Mailbag! More Datamarts - Ep. 375

Mike, Tommy, and Seth dig into listener questions - focusing on datamarts, citizen developer workflows, and how teams should think about tooling and access constraints.

News & Announcements

• Chicagoland Power BI Meetup
• Link
• Subscribe to the podcast
• Link
• Connect with Mike Carlo on LinkedIn
• Connect with Seth Bauer on LinkedIn
• Connect with Tommy Puglia on LinkedIn

Main Discussion: Mailbag - More Datamarts

This episode is a classic mailbag format: the team works through listener prompts and clarifies a few recurring themes - especially what “citizen developer” means in the Power BI/Fabric world, and how tooling choices change based on access, scale, and ownership.

Key themes they cover:

• Citizen developer is a spectrum: It is less about a job title and more about the context - building fast answers for a team vs. building scalable assets for an organization.
• Access constraints shape workflows: Many organizations restrict features or environments, which forces creative solutions (and impacts adoption of newer capabilities).
• Datamarts and “where work happens”: They contrast Power Query + Desktop-centric approaches with more Fabric-first patterns (lakehouse/notebooks), and when each path makes sense.

Looking Forward

As Fabric and the Power BI service keep adding more “developer-grade” experiences, teams will have more options to move upstream work (data shaping, semantic modeling, and governance) out of individual Desktop files. The practical win is repeatability - fewer one-off hero builds and more shared, reusable assets.

Episode Transcript

0:35 good morning and welcome back to the explicit measures podcast with Tommy Seth and Mike good morning everyone good morning and a happy Tuesday beautiful it’s another Tuesday we are now in the week of Thanksgiving so just to orient ourselves here this is a pre-recorded episode so we Rec for this earlier beforehand but in the US we celebrate Happy Thanksgiving a happy thing we don’t celebrate just Thanksgiving it’s a happy Be Happy Thanksgiving we get to eat turkey and stuff I wonder guy now on

1:08 turkey and stuff I wonder guy now on Thanksgiving yeah well we we we go with a larger group of people so of course we’re having turkey but of course I found this really good place really good meat shop that does their own pigs and pigs and stuff so it’s good we we have started a tradition with our family we when the women of the family would cook a really nice turkey and have this big meal but it was a lot of work a lot of stress on everyone to just get all the meal and the meal prep together we have fallen

1:39 the meal prep together we have fallen into a new pattern where we go to Mission Barbecue have you heard of yeah and they have an incredible like turkey roasted slow roasted all day long kind roasted slow roasted all day long thing and it is amaz they must get up of thing and it is amaz they must get up early to do it I don’t know what they do but they you take it to order it beforehand and it’s like super soft very tender it’s not dried out any way and it’s very consistently good all the time so we we got to the point of like I think the when we got together we’re

2:09 think the when we got together we’re like if we just keep doing this we don’t have to cook turkeys for the entire day so we’re like yeah let’s do that so it’s what we’ve been doing for the last couple years it’s been going to like Mission Barbecue it’s it’s definitely probably quite pricier then getting your own turkey and do it yourself but yeah the time everyone’s happier I’ll just say that everyone’s happier by doing that I I used to be lasagna guy myself for Thanksgiving for Thanksgiving Tommy oh yeah but what when people say turkey’s overrated maybe you’re overrated you’re not doing the turkey right what I found is you salt

2:42 turkey right what I found is you salt that turkey for a day literally in a bath of salt oh man even heard of you’re brining it yeah it yeah so so eat a pickle turkey and you’ll be happy more or less but it basically soaks the flavor if you want a dried turkey just put it in the oven don’t do anything to it you to your obsessed point you brine the turkey just basically give it a bath salt leave it in the fridge for a day Turkey bath salts yeah yeah that’s funny I always I always remember around Thanksgiving time

3:12 remember around Thanksgiving time there’s always like the the warnings I guess it would be warnings or like cautions everyone who people fry their turkeys fryer and so I every every year every year for like the last couple years it’s been like warning don’t fry your turkey and there’s all these pictures of people their turkeys in that I guess have moisture on them and they put it in the super hot oil and the oil just like decides to like turn into a flame so the oil the steam it like instantly steams this the water the water steams out the oil turns into a mist and then it hits the flame and it

3:43 mist and then it hits the flame and it turns into this Fireball and you’re like yeah I don’t look like I don’t think I’m ever doing that sounds delicious though very slow very slow drop it in and dry and as dry as possible yeah was funny to me so anyways this is this is that time of year we get all those weird messages I was also talking with someone in Europe trying to schedule a meeting and I’m like oh yeah well you a meeting and I’m like oh yeah well in the US we have Thanksgiving we know in the US we have Thanksgiving we can’t do that they’re like what are you talking about I’m like oh yes that’s a that’s a us us thing we have a weird

4:13 that’s a us us thing we have a weird holiday okay I keep forgetting talking internationally other countries don’t really do our holidays I keep getting confused and there’s other countries that do a lot more holidays than us I feel like Canada is a holiday every other week it feels like to me and they’re just off a little bit yeah I’m curious of our nonn North American friends have turkeys is that just a north and I know that sounds so naive if someone probably listening going you idiot it’s not a thing it’s not a thing in other places no just not Thanksgiving Just Turkeys oh just eating a turkey right gotcha okay

