right? , I’ve heard of things like notebook LM for like research and doing like taking notes in like a more AI centric way would make sense. That way I could have like a a more of like a knowledge collection of things that I understand. , I need to do a bit more
13:05 Of like I have a lot of knowledge. It’s all stuck in my head. it just hasn’t been materialized into like an agent or something that I can collectively put a bunch of information around and ask questions to it. So, I’ll have to spend some more time figuring that out. You’re the one that’s really taught me the be the the approach to this because yeah, there are so many tools like, oh wow, I’m going to learn any history book I want. I can learn it in two hours thanks to this new GitHub repo. That’s well and great. But then the thing with is just utilizing those hey quick thing need this get that move on our way. So
13:40 Yeah so let’s jump into our main topic. I think that’s pretty good for our inter Oh sorry one last note Marco put at the very bottom of this thing. He says one other performance indicator and I did not know this. This is something that was new to me. Everything else he said so far was I agree with I I align with the last one. Oh yeah Aaron thank you for the the comment here. Obsidian is one of the ones that’s for like organizing knowledge. That’s Tommy, you use Obsidian, correct? I use Notion. That’s use Notion. Okay. Surprisingly, yeah. So, so that’s I use a lot of loop. It feels like these Obsidians and notion
14:14 Should could should take from like what Loop is doing. Loop should just migrate into like these other technologies. It should it should help you out more things like that. So, not sure how those are all working, but yeah, I’m I’m going to need to check out Obsidian. And I’ve heard a really good idea around keeping that for knowledge organization as well. Anyways, moving on. Sorry. The last note from micro says, “Oh, and one last note just for the future, not myself remembering these details, but it’s always better to convert a CSV file into a paret file one time.” And he’s talking about CSV to parquet
14:48 And not a delta table. Here he’s just saying create just the file. So taking the CSV file, transforming into the parquet file and then reading the parquet files. He said he was consistently getting a better performance and it wasn’t it wasn’t like a small performance increase. It was a substantial performance increase as he was doing things inside that table. So maybe just one side note here. If you’re using data flows to write data down to a lakehouse, if you’re using other experiences, you can write them into
15:20 Delta tables if you want. , but you also may just want to consider taking data and just transforming it from CSV to parquet. Again, just a parquet file, which is a columner stored file. Just use that alone. And that alone will help you build out better, more efficient pipelines as well. Again, I also feel like this is we’ve done this. I’ve I’ve heard this before like this is already there something that’s been proven in the community and what data engineers already do, right?
15:54 Yeah. Yes. Okay. Enough of my rambling about around about around how much I love notebooks and how much you should start using them. So, if you don’t if anything you take away from this, if you’re not using notebooks today, you should definitely spend some time Marco recommends them. I recommend them. Tommy recommends them. You should definitely spend some time getting more familiar with notebooks in 2026 if you haven’t already. That’s that’s your that’s your easy win here for 2026. All right, that being said, let’s go into some of our awards. I don’t know how we do the order of these awards,
16:25 Tommy. I’m going to let you just pick whatever you think is good to start off with here. but maybe state your reward, your award, and then I’ll state what part of fabric applies to that award for from what I observe. Oh, thank you very much. And I I’ll give it a little brief more context on what we’re doing here. So again, yeah, go ahead. These are not real words. You’re not going to find this on Gartner’s magic quadrant or someone’s or maybe we should. Maybe we should. I Microsoft may not like all these explicit measure magic mushrooms something. I don’t know.
16:59 [laughter] Someone’s been playing Super Mario. exactly right. So, , honestly, what we are just doing is having a little fun. We’re if we ever wanted to give out our own rewards, both on the things that are frustrated and cool, these are not MVP rewards. They’re just more fun. And you’re going to see by the names of the rewards that we literally made all of these up. Honestly, it’s about, but Mike, it goes back to our day-to-day when we’re in the grind, when we get frustrated, things like, man, if there was a word for this,
17:31 This would win. and just give give a shout out to what 2025 was. So, let’s go with ah [sighs] we have a lot to choose from. let’s go with a fun one off the bat. This is going to be called the most likely to be disabled by it award. [laughter] This is a good one. I thought this award was a really funny title here. So, yeah, thank you very much. And again, the idea here is Microsoft’s always coming out with new features as quickly as possible. But if you ever worked in an I with an IT team, they are
18:06 Very adverse and there are some things more than others that are going to make them go screen to the disabled button. So Mike, I got to hear from you. when you looked at what came out in 2025, what was or some of the things that immediately got you out of or into the IT mood? Yeah. And let me let me slightly preface this here right now. My understanding is there are some ways for you to turn off specific workloads
18:39 Inside Fabric, right? If something’s in preview, you can you can turn it off and not let users use it. But once that workload gets into G, I don’t think there’s really any way to turn stuff off at this point. I I’m not really sure. So, , so I’ll just preface this with this is like a fictitious feature anyways because it technically can’t turn everything off yet. I don’t maybe that’ll get there eventually at some point, but , one of the things I think that it would most likely turn off is the fabric userdefined functions. So I I think that’s a feature that it
19:14 Sees a lot of risk around just because anyone could write any function. Functions can talk to external APIs. It can do any data transformations. And also I think if we’re talking about like a business intelligence team, I don’t think our first go-to is hey let me build a userdefined function and use it in my workflow. I I think that’s more of an advanced feature. I think prousers or professionals, developer professionals would be the ones to explore, learn and use that feature. So I don’t think it’s a very commonly used
19:45 Feature yet. I have done a ton of demos around task sorry it’s it’s called the oh no I’ve done done a bunch of demos. I’m drawing a blank on the name here. It’s rightback basically or rightback or using oh the trans analytical translitical task flows like so that’s that’s the translitical task you need a userdefined function a fabric userdefined functions to make those work so yes it’s part of the system yes it does some really powerful things but I think there’s a lot of we’re not it doesn’t
20:18 Understand like where that security lives and how we secure them correctly and therefore I think people will turn them off because people are are worried about the userdefined function modifying tables or data or or messing things up. So I think that’s one of the features that I would say it will likely turn off. What do you think Tommy? Interesting. I can’t believe that you didn’t think of the most obvious one. And Mike, the most obvious one to me is fabric MCP. Oh, that’s another one. Yeah, I didn’t think about that one either. That’s another that’s another one, too. That is that is literally if this was a
20:52 Sports team, this was like he’s got promise. like it is. He was made to be turned off by and disabled by it. Wait, so it’s a plug. You’re a plugin. This is a plugin by Microsoft. Where’s the authentication? Well, it’s built in. MCPs, not that they’re not secure, but here’s the thing. Nobody really knows because it feels like a black box. It’s very much a black box and everyone just wants to use them. So, and especially that most of them don’t have to do with a authentication. So MCPs because they can interact with so many
21:25 Things while amazing as they are if you’re in it that is being vetted big time because yeah I actually should thought thought of that too in terms of no one really knows the security risk of MCPs because they’ve been out for so for so short and but what they allow someone to do is really cool. I don’t know. But yeah, MCPs are definitely the one where it’s I I think that as fun as they are, if you’re in a larger organization,
21:59 And maybe you might disagree with me on this, if you’re in a large organization, you shouldn’t expect to be using MCPS and fabric anytime soon. Yeah, I don’t know. that that’s one. I’m not sure if I agree with that one. There there is. So, you say the word MCP. There’s a lot of MCPs. I think there’s like an MCP for like data warehouses. I think there’s an MCP for VS Code. There’s an MCP for PowerBI desktop making model manipulating the report. There’s a lot
22:32 Of MCPs around it. It so the MCP space is large I think. so that that’s one I’m I’m which MCP are you specifically talking because I know there’s a fabric a generalized fabric MCP as well, right? I guess that would again I haven’t used that one time so maybe fill me in a bit here. Is that one where you can like create workspaces create report like what what does it do? What does the fabric MCP do? So it just simply allows you to touch there’s the fabric MCP and there’s a PowerBI semantic model MCP that are the two big ones. The fabric MCP simply
23:05 Allows you to actually monitor and through conversational piece interact with your fabric environment, see what items are out there, create etc. And then the semantic model one is simply that interacting with your model. You can update things and get information from it. Okay, so we’re on the same page here. So I I guess I would say I’m probably more on the side of the fabric MCP the IT team might turn off or only turn on for a select few. I I I would argue that MCP if you’re going to turn that on in the admin settings, that
23:38 Would be something you would definitely want to govern with a security group, create the security group of fabric MCP users and then attach people to that and then put that user attached to that specific feature. yeah, I could see that. I could I could see that one. However, I don’t know if it is going to be able to turn off MCPs inside VS Code and doing things like local changes on your semantic models. I think that’s really different though because that’s again the individual when you’re thinking the fabric MCP that’s talking to your enterprise data.