4:45 just eating a turkey right gotcha okay so say like I’m pretty sure they don’t I’m pretty sure South America doesn’t celebrate Thanksgiving but that’s okay all right all right well moving on then so so a couple of quick news and announcements first news announcement thing is we are going to be down at wait a minute we already did this didn’t we I think this has already happened actually just kidding we had a great time at the Chicago Meetup I was going to announce having us guys but because yeah thanks for having us we appreciate it I think it was a good meeting U I’m hoping that

5:17 it was a good meeting U I’m hoping that it was it went well so that’s thank you for attending for those who came that’s a weird news item okay well this is what happens when you record things in the past for something in the future that being said any other new or topics that we want to talk about anything else that came up we can’t really talk about anything at ignite because we haven’t been there yet and it doesn’t there’s been no announcements so there’s nothing that we can talk about there great time I had such a great time I had a good timeo yeah yeah people Tommy yes there a little bit rainy because

5:47 little bit rainy because Chicago the fall it may it actually may have snowed when we were there it could have snowed I don’t know it supposed to there was that one time when Tommy was walking he fell right on his bum because he was slipping on the snow that time that happened last talking about things that may probably never will happen awesome all right well with that being said then we should probably just roll into our mailbag so today our topic is just a mailbag and we’re going to talk about tools for citizen developers would our reader Seth would you mind

6:19 our reader Seth would you mind reading us through our question today sure sure qu if it’s a question starts off great no real question here comma all right we’re already on a winner who chose this just a comment on your thinking about data marks oh gu we’re in trouble wait doesn’t everybody just agree with us does this come from a mirror what’s who’s this from okay I have no idea what Microsoft is up to with these nope not air but I will say that as a citizen

6:49 air but I will say that as a citizen developer user who has zero input into the development of our bi infrastructure and consequently no access to fabric or literally any other data warehousing solution data Marts have been a real Lifesaver I don’t have access to fabric so I don’t know how data Marts compare exactly but for me data Marts allow me to build build fairly extensive data models with a huge variety of data sources and then build a bunch of thin reports on top of them they load quickly

7:19 reports on top of them they load quickly and are easy to troubleshoot anyway I thought data Marts came in for a bit of an undo drubbing in the kill it or complete it episode so just wanted to pop in and defend them would I like to have fabric sure but I actually haven’t come up against a single thing you guys have described in the podcast that D Mars couldn’t do just pulling that out there putting that out there thanks again for everything you’re doing Andrew and then wrote writes in that he has previously been I guess part of the

7:52 previously been I guess part of the mailbag so thank you Andrew again for this this disagreement that we will now discuss and for your previous mailbag submission wow this is a great question so one thank you very much for your question this is going to be one to unpack a little bit here because this is we’re now talking the break point between I think powerbi and what is fabric moving forward and so this is an this is potentially an organizational decision

8:23 potentially an organizational decision and something that may be taken out of your hands you may not be able to get anything beyond the scopea data Mark so let’s let’s maybe give a little bit of History here at least how we see it harbi is being built people are still trying to build this architecture of hey I have data in Source systems I need to manipulate or shape or create changes on that data power query either in powerbi or in data flows gen one actually would be your only way to transform data so those are your two transforms of data that’s all you could do initially in

8:54 that’s all you could do initially in powerbi the data flows gen one would save all the files into a file stoat format it with CSV it’s in the clouds if you if or you could use your own storage account if you wanted so there was some availability to store things but you’re you were basically relegated to only using power query for those things and I think data Mars

9:14 those things and I think data Mars was opened up for powerbi premium so premium per user right and if you had a pcq or a premium capacity you could turn on data Mars but data Mars is not a fabric only item it is a pure powerbi item and it feels a lot like a SQL Server is that a fair statement anything I missed there about like the history of like how how we got here to data Mars no I think it I think it was those it was the ETL pipe but also the the way to create an extensible

9:44 also the the way to create an extensible model that like was easily queriable and you could build on to like it it was the mini warehouse right like yes in the mix I think we had an issue with like I I think I’m not sure which came out first but I feel like semantic sorry what am I drop the name on not semantic I’m synapse I feel like I feel like synapse came out just before data Mars came out so I feel like it was like synapse was released synaps was supposed to work

10:15 released synaps was supposed to work well with powerbi but it was a separate paid thing it seemed a little bit too expensive and so people were like ah maybe I won’t use it one of my issues that I think we had here was there was no place to like have a sequel like a sequel on demand mhm right and if we think about like the loading pattern of what we’re talking about here we have data in Source systems we need to transform it and put it somewhere we need to pick it up again and put it back into po semantic model and I think the pattern at the time was a lot around doing it all inside SQL servers that’s