24:11 I’m just saying you said the word MCP. I’m just thinking about like that term to me comes with a lot of different pieces, right? So you say MCP that could mean a lot of different things to different people like what are you interacting with right? Does that does that make sense? So I’m just I’m not debating your I’m not debating your comment. I’m just saying that term seems loaded to me. Okay. All right. And that that’s fair. I let me put it this way. I don’t think anytime soon we’re going to have that integration with like other MCPs where hey I’m talking to Slack and I’m actually interacting with my fabric environment. I think it’s still the
24:44 Individual’s not call but if you have someone who’s skilled enough that’s working on their own local machine. Okay. Yeah. They they would be able to interact with it but it’s not going to be anything that’s going to be deployed by it anytime soon I don’t think. Oh heck no. That that’s I would and this is I think also these two features Tommy particularly the UDFs and then the fabric MCPS. These these are all like premium pro, very skilled individuals you’d give these accesses to. Like these are these are not common like launch it across the company and just have everyone go to
25:17 Town with it. A thousand%. So, next award. Let’s go to our next one. All right. So, why don’t we [sighs] I’m going to save that one because I really like that one. Okay. Well, let’s say let’s do the it’s the best feature that no one has ever read the documentation for award. So, this is one that it comes out and we’re going to start using it. We’ll figure it out. We think we know enough information about it. I’ve never looked at the Microsoft docs
25:50 On this at all. let’s be real, Tommy. This is everything that comes out from Microsoft. I never read the documentation. I just start clicking buttons and see what I can figure out. Only when I get stuck do I go find the documentation. So I’m not going to answer everything in fabric. So [laughter] I I won’t answer that one yet. Good choice. Good choice. I will give very specific things. The feature that nobody reads the documentation for is licensing. Licensing? No, [laughter] not until you have to pay. I think everyone’s confused about
26:23 Licensing and the licensing seems to be changing all the time and there’s different so I feel like everyone has an opinion on licensing but no one actually sits down and like understands what’s going on with licensing and the very specific this is why part of the reason why we have jobs Tommy there licensing is complex how do you optimize it for your business are you buying a bunch of fabric SKs you giving it away for you are you buying a F64 level what what turns on at what level there’s a whole bunch Wait, I thought I could still use co-pilot with this. What gives? [laughter]
26:55 So, , that’s my funny one. So, so licensing is the thing I think no one reads the documentation on. The other one I’ll give is probably because you could have said that eight years ago and it still would have been right. It still would have been true. Yes, that that award has been for, for the for the eighth year in a row or 10th year in a row. Still no one’s reading licensing. That’s good. , I will also say data agents is another one of those that I think no one reads the documentation on and I’ve had to go like play with it a bit and then like okay what are we really trying to do here? What’s going on with this? So that’s another one I would argue is another space
27:28 That no one’s reading the documentation on. Those are my two features. What do you think? Because there’s not too much documentation on data agents for you to read. So, it’s interesting and I think these products are still so dang new where when you looked at when PowerBI came out with the new the before dataf flows gen 2, they were really updating Gen One and they had this a ton of articles on the like not just hey click this button do this and now you
28:01 Can start using your dataf flow. It was best practices guides. , it was how do you work how do they work together? It was different methods and things to look out for. And there it would take you a long time to read through all all through it all. Right now data agents really just has, hey, set up a data agent, write a good prompt. Hey, connect to your semantic model. Just so if it’s semantic model, you’re doing DAX. If it’s a database, you’re doing SQL. Step two, deploy. and you’re good. So
28:34 The documentation is very wanting to me where you see what I’m saying where we don’t have that comprehensive and again it is new but that comprehensive like not just okay I click on this I write a prompt but Mike for crying out loud when they had the Q&A that they’re deprecating or they’re removing the the built-in AI features in PowerBI they must have had four documents four articles on the learn on how to prep your data for natural language.
29:07 Yeah. it was like hey things to look out for in your columns. Hey this is the best way to ask questions and they were very comprehensive and that was something that nobody used. So we’re still at this point where the documentation is very slim. I think it’s also maybe too like the Microsoft I don’t know if they want to give a ton of guidance around what you can and cannot do with a data agent because maybe it’s relatively not relatively it’s slightly unknown. It’s like very open-ended like whatever instructions you provide back to the agent can give you very different
29:41 Answers. Like you can you could say be very direct, be very add no fluff and it will change how the agent behaves. And maybe that’s just something where as we’re exploring like what agents are and how do you talk with large language models that might be just something in general that we have to figure out and learn on our own. But I I feel like data agents have a wide tool set can can do a lot of different things. I don’t know how you describe it, but it feels like it’s a much more of a a free floating idea as opposed to something that’s a bit more locked down
30:14 To give you very specific things like a Q&A visual. So I I think that’s an advantage. It’s also maybe a disadvantage because then it’s harder to write documentation around it. And maybe we just need some more best practices and guidance guidance documentation to show up for that which again back to your point Tommy Microsoft’s not supplying that today currently but maybe they need to supply that over time. And I honestly like you hit the nail on the head too. it’s not so much on all the like the style of the prompt even though they should be able to at least have what each system instruction should have but also
30:46 How do you deploy it what is the best way to get people to start using it right and I think that’s the hardest thing with a lot of these is okay cool I can create a data agent and I got to write a good prompt which again they should probably have a few pages on prompt styling and semantics but then what like okay how do What are some examples of deploying an agent out to consumers? So things like that I think would be really helpful. I’d agree. All right. You want to pick our what was your So mine is what you would never think of
31:20 It is mirroring and shortcuts because it’s so Yeah. Yeah. I thought that was a good one. They’re so dang easy to set up. , they do have configurations, but once you heard about it, you’re like, “Oh, so I just click that button and it’s connected.” Cool. I don’t That’s a really good point. I have been caught a few times with where a shortcut had to get updated and then you go to the configurations of the shortcut. You’re like, I’ve never seen nor know what I know what any of these properties are. So shortcuts does work
31:56 Marvels, but and it’s meant to be easy. However, make sure what you still know what you’re doing. H interesting. Yeah, I can see that one. I there’s probably some edge cases or things that it can or cannot do that you need to be aware of. I do like a lot of the mirroring. I don’t know how many people actually experiment with it. One of the things I think that’s interesting with mirroring is when you talk start talking about mirroring to SQL, it’s it’s to some degree mostly free of what the mirroring is doing. I think that’s an underrated feature of mirroring that
32:30 You get a lot of free storage of what mirroring’s doing in order to get that feature turned on and used within the system. I think that’s interesting. It is a marble that we have mirroring and maybe because it’s just not the fanciest feature. We don’t hear too much about it. But the fact that we’re not paying for getting that data into fabric or maybe Microsoft’s paying an arm in a league just so data is in fabric still it is something that should definitely definitely be utilized. I’m not sure. So I would maybe argue it a slightly different way Tommy. I think
33:02 One of the reasons why Microsoft is doing that is because Microsoft is SQL server. That’s like their space. I think they really want to incentivize people to use more SQLesque things and in order to reduce the friction between using SQL and getting it into fabric really we’re talking like transactional systems SQL versus analytical systems fabric they want to make that if the the less friction you have between those two worlds the more you’ll want to just go buy the rest of fabric and I think
33:35 Microsoft is very wise in doing that because the spend end that comes through other things that happen inside fabric like houses other items makes a lot of sense to have that mirroring just available to you and I I believe what’s happening there is get started yeah and I believe what’s happening there is the SQL server is doing like transactions and then fabric is reaching in and going to grab those transactions and pull them in directly to fabric so that way it’s reading the transaction log and making sure records are up to date I like this the more I feel like a lot of times there are very consistent patterns in data engineering. Hey, I’m
34:11 Going to go I need to get this table. This table needs to be fully copied every day or this this table is a a full copy every time. Delete and replace. That’s a pattern that’s very wellnown. I want to incrementally load this table. Any new records that change, here’s the information that decides when a new record is updated or changed. I only want to grab the newest or the latest or the changed records of data and then want to go update my original table. that pattern is also very well known. So I feel like a lot of now that I’ve been doing data engineering for a lot longer now, I feel like the patterns of what I see data happening are becoming more
34:45 Repeatable and I want Microsoft just to solve more of those common patterns in an efficient way. And I feel like mirroring meets a lot of those needs. Thousand%. So dude, off to a good start. Do you want to ch pick the next one or do you want me to? yeah, I’ll pick the next one. Let’s see here. let’s let’s do this one. So, this is going to be a bit more of like a Okay, I this I understand this one. This award goes to This award is this is going to matter way more in 2026. What feature of fabric is going to matter way more inside 2026?
35:22 So, I guess this would be like the large impact award. So, for me, this one is what’s going to matter way more in 2026. I think the proliferation of workloads are going to be much more impactful this next year. I think people have been starting to explore around it. Microsoft seems to be putting a lot of effort behind it. There’s currently a competition going where people are able to build their ideas around what a workload is and decide what that looks like. but I think there’s there’s more thoughts
35:54 Around what we can do with workloads. So I think this is going to matter a whole lot more inside 2026. That would be my choice around workloads. Maybe I could throw in a secondary award here for this one as well. I think the one lake catalog is an underrated feature today. Catalog. Okay. So one lake catalog and there’s like security and things for admins in there as well. Purview doesn’t really tick the box for me. I don’t really like purview. It seems like a whole separate system that doesn’t work very well with fabric. However, I really like the one lake
36:27 Catalog. The one lake catalog makes a lot of sense to me. I can see all my tables. There’s semantic models there. There’s a little bit of lineage. I can start adding descriptions to stuff. There’s a little bit missing I think in the one lake catalog. We don’t have descriptions on lakehouses, lakehouse table names, and lakehouse column names. But I think with a little bit of enhancement, the one lake catalog could be very impactful for the organization for like data discovery, organizing your content, using tagging, adding domains to things. I think a lot of these other elements would make a lot of sense for
37:00 People who have large PowerBI environments where there’s going to be a lot of data pushed together. So I think the one lake catalog could be a great discoverability item for people as well. So I think the combination of workloads, fabric workloads and one lake catalog are two features that are going to be very impactful for this next year. That’s I hope so Mike I and I I really I have so much hope for the one lake catalog. I I still use it but again it’s I’m waiting for that governance speech that they talked about to be part of it because I would love for that to be the hub for data governance to go to
37:34 And to make that more friendly than rather list of items. But I do hope 2026 is year’s the one lake catalog. However, for me I have a few Mike I wanted to say fabric IQ as going to matter way more than 2026 but it’s in basketball and football if a guy doesn’t play long enough even if they had a great season they’re not eligible for the MVP or the eligible for the award. M. So for me, as much as I would love to
38:06 Say Fabric IQ is definitely the front runner here, just didn’t have enough reps yet. It came out what in the end of November if I’m not mistaken. It was right at the end of the year. I think it was announced at the Ignite conference. Announced there and then like like very soon thereafter you were able to start playing with it and start doing things with it. And again, I it seems like an interesting concept. What’s actually been deployed and developed [clears throat] doesn’t seem to me to hit the mark. And I’m not sure quite yet how to use it.