10:45 doing it all inside SQL servers that’s because that’s what people were used to doing and I in Seth I remember very vidly at one of the MVP Summits you and I were talking about data bricks and I was like oh my word this is going to change how people do things because I was like I why do we need the SQL Server I kept I kept having really big push back against why is the SQL Server here because it’s just reading tables and it’s got to be on all month long just for me to read the tables once a day when I load them in import mode and I was having a lot of like this makes no

11:15 was having a lot of like this makes no sense to me and I think that’s where data Mars came in as that that SQL serverless part well it it I I kind I I I get what you’re saying but I don’t I agree right synaps was the Enterprise yes yes solution at least I think a year or two two years two years before yeah but the Enterprise side right like Hey we’re gonna we’re going to the cloud with our warehousing solution Etc data Marts was more how

11:45 solution Etc data Marts was more how can we take the Enterprise side of things and give it to the business yes in a way that they could interact with it and and and it was like I think yes amongst the Enterprise side conversation is there a lot of was there a lot of I think rapid development than like open opportunities that data bricks brought to the market 100% yes right but in the context of where I think I’m not going to call it

12:16 think I’m not going to call it frustration but a little level set of where where we typically talk in this podcast is focused on a lot of the areas that we’re engaged in no which tends to lean into the Enterprise side in space yes and I think it’s valid that Andrew is pushing and and we should refocus a bit more in in what is available to some of the folks in the ecosystem and I think what he clearly articulated here

12:47 think what he clearly articulated here is is like great guys love love the chat love what you guys are talking about I I don’t have access to it so this this is the solution that people like me not just me right yep labeled citizen developers are getting a lot of value out of data Marts and then in my mind I was like who who was it that poo pooed so hard on data Marts was that you Tommy we all I think we we all played a part we all we all contributed but this is I

13:17 we all we all contributed but this is I agree I’m not sure I poo pooed it I just said I wouldn’t build anything new in it because I don’t think they’re going to build anything new like so assumption that we can with that we can leverage the environment right next like and the frustrating part about that and I’ll let you talk to me in a second frustrating part about that is it’s like in the past I could I could understand right likey oh yeah you got to spool up these OU services but it’s like as an organization like you not having access is literally just it’s all right next

13:48 is literally just it’s all right next door yes right like it’s a knob now it’s literally a click of a button now to get to get it up and turn it on yeah and so this is important from the unpacking point because I really do want to reiterate what Seth is saying for two of us here if I wanted to turn anything on in a Microsoft capacity for three of us all three of us have this ability Seth Seth answers to no one if Seth turns on things in FA Seth turns on things in fabric let’s just let’s be clear here all of us are the data Zars of our whatever our world is right we can all

14:18 whatever our world is right we can all turn on things whenever we want however we want so like we don’t have the limits that organizations put on their people they yeah and no you’re point I love you’re going with this one yeah and you’re so you’re spot on so and we’re also MVPs with that nerdy attitude so we’re always going to be trying the latest and greatest I can’t imagine how many people listen to us and we keep saying Jupiter notebooks oh my gosh this fabric notebook like yeah yeah talking about again yeah well guess what I can’t and how many of the features I actually wonder how many of the features we

14:48 wonder how many of the features we actually talk about people are like that’s great but we’re still on Pro like yeah we’re still like I have gen one and my company’s running on that yeah that’s still the case and I think it’s important to us to think about the world we live in compared to a lot of people I I’m five years or four years mean I I’m five years or four years removed from being in a company was like we need to turn this on they’re like well we’re not going to because X Y and Z and you’re like okay I can’t play with this I’m not going to spend my own personal money to do so no so the

15:19 personal money to do so no so the solution that Andrew is talking about here and data flow is Gen one even if you just take premium per user right not even a premium capacity this is to his point it’s a he used the word Lifesaver and he’s not just like no it’s cool he’s like no it’s runs this runs our core stuff this is business critical to what we do and I think a lot of times yes I would recommend someone to move forward as much as you can but that’s not always the like that’s not always going to be applicable that’s not always a reality for a lot of people

15:50 always a reality for a lot of people you’re bringing up a really interesting point here so that there’s there’s this is there’s there is no real question here there is just discussion points around this one but I think one of the points I’m hearing around the discussion here is how do you work with an organization who doesn’t understand the value of what your team is producing and therefore doesn’t put the money in the right places and and what by this is if you think about what we’re doing here there’s a certain amount of budget that’s being allocated to what we’re going to do in PBI and we’re saying at this level again whatever this organization is hey we’ve

16:21 whatever this organization is hey we’ve decided that premium per user is the thing for us 20 bucks are now coming April 1st will be again the weirdest date ever $24 per user so at $24 per user your organization is finding enough value from what you’re producing to say let’s just do that let’s just make it $24 a user you get all the premium things of parbi data Marts just show up you can now have this full really what this is is this is a full SQL Server that’s now inside fabric so do data flows load data to tables somewhere

16:51 data flows load data to tables somewhere get it into a SQL Server does data Mart do store procedures or something no it’s more I think just access to just it’s just table access query the query the tables in ways you can’t do with data flows right and you can do views inside a data Mart is that a correct statement you can tell how how little I actually use data Marts because I don’t know these features of data Mars I’m not sure we might have to Google that one maybe Tommy can Google that one in real time I don’t think are the stored