38:39 It’s still confusing to me. Yeah. And honestly, I think it is so it is so new because Mike, you and I were like, “Hey, we’re going to do two months of episodes on this. This is changing everything now.” And then our first few episodes we were looking at the two or three pages of documentation going okay not much here. What do we use this for again? Like oh so it’s just for a data agent. Okay. What’s an entity and and just going through that. So yeah
39:12 I think there is an award for fabric IQ. I just don’t know what it is yet. But I am then going to shift and then say it’s going to be data agents and they’re not going to be demos. They’re not going to be things that you’re just showing the team. The goal is the perfect synergy between a fabric data agent and co-pilot studio. That’s where a data agent should shine where the user no matter where they are can interact with their data as via the data agent in co-pilot studio
39:47 Wherever they’re at on on their in their workflow. if they’re in Teams, if they’re in Outlook, if they’re in fabric, heck, they should be able to on their phone if they have the questions are actually have start actions getting moved from that data, not just asking, “Show me a bar chart.” That’s where we’re trying to get to with a data agent. It is not just to be a more conversational way to build your dashboard for for consumers. it should be able to trigger responses to again what the data that’s
40:23 Actually being worked on by that team. Yeah. you can get those latest numbers but if that’s all you’re doing we missed the mark. So I see data agents with in in synergy with copilot studio and co-pilot agents to really be at that point where this is going to be something that everyone’s talking about. Yeah, I I would agree with you Tom. Think these are a lot of clear points and again I would also agree with you the fabric IQ I think it seems the sales pitch of fabric IQ seems really relevant
40:57 Like what what Microsoft markets it as makes a lot of sense how you actually build one and how much time is required to be spent to build one I think that’s a miss in my point right today currently you can start you can go into a semantic model and say from this semantic model build a ontology right that that works and it just kind builds out a number of items for you. But where’s the other automation? Why can’t I add a second ontology, a third one, a fourth one? There’s a lot of buttons and clicking and like it seems to be like a
41:30 Lot of manual effort to make it work to get the ontology that you want inside the fabric IQ pieces. And so I’m not necessarily confident that I want to spend a whole bunch of time spending building a bunch of fabric IQ things because I don’t even know if that’s going to add a lot of value to the data agents at this point. So it it feels like it’s something that they’re trying to prepare for. I I hope they spend a lot more time in 2026 automating how to get things into the ontology. How do you bind things together? How do you bring more tables together inside that space
42:04 To make it easier for people to use? So, we’ll see. I’m not sure about I’m I’m a little to your point, Tommy, I’m skeptical about that feature. Okay. No. So, but yeah, I think data agents is definitely we’re going to be leaning. I am intrigued, Mike, where we’re at at the middle of this year in 2026 where Fabric IQ is. , is it is this something that we’re putting on our LinkedIn profiles because it’s taking the world by storm? , I really hope it’s not one of those one and done. Obviously, you can see they’re investing in it, but it’s a it’s a bluecollar
42:39 Prospect. That’s what we’ll say. Just keep going with the analogy. All right. So, we got a few more left. Do you have another one that you really like or No, I’m indifferent on the last three here. I can I can do whatever. Maybe you pick the one that you’d like to talk about next. All right. So, let’s go with this is why we can’t have nice things. I love this one. I love this one so much. But what this award is is it’s an amazing feature, but people are going to
43:12 Break it because they are not either have the skill or the experience in that background. We’ve talked about this when it comes to Microsoft making a prettier UI, but it’s missing all the bells and whistles and all the configuration features. just for someone to do copy a copy job. There’s a lot that goes into that. So with all these things that are coming out that are easy shiny to look at and do have some of that wow factor. Yep. Well, if you don’t have the necessary background and experience, things break
43:47 And I am going to take actually I do want to hear yours first. Okay. I’m I’m going to go a little bit on the off the rails this one. So Tommy, the way you describe the feature is like you have to have a lot of knowledge, you can’t use it, it won’t work right. I’m just going after implementations that in my mind just missed the mark. Oh, interesting. Okay. Things that So when I wrote this one, you this is why we can’t have nice things for me. I looked at this question of this award and thought this is where we just flubbed flubbed the deployment of it. We just flubbed the whole item of it. , so one of the things I will I’ll probably say around this one is the
44:23 New feature that was highly touted I guess was the in inside a fabric in environment. So when you’re building in fabric, you click at the little icon in the bottom lefthand corner. You change from PowerBI to fabric. The whole new workspace tabs feature. Whatever that is, I don’t use it. I don’t like it. That’s just a to me that’s just like a lot of noise. There’s a whole bunch of like colors that appear and then you have a whole bunch of workspaces that are like colored. That’s just to me a flubbed feature. It just I understand maybe why they did it of
44:58 Switching between tabs too much, but in general I can’t stand the feature. I don’t like it. It seems awkward. I don’t want to use it. I just stay in the PowerBI experience as much as I can because those colored tabs just leave it alone. Like what the heck? That was just a miss of a feature. I don’t know what it’s there for. Hilarious. That’s a pretty good That’s a pretty Yeah. No, because you’re like, I will keep my tabs. I will keep showing on multiple monitors. Yeah, I don’t have a problem with this because I have a lot of screen space, but I I
45:30 Just in general can’t stand I when I go from the PowerBI space and then over to the fabric space, I get messed up all over the place. I can’t find where things are. Things are reoriented. If I go from the PowerBI workspace over to the fabric workspace, the workspaces icon gets stuck in different places. So, in PowerBI, the workspace is at the bottom. over in fabric work pieces are at the top. Like it’s just make them all the same. I just get messed up. I don’t like it. And on top of that too, the tabs are so small that you can’t read anything that it says. it’s interesting though too
46:05 Because initially I was going to love the idea because I I could see where they were coming from because I would get annoyed going to fab like, “Oh no, this browser refreshed while I was waiting and now I have to rerun things.” It was a nice idea but they changed the way people work with computers and how like our experience working in browsers and try to change that complete workflow. So yeah I I didn’t want to admit it Mike but because I really do feel bad for the
46:37 Team because it was something that I was hoping to have but just I guess not that execution. I would be intrigued if Mike I wonder if the number is higher than you think if we pulled all of our listeners. Well, I know already Johnny Winter is already, is going, “What? What are you talking about here?” workspaces tabs are awesome. So, I could be just a stick in the mud here. I think it’s a very polarizing feature. I think you either love it or hate it. I’m of the opinion I don’t love it. I don’t think it’s a very useful I think trying to force PowerBI or fabric to live only in one web browser window is a
47:14 Mess. I I I don’t I just don’t think it’s one of the things I think I wish I could do a lot more of when I’m in data bicks and I’m going to I’m going to directly throw some shots fired here. If I’m in data bicks and I need to open another tab, almost anything I click on in data bicks, I can controlclick that item and it will immediately open up a new tab. And yeah, I may have a couple tabs open in the browser for what I need, but for whatever reason, Tommy, I cannot seem to control-click a workspace. You can’t control-click on a workspace and get a new and a new tab to open. So, just Okay, I understand why you made the whole tab
47:48 Features in Fabric. Get rid of that. Just allow me to controlclick any of the workspaces and allow me to open up a brand new tab for that workspace. I think that would actually solve a lot more of my problems. I think that would be much more helpful for to me than having to add this whole new tabbed feature. Just seems like wasted effort in my opinion. Dude, I had never thought about that. But how frustrating is that with the workspaces? Yeah, there’s a lot of things you can’t just or or control-click or right click because I want to know I not that I get scared when my browser opens a new tab, but I like to control that, and I
48:21 Like it going in the background first, especially when I’m working at things. I don’t want this like pause going, “What happened?” And then a new tab opens up without me wanting. So that that’s the feature. That’s your feature Microsoft that you work on. Yes. I’ll also let me give you So that was my one thing that I complained about here. This is why we can’t have nice things. the other thing the other reason why we can’t have nice things. Okay, Tommy, this is going to be very nitpicky here as well. You remember how we had workspaces and when you’d open up the workspaces pane, there is like the cut off or limitation of the workspaces names, right? So the
48:56 Names would always be shortened because the name of the workspace was too long. Well, Microsoft said, “Look, we’ve listened to you. We are we’re going to make this a lot better for you. Oh, and by the way, we’re going to give you a button in the upper right hand corner, and that button will allow you to expand the window, but there’s no shrinking and growing and like I can’t make it as wide as I want to see names of things. It just doubles the size of the window.” And I look at myself going, “You guys just didn’t think listen at all.” Like, I don’t want a button at the top of my workspace
49:29 Window that makes it wider, but still sets me to a specific width. Every other UI that I’ve seen around like things that expand and can close or, everything I’m used to in PowerBI desktop, all those panes are fully expandable up to a certain maximum length size and can shrink down. I don’t want a button. I want to handle on the on the workspaces that I can grow and shrink it as much as I want. So to me, this is another one of those, flubbed features if I look at it. Okay, someone put this in. I I feel like the
50:02 PM missed the mark on that one. I I didn’t want a button to just expand the the the workspace pane. I wanted a handle to make it open or bigger, like smaller or larger. Like that’s what I wanted. It doesn’t help me. And it’s one of those almost nearly frustrating things. because I feel like a lot of other UIs have that. So it’s like I this is still built on PowerBI from 2015 which I guess is old now and it’s fine but I’m usually working Mike on my multiple monitors, right? So it’s not always the
50:37 Most frustrating. Yeah, I got space. I got space. I got a lot of space. But I do know I remember it was actually when I first started working with PowerBI, someone’s like, “The report’s wrong. All the things are bad. you made it terrible.” I’m like, “Wow, okay, let’s go through it.” And then I didn’t I realized that the person was zoomed in 200% resolution. Oh yeah. And it was already on like a 12-in laptop. I’m like, “Well, you’re going to have trouble working with this. You’re gonna [laughter] have trouble working in anything.” Yeah. So, , it’s not just PowerBI, but you had to