17:22 time I don’t think are the stored procedures I’m 95% confident it’s not there okay yeah I don’t think store maybe store procedures are not there but maybe views are so tables and Views maybe is that something but I remember when so I know the feature of this data data Mars came out of Charles web that was the team that was doing it and Charles was describing he’s like hey I got get this really cool thing let me let’s talk about this thing and you described this whole like merge of like analysis services and SQL

17:52 merge of like analysis services and SQL and being the same thing all the same item here and it was really I got to be honest it was was very intriguing to me but I’m like I’m I’m getting the vibe that they’re taking the analysis Services engine and the SQL engine and they’re really merging them together in a very weird way that was like okay we’re going to give you one UI that lets you jump between the two different like engines basically a SQL engine and Analysis Services even though it runs on the same machine and so the

18:22 it runs on the same machine and so the whole data March was very intriguing to me me personally was like oh but I have my data flows gen one I don’t need that

18:28 my data flows gen one I don’t need that why do I need to do all that data engineering why do I need that again it was more of like what I’m comfortable with I was more comfortable just doing things in powerbi I wasn’t as comfortable inside SQL and I’m seeing the same effect now in fabric like I go into companies that have a lot of SQL knowledge they all want to start up with the data the data warehouse and they want to run store procedures they want to run SQL statements they want to make views like this is a place where people are comfortable yeah and I’m not disagreeing with here in no way am I trying to disagree I’m just saying that

18:58 trying to disagree I’m just saying that there’s a there’s a comfort that your organization has about the skills that you’ve hired and so data Marts I think met that need for that SQL Developer skill that needed to be brought to powerbi because it is a it’s a data skill can I can I challenge you to think a a little more open than just data march here because I’m assuming no I’m only thinking about data Marts that’s it think about was part of the tool belt for Andrew or someone in this position right you’re using premium per user think of the solution they’re probably already using gen flow data flows I’m

19:29 already using gen flow data flows I’m assuming I would and I would make a very good assumption here I feel that Andrew and his they have to be using gen one Gen 2 doesn’t exist inside they’re using this as Mission critical I’m not saying they’re some using I’m saying that it’s part of the workflow so it’s not just data March but it’s going to be data flows gen one right yes and the a company can run on this it just because you’re not on fabric you can have citizen developers utilize these I I don’t want to call them older tools because we know that the technolog is

20:01 because we know that the technolog is always going to run faster than the business like yeah but it’s still in preview like it’s been out like I understand I understand but like yeah you’re the for two years now two years three years two years sto the company though they never they never it’s still in preview so the tag of preview means unsupported by Microsoft if something doesn’t work they’re not going to support it like there’s less helpto knows that what’s that do you think the CTO knows that I bet this team right I don’t know but okay

20:32 this team right I don’t know but okay good point going back to what I was saying earlier around this this Camp of like okay does the business really understand and is aware of what Andrew is doing is it right yeah what Andrew is doing and does if Andrew said to the leadership say hey look guys we’re using this thing it’s called Data Marts it’s working for us all of our major certified reports they’re running on on data flows I data Marts we’re happy yeah FYI leadership this thing’s INE in preview and FYI this is not

21:03 INE in preview and FYI this is not supported by Microsoft and it hasn’t been for 2 years and I don’t see it getting any new features here in the near term I wonder what that leadership conversation would be like and I doubt but Tommy to your point though I don’t think I would I would argue that leadership is not acknowledging all the the maybe potent or or knows about hey it just works the report shows up it’s happy everything’s going good no let me let me ask you a question Mike if you’re Andrew and you you I’m looking

21:34 you’re Andrew and you you I’m looking for new for new company back you back off all of all of your experience right and Leadership and you’re running reports for an organization workload is way high right sure and and you now are presented with I could go build things separate and outside of data Mars M or I could take this preview feature and and save 40 50% of my time yeah you’re going

22:05 save 40 50% of my time yeah you’re going to take it right so like to make an argument that you would want to go pull that forward even even though Mike you’re not validating that with leadership are you doing that I don’t so in my for in my former self if you talked to me before I do a lot of Consulting if you talk me like year one year two of Consulting I probably would have said Microsoft has been my story would been Microsoft is pretty good about putting things into preview and getting them out cuz I think up until when data March showed up most things came into

22:35 showed up most things came into preview it would live a little bit and then would get kicked out into an actual thing and the whole idea was like it’s preview because it’s not going to live there forever that was the point and I think if I looked at the track history of what microft was doing in just powerbi earlier on a lot of things would go to preview goals was a preview for a bit and then it like rolled on and did its own thing like so I I would I at the time I would not have had enough input or information in my mind to say it’s not going to stay preview forever I wouldn’t been I wouldn’t have known that so I think in

23:05 wouldn’t have known that so I think in that in that regard I would have said hey look we’re a little early in the curve here let’s just do it this is going to add a lot of value for us it’s already in a space that we understand let’s do it so you’re saying that to yourself nobody nobody nobody is telling you how to do your job you’re just in charge of building all the reports I’m just gonna I’m trying to reports I would have said yes let’s do it Microsoft will eventually make it go it will be going GA so that’s where we’re at though but then I’m looking at it now and so again if I had to look at my so