51:12 Think about that and I anytime I’m on the road, , or I’m thinking of other people here too where just make it customizable. Just drag and drop. We all love clicky clicky draggy droppy. Do that with the UI rather than just data. , I like that one. All right. So, I am going to take this a little different direction because that was that’s a really good one. I’m going with GitHub, Mike. I’m going with Git and Fabric. Oh, interesting. Yeah. So, I the number one question that I get on LinkedIn when I don’t work with
51:48 A client is, would you mind sitting down with our team for an hour or 30 minutes just to walk through some things we’re going through with Git? It’s the number one thing why people will cold like just pick your brain and about how do we navigate this in fabric? how far should we go? , and just hearing the frustrations and the nice thing though, Mike, is most of the te people on the teams, they want to handle it themselves. Like the reason they’re trying to learn it. Yep. , but they’re trying to learn it. And
52:22 It’s really going through with not just an individual member, but that leader or that manager and the rest of the team and then trying to delegate that and th those people going, I’m barely touching this and I have to be part of all these get these meetings and basically everyone is really just trying to understand how deep the well goes. So let me react to this the GitHub piece. I think this is a very advanced feature that we’re bringing to a lot of business intelligence developers that have not
52:54 Been required to use this. Right? The most amount of effort that we’ve had to spend as business developers or BI developers is where’s the report? What’s the Excel file? How do we store back of it backup of that file on SharePoint? That’s really it’s backups. Basically, we think about backups of things. We’re not thinking about deployment pipelines. We’re not thinking about deploying things. I think this is this is definitely a borrowed experience from the IT realm. As your PowerBI environment grows to really large enterprisegrade things, you’re going to have a lot of more a lot more people or
53:27 Education around what DevOps is, how to store backups, and how to deploy things, right? You’re going to you’re going to seek out a lot of automation to deploy. We’re working on some really large customers right now, and that’s one of our main stories with them is don’t do things manually. let’s figure out how to automate the deployment as much as we can through different environments and really focusing on that effort. I think that’s a very solid skill that’s required when you get to large scale bigger enterprise deployments because you can’t have you can’t you can’t rely on a Tommy
54:02 Pushing through a report from dev to test to prod on a Thursday afternoon and assuming everything’s going to be just right. Right. We want we want as much of this automated as possible. So, we spend our time working on the automation. And I think this is where a lot of get stuff starts. It’s just a new world for a lot of business developers. So, I think you’re right, Tommy. It’s it’s a really good thing to have. There’s a lot of questions around how to use it, how to how to use it effectively. yeah, I agree. That’s a good one. All right. So, I don’t know if we’re at time, but I’ve I really enjoy this.
54:35 We do have one more, but it’s up to you if you want to do it. Let’s do the last one. I think we got we actually have two more, right? Oh, so we have this like I think you may be right. So So I’m going to do this other one here. Okay, Tommy, here’s the next reward. The best feature that requires a long awkward conversation. I think this is a funny one. what what feature, Tommy, would you say is a a feature of fabric that would be requiring that long awkward conversation? Well, you already took my thunder because I was going to say fabric capacity billing and the
55:09 Model changes and all the different changes that they had and whether you’re paying for a database and you don’t have to pay for copilot, but you probably should. , so it’s already getting awkward there. But, , in an effort to not be too redundant here. Yes, that’s the first one. That that’s the first awkward conversation. So, I was going to say, Tommy, I thought your your awkward conversation would have been metric sets. [laughter] You see, no, I was s I was saving that one. So, oh, I was saving that for the nice things one. Dang it. So, [laughter] that’s right. I had a Oh, I wrote it down, too. And I was like, when am I
55:43 Using that? So, metric sets. Ah, no one really cared anyways. So, I don’t think it’s that, [laughter] dude. It’s that big of a loss. So, apparently, dude, I don’t even know if anyone was actually using them productionwise. I never heard anyone who was like at least from a larger organization who was like, “No, we rely on metric sets.” Did you ever hear that at all? No, it was I think it was a novelty. Again, it’s one of these features that I think had a lot of excitement behind it when it came out. I thought it was a really neat feature. I really like this idea of building the metric and then
56:17 Applying the dimensions to the metric. I think maybe where we got stuck a little bit on that feature was it just took too many clicks to get one of one metric built and we needed a bit more automation to figure out like what if you could go to a report and say in this report here’s all the metrics you should be using like here’s all the here’s all the measures that are being used in different visuals here’s all the dimensions that apply to them we’re going to auto build you a handful of metrics that might have been a bit more useful take an existing report with information in it to help you generate that information would would have been interesting I think the only
56:49 Reason they deprecated is because I had the bet with Emily. With Emily on No, no, it was Emily. It was Carly. You You were betting Carly and you said this is You said I think half of all usage from fabric was going to be No, it was even worse than that. I said in five years once metric sets become GA. Yep. It will have more usage than reports. Tommy, I think you really missed the mark on that one. Okay. Did you bet something on that one? Did you bet like a a steak or some food?
57:22 I think it was a It was one of those just gentleman’s friendly one because I’m like I also don’t know the next time I’m seeing Carly. , so it was probably she probably knows because she was right about this, but I think it was more of a a pride just gentleman’s bed. But we’ll make we’ll make a t-shirt and say Carly was right, Tommy was wrong. That’s [laughter] fine. That’s fair. I Tommy cared more about metric system than he should have. So, [laughter] but to me it was the the initial idea it was what was what scorecards were trying to be right where at the end of the day
57:58 A team a user a leader can quickly have trust and understanding on what the core important numbers are and what metric sets did was take that up a level to actually say no you want to know the the the problem I’ve had for years where what’s the number of leads what’s the number of members Well, you got to look at this report, but don’t look at his because he has a filter on it. And everyone gets frustrated that way. So, no, here’s your core numbers. This is a governance piece now. So, these are the, based number of things leadership where you want to look all
58:30 The reports have filters and things that we do and you’re grouping by things and we have to make it look pretty. Here is how many members we have. Here is what our pro number of sales or our units in or whatever are. And I love that because but I obviously people have to use it and want to go there and I think that’s what Microsoft’s finding is they may get people complaining about these things but no one was gravitating to it. Let’s go to the my other one. So you you said metric you said licensing right
59:03 And metric sets of those awkward conversations. So my feature for the best feature that requires a long awkward conversation is drum roll please task flow diagram [laughter] it’s there every time I do a demo the task flow thing window is open it’s there every time I say we don’t use that we close it and I just click a little button boop and it closes itself that’s a very visible feature that and then people ask ask me well how do you use it and then I go in and say, “Oh, well, well, it helps
59:37 You group things. It gives you like a a general broad diagram of like how things are linking together. you could have a lot of things in the workspace. It does this automatic filtering and people are just like, “Oh.” And then we move on. So, for me, explaining what a task flow for for the workspace is just awkward. It’s the whole thing is just awkward to me when I talk to people about it. It’s there. It’s part of the UI. most often I do not use it. I’m using a lot more folders and I’m using a lot more or organization in other ways. So, to me, that’s one of the areas where I see there’s a an
1:00:11 Awkward conversation with with people on the phone talking about PowerBI. It is even worse, too, Mike. Especially when you actually have a halfbuilt task flow that you have to quickly hide out of. Yeah, correct. Like, no, no, you built something. Show me. You’re like, it’s really not worth it. It’s not worth it. It doesn’t there’s the lemon is not really worth that squeeze at the end of the day. And I I don’t think that feature will ever go away. I also don’t see any new features coming to task flows. No. So I I don’t there’s nothing that the task flow diagrams is not getting any net new features. So it’s just going to be there and it’s going to be in part of the product. So
1:00:42 The day that the Tesla’s died for me day one and it I [laughter] think it was day one when the it was so buggy where I want I like to rename the tassels obviously. Sure. However, it would not like you would type three letters. Yeah, that was the name of your task flow. It would be it was overaggressively like accepting the name too fast. So, what I had to do was write the name on notepad and copy paste. Copy. That would be weird. I’m like, what am I doing?
1:01:13 That was in the early days. It doesn’t do that anymore. They fixed that. No, but it’s still like you want it to be a little more creative, I guess. it is so not [snorts] static. it is so rigid and and you’re like I already deal with rigid stuff. I don’t need another wannabe. yeah, what ? Like it’s where like make it like muro thing. in terms of the because I think that’s what it’s trying to be especially if I’m just doing this for myself to brain brainstorm or plan out like you want that at least that
1:01:48 Freedom to be a little more creative but that this is what you’re going to show everyone. So task flows, that’s a good one. That’s a good one. So I I will say that one. And then the other one I will just another tag on here would be one around I’ll just call it dashboards. I don’t I don’t ever use dashboards. That single pane of glass where you have a whole bunch of visuals and you click on one and it goes directly to the report where the visual lives. Oh, people people still use the word dashboard and report interchangeably. But dashboards in PowerBI are a very
1:02:21 Specific thing and I don’t I don’t even talk about them anymore. and people if you upload a file like so the only time I see a dashboard appear is like when you upload a file from your local machine using the upload files dialogue you get the semantic model you get the report and you get some very simple dashboard. I don’t know how many people actually use dashboards. I think it’s very very limit limited now in in these neat features here. So, that’s one that I’ll say is another one that is it’s just an awkward conversation. Again, it’s this is a feature. It exists. It’s a single
1:02:53 Pane of glass. You can click on the visual, go into the thing, but I don’t ever use it. And so, just we move on. Yeah. But that’s also Mike, I think we’ve been saying that we haven’t been having that awkward conversation since 2018. It’s been out for a while. Yep. I agree. So, anyways, that’s it. I think that’s all we have for our awards. I hope you’ve enjoyed our awards today. just a little bit of poking of fun and things that we like and things that we found that were interesting, things you may want to look at and things you may want to you have also felt some pain around talking through some of these
1:03:27 Features as well. That being said, we hope you’ve enjoyed the podcast. We really appreciate you listeners. We do this for you and have a lot of fun with the community. Thank you community for having a lot of chat here in the window on YouTube. We really appreciate that as well. If you want to join us on the chat on YouTube, feel free. Come on over to YouTube. We have a chat. We do these live every Tuesday and Thursday and we’ll be online. With that being said, Tommy, where else can you find the podcast? You can find us on Apple, Spotify, wherever get your podcast. Make sure to subscribe and leave a rating. it helps us out a ton. Do you have a question idea or a topic that you want us to talk about in a future episode?
1:04:00 Maybe this whole make up your award thing takes off and you have awards that you want us to do. Well, head over to PowerBI tipsodcast. Leave your name and a great question. And finally, join us live every Tuesday and Thursday, 7:30 a.m. Central, and join us on all of PowerBI tips social media channels. Thank you all so much. Appreciate you, and we’ll see you more in the the year 2026. [music]
1:04:49 That’s the battle. [music]
Thank You
Thanks for joining us for the first-ever Fabric Awards! Got an award category you think we missed? Head to powerbi.tips/empodcast and submit it — maybe we’ll make this an annual tradition.
Want to catch us live? Join every Tuesday and Thursday at 7:30 AM Central on YouTube and LinkedIn.
Listen on Spotify, Apple Podcasts, or wherever you get your podcasts.