23:35 and so again if I had to look at my so again fast forward to what I’ve learned now no no I said I have to you no you can’t why can’t I right because you’re giving recommendations in the cont in the context of this is this is Andrew’s situation based on what he’s describing yeah right that’s true I’m saving exponent I’m producing more value to the business by using this feature that is in preview and this is this is why we talked about that the whole episode and why I leaned in a little bit because I think the vast majority of people who

24:05 think the vast majority of people who have been using Microsoft products have been shooting preview features out into their production environments and this is where it’s been like to your point well typically it’s on this Caden SOS preview and then we do these things and and that’s where I was just like you and and that’s where I was just like I I think we need to commit it know I I think we need to commit it right we we need to get it out the door or we need to kill it because if this this is the mentality that is been pushed out there now and I think is the responsibility of Microsoft to go back

24:36 responsibility of Microsoft to go back and clean some of this stuff up so that people don’t find themselves in situations where oh my gosh we’ve implemented a feature that’s extremely valuable for us as a business and then you you’re you’re going to be you let it go for years and then go yep nope we’re not going to support that because you’re all you’re doing is like you’re You’re Building bigger and bigger and bigger Solutions on top of preview things right and then it’s going to be disruptive so so to my point here and and not to heart but I’m

25:06 point here and and not to heart but I’m saying like you’re not going to preemptively just bring this up a as like the core part of your Solutions and say you you have to get me fabric because I built all this stuff on preview right like you’re you’re going to say that only when Microsoft says we’re killing it right because then it’s not really a solution then it’s like hey we’ve got this fantastic solution I’ve been been using right it’s saved me tons of time it’s been it’s allowed me to put out

25:36 it’s been it’s allowed me to put out reports X faster than I I did before they’re discontinuing the solution that we did which I was also doing cost savings because it fell under the the licensing we had we didn’t need to grow etc etc but now we don’t have an option now we have to go the fabric route and and I have a lot of rework to do to onto this new platform I don’t think that’s something you preemptively say like hey I wouldn’t build previous two major things two major things I think maybe I’ve been

26:06 major things I think maybe I’ve been having C here since but I am fired up about this so Tommy you would like data Mars I could see this Tommy no it it’s the confines of which a person’s working in and because I’ve been in the in these shoes where solution needs to be done the CTO is not the one who knows everything about powerbi and we care about results we care about results and yeah you have to work within the confines that you have you can so you’re going to use the preview feature because guess what it works and it it’s obviously it’s been working for the

26:36 obviously it’s been working for the situation and I guarantee you others and if you go in that conversation okay we need fabric go yeah how long is that going to take what migration is that oh by the way you have 20 tickets behind where are those resources coming from because if you’re going to make that type of migration and again I don’t want to blame anyone for saying you started using the preview future you dug your own grave that’s not the case because it is it literally says on every single preview feature it literally says on every single preview feature be careful this may or may not live on all

27:07 careful this may or may not live on all the documentation anything guys I’m saying if you’re going to make the migration I know I’m with you I get it people are in positions where they’re like we need a solution I don’t care how you do it oh by the way you have no time to do it and that’s why data flers can be so I almost use it screwed up that’s why there are some solutions that can be con polluted and don’t have great ground because people are already flying by the seat of their pants as it is in our industry in what we do so the citizen developer tools with data marks with

27:38 developer tools with data marks with data flows gen one yeah it’ be great if

27:41 data flows gen one yeah it’ be great if we had lak houses and warehouses and ponds and rivers and whatever the you want your data to flow through but that ain’t the case for a lot of people from resources from skill and from money and again again go in that room tell them we need fabric because we’re in a preview feature they’re going to say okay how long is it going to take and especially if you’re already running a ton of things on it they’re either going to look at you go can we get Tableau or we need to bring in a consultant in or whatever the case is but they’re not GNA go yeah spend next

28:11 but they’re not GNA go yeah spend next no one’s ever gonna pick Tableau over P just I’ll say that right now someone’s going to say it in Happ what they’re talking it’s a dead conversation it’s a dead that one that one now other tools maybe T it’s not that’s not the I hear I hear of no one saying wow we’re going to rip out our power go if you’re going to bring up we need to move away from a preview feature this is something with your boss before you go in a room say this is something we need to plan out and we need time for because working on this migration means you’re not working on other things and this is a huge issue

28:42 on other things and this is a huge issue with a lot of companies couple couple reaction points here like I’m not disagreeing with you guys I think you guys are right about like sometimes we just made the decision I think based on our prior knowledge we we were under the assumption that most things that went into preview would eventually turn into real solutions it would go ga it wouldn’t be a problem I’m also thinking about this too as well from the Microsoft side what is Microsoft’s motive here around all of this data March thing right the whole thing was keep people inside powerbi longer the reason that’s under premium per user

29:12 reason that’s under premium per user because you’re paying more money for whatever the compute you’re using is to run that thing so if they’ve learned so maybe at some point in here they’re going to take things that they’ve learned from fabric and they’ll maybe pull something down and maybe give you pull something down and maybe give data Marts A Min upgrade more than know data Marts A Min upgrade more than likely what I think will happen again this is just me reading based on the the past experiences of Microsoft again this is me this is Michael looking at what Microsoft has done historically as opposed to what they may do in the future I don’t know but looking

29:42 future I don’t know but looking historically it feels like they would say hey we’re going to start deprecating data Marts we’re going to give you some tool to migrate you or help you migrate most of what you’re doing so the only way they can ever announce that they’re going to go out of G so guess the reason what I’m saying this is listen for this this is the point you need to listen for listen for when Microsoft says Hey we’ve given you a tool to migrate anytime they say that that means there’s some plan of them deprecating something I think or there’s

30:12 deprecating something I think or there’s a better way they they have figured out and they want to get you off of the old thing and into the new thing so when I hear things like that I’m even hearing some language around how to how to upgrade your P upgrade your patterns from data flows gen one to data gen data flows gen two at the last fabric conference I was in some of those sessions where I was helping teach like hey here’s a pattern you used to do in data flows gen one here’s the upgraded pattern you’d want to use in data flows gen two they’re not deprecating gen flow gen one in any way they’re just saying

30:42 gen one in any way they’re just saying we want you to start using something else and I think yeah depending on how things have turned on and who’s turned on what items in whatever if you try to turn off data Mars right now I think there’s going to be a huge lashback from the community there’s going to be companies that say look we use this you can’t turn it off but there is is I don’t see right now any path where they’re making more improvements to it to make it more in alignment to a lightweight thing of fabric it’s not getting any more love it just seems to be like okay this is a it’s good enough again to our point when we talked about kill it or complete it just just

31:14 kill it or complete it just just complete it just say it’s done and just get it through GA and just say we’re not going to add any more features to it this is it this is what you get yeah but I think that’s the completed part of that conversation while while we see that there are potentially better solutions that fabric provides now yes like it Andrew makes a really good point right just because there are better Solutions right on the other side of a toggle doesn’t mean that

31:45 of a toggle doesn’t mean that data data Mars doesn’t need a decision point of it could go ga at which point it’s serving the purpose of what what it what business needed to do for that PPU license yes that that doesn’t mean that there aren’t better Solutions yeah right and it also doesn’t mean that there’s there’s not value but I do agree with your point that the in the additional investment in that area probably won’t be as accelerated as

32:15 area probably won’t be as accelerated as what you’ll find in fabric yes they’re giving us a year and a half to get off powerbi premium wasn’t that 2026 is standing deprecated so if they’re not there I don’t think I’m getting a Blog today saying data marks will be unavailable after the thanks you abbrevi a lot of that comment they’re giving us a year to get off a premium or premium tier the P skes the P you to migrate away from PC to fcq isn’t it getting deprecated in 2026 wasn’t that the date no 2020 well deprecate no 2025 is in mid 25 is when

32:48 deprecate no 2025 is in mid 25 is when you re if you renew if you renew it would be your last so so deprecated meaning no other licenses provided no other and buy it again yeah PC would like on their end be able to be shut down yeah 26 and that’s when the okay so still it’s still six months of time available before but think about it no longer available but think about it P to fcq everything you can do in the pcq is available in the but that’s my point though my point is Andrew’s like I know

33:18 though my point is Andrew’s like I know I agree I know I get it understand but like my my my pre my yeah pre premium per user that’s what’s missing here but like the the P skew to f story is like Microsoft is like look we’re not going to give it to you by the way you’re going to get that everything you could do and more with the fabric skew and by the way if you’re already using premium here’s all the fabric things anyway so that’s a very easy migration pattern U it definitely does take some time you’ll need to take things down for a weekend turn on a skew and flip it over but like okay fine there probably is some testing organizations need to do to turn over their workspaces

33:49 need to do to turn over their workspaces from the P SK to the F SKS I’ll be it yeah fine it is what it is anyone listening disagrees with you on wanting to do that but the equ of that that lives in data marks is not the same so that’s where my point here is like if I wanted to if Andrew said look Andrew let’s do this let’s move you over and maybe answer again this question came out a little bit before I I’m assuming this question came out before they gave the price increase I would I would Reas the question back to Andrew and say hey Andrew send us another send us another comment in the in the podcast does the

34:20 comment in the in the podcast does the price increase of pro users to $14 or the premium price per user going from $20 to $24 does that change your mind at all is there a breakpoint again I think there’s a a situation here where if you’re on a small enough organization where you’re just getting things to happen yeah it doesn’t make sense for you to go up to a fabric skew but I’m thinking now maybe there’s potentially a need for uproll rolling up to this one and I think to me one of the major break points of whether or not I can go from where I’m at in premium per user to

34:50 where I’m at in premium per user to fabric is what is the size of my semantic models because that’s to me that’s the one not that I can’t get with a lower fabric skew unless I’m paying a lot of money for fabric let me just let me just unpack this a little bit here for you I’m thinking here the the premium prod user gives me 100 gigabyte size memory models which you can’t really shove 100 gigs in there anyway so if your model goes above one gig you’re out of pro you’re into premium per user the top end of Premium per user is 100 gigs it’s a

35:20 of Premium per user is 100 gigs it’s a huge gap to get 100 gigs in a model you’re looking at a f256 the same size machine that you’re going to get to get a premium per user model that’s 100 gigs so think about that like that is the size model like that’s the size machine that’s an expensive machine to run per month to get up to a 100 Gig size model so if you’re in that lower end tier of like my model is five gigs so I’m seeing a lot of things where companies are jumping from one gig to

35:50 companies are jumping from one gig to like 1. 5 gigs or they’re getting their models get a little bit larger and they’re like hey we either need to tune our models take data out of them so they continue to work or we go up to a bigger premium per user skew and that potentially is a big jumping cost for them because they’re going from a handful of users that are already premium per user and they’re going up to premium per user so to me there’s a pretty big step between those and that’s my main consideration for where I would say yeah you don’t really want fabric because fabric will give you a little bit larger data sets five six 10 15 25 like there’s going to be other

36:21 10 15 25 like there’s going to be other steps there that are going to give you more gigs in those models but that’s maybe my point and so like if you’re using Mars and you’ve got really big models that you’re doing things with yes you’re not going you’re not going to be incentivized to move to fabric for that reason because you won’t get the same performance let me pose a question to you guys and let’s take this situation that Andrew’s in and let’s say Fabric’s not the option because I we know Fabric’s always the answer and like I said I recommend it but it is the way okay it is the way it is gonna be the way so we all know that but for this

36:53 way so we all know that but for this situation if you were in Andrew Shu if

36:56 situation if you were in Andrew Shu if you were in the situation with the team that you’re already utilizing data marks are you trying to move off it right now or are you keeping business as usual are you trying to push for something else a database or some other service or something that maybe is still not cost intensive or not a migration but no back yeah yeah no I I think the way he’s presenting it is in such a way that he’s saying he’s leveraging this

37:27 that he’s saying he’s leveraging this and finding success in it and has Solutions built on it and no I don’t I don’t think he’s proactively looking to switch up infrastructure and nor should nor should he would I yeah would I would I Pro would I proactively blow up my world for infrastructure and slow down the development of what seems to be a fast Dev org right like where you you’ve been part of that Tommy the expectations of you turning around reports really fast how how much of that would involve

37:59 fast how how much of that would involve like yeah let me bake in another day so I can like redo my infrastructure and I think the other part of this is there are no other options for him at the moment that he sees now I want to push into that a little bit after I like we we get through this question but I think no you’re not just picking up an an infrastructure or change kind infrastructure or change thing when you’re you built this of thing when you’re you built this solution and it’s working really well like your goal is to bring insights to the business it’s not not just to play

38:29 the business it’s not not just to play around with the infrastructure Mike same question to you I I’m trying to unpack what what we’re doing here on like the the size of things so things so I’m I feel like what needs I feel like what’s happening is Andrew is in a place where someone else is making some of those larger harbi Administration conversations and it’s less around a a center of excellence that’s actually making these decisions I’ve seen this before where there’s the decision point if Andrew had the decision right if Andrew was making the calls he would

38:59 Andrew was making the calls he would have said let’s just I feel like the reason we got the data march to become the first time was because Andrew said I’m going to be look at this cool new thing this is in preview let’s check this out oh wow look I am making the and this is what we talked about in our other podcast around leaders High performing individuals like he came in he said look we have a problem here’s a data Mart this solves our problem look what I built I built this thing this thing is now adding value okay great we’re going to invest more in data Mars and build more on it great everyone’s happy stuff’s working he get things done quick it’s sequel great we’re all happy

39:29 quick it’s sequel great we’re all happy in the same way he would have done the same thing for fabric but he can’t this is and this is where I’m like the the point here is he want he would he would be that individual to go ahead and do this to pull things to the next level for his organization but because the decision doesn’t live in his hands it lives at a higher level in the organization that decision point is being pulled away from his hands and it’s not allowing him to like step in and do the Innovation that he needs to do to keep the solu current with what stuff that’s not in preview so I don’t

40:00 stuff that’s not in preview so I don’t know how you handle this conversation but it feels to me like there’s a a resistance level at the top level of the leadership of the company that is causing some push back which is inh inhibiting him from exploring the next thing or I I would I wouldn’t go I wouldn’t go that far I I would say or or there just hasn’t been the engagement across the teams or he he’s in a silo of Building Solutions and not integrated with that team for whatever reason let me let me

40:30 team for whatever reason let me let me challenge I’ll say this is a great opportunity because data marks came out well before fabric right or was it the same year it was well no it was the year before it was the year data Mark came out a year before right so I guarantee you what was happening was Data Mars got built it was getting started and Microsoft made a pivot at some point so some somewhere in here somewhere in this process Microsoft said hey we’re going to bring we’re going to bring more SQL like features to to powerbi with data Marts and then someone at the leadership

41:00 Marts and then someone at the leadership said L said no wait a minute the real place here is how we’re getting eat people are eating our lunch in like Delta lake lake houses and so someone at the leadership level said wait a minute we need to also pivot and say look the verta PAC engine and Delta tables actually can play really well together so what we’re going to do is we’re going to respin this and say we we now need to have a bigger vision of what we’re talking about for data platform and now we need to bring this new format to to me that’s the that’s the change that happened because yes whether you’re not your data

41:30 because yes whether you’re not your data warehouse we know what Microsoft saying I’m saying for the situation where people are building data Mars that’s the sand that shifted underneath these people’s feets yeah but you’re saying that but I don’t agree with it right BEC the sand is they’re bringing interprise tooling into fabric we talked about how do you bring a citizen developer into an ecosystem that they don’t understand and you’re still you’re pretending that that that transition is just super seamless and it’s I I I don’t know I a little bit of push back in that regard understand

42:02 of push back in that regard understand your P not worth it that it’s not worth it but data Marts is a very straightforward way of like Hey we’re going to add this additional thing for you and get a lot out of it do that’s a big point because honestly data when data flows gen one came out in 17 that was my first introduction to the data engineering world and one of the first introduction to like how do things actually work when it comes to Shifting data through a process because that’s the first time that Medallion

42:32 that’s the first time that Medallion came up to me because it was like okay we’re not just dealing with this data set okay what’s this world now that they’re talking about about loading data in and then transforming it so for a lot of people the tools also introduce to them some of the skills but here’s what I’m saying this is an opportunity if I the same question I guess asked you would you change gears or would you st as as the ship sailing well if data March were being built preab and everything’s working I may be happy but I’m probably waking up every other

43:02 but I’m probably waking up every other week going data Merchant preview because there with all of the growth of fabric at this point to me this is an opportunity where I would challenge that a lot of the features that they’re making in fabric are Citizen developer a lak house is so friendly to to create and you you also can use data flows where there need to be the slow conversation because it’s not going to happen overnight it’s not going to happen over the next two months where you’re going to just shift everything over but I don’t know if I if

43:33 everything over but I don’t know if I if I’m in this situation where my companies running on these data Marts and I’m seeing I haven’t seen an update for data Mars and that’s where I would get nervous I haven’t seen a Blog article a Google search anything by Microsoft with the words data Mart in two years let’s say probably I’m getting nervous I’m starting to I’m getting some beads of sweat at this point and it be very hard to continue the gears without saying Hey to leadership or the bosses they might not be pushing back and that’s where I

44:05 not be pushing back and that’s where I I’ll argue I don’t think they’re pushed back we don’t want fabric they might be aware of the value they not be aware you aware of the value they not be aware they know there’s a cost involved know they know there’s a cost involved sure but this is a great opportunity go look we’re playing in this in a playground that is going to close down we don’t know when but this is not Microsoft’s path so what are we going to do so so these are really good point because I think this is this this would morph into recommendations that I think we would have for Andrew right I think so we we

44:36 have for Andrew right I think so we we don’t talk about all of the things in Fabric or easy replacements for data Marts be just to talk about them right they’re much more extensible solutions that’s where all the time in development is gone and they’re Enterprise tools yes agree with that business as it stands especially when you’re in the burner which is like I don’t know if you have development processes right or is it just the you processes right or is it just the con Bond board that never ends know con Bond board that never ends where business is just throwing report

45:07 where business is just throwing report requests your way and that’s the the life he leads and like that’s a rough life it’s rough they don’t care they care about results they don’t care about infrastructure they don’t care about how you built stuff what they would notice is if you deviated from that to go fix a bunch of things for for infrastructure right so he I I think Andrew needs to understand the Nuance of preview that we pointed out right the the lack of Engagement or feature updates and data Mars and what does that speak to his

45:40 Mars and what does that speak to his Solutions and I think the nudge here from my perspective would be I think now is the time that you start to engage the the other technical team or the other leader and present a solution for a potential better future or one that you potential better future or one that is more sustainable because if it’s know is more sustainable because if it’s still in preview if there’s been no features and all this other stuff is happening that we’re talking about in

46:08 happening that we’re talking about in fabric there’s a compelling reason why you would want to go that in that direction because it’s scales right so what are what are the points that win that that argument so I talk I I know exactly where you’re going with this one I’m going to just give you maybe one point that I would look for in if I was in Andrew shoes I would one try to educate myself either reading documents or getting into things as much as I can about what fabric is capable of and potentially even some of the differences around what you’re doing today to what fabric is doing the reason I’m doing that is because I’m listening for when

46:40 that is because I’m listening for when the business comes to me with a new request hey we have this API it’d be really helpful if we could add that data to my thing or hey we’ve got this new non-sql server it’s a thirdparty service that lives in the cloud somewhere how would we best connect to it and full well yeah I could probably do that in data flows gen one yeah I could probably land the data in our data mart but you in my opinion here is I would try to know enough to say okay we could do this with data Marts but what you’re really asking for is we need to land

47:11 really asking for is we need to land these small files somewhere or we need to use these other API calls to go get some data and it’s actually really difficult to make an API call with data flows gen one we have to go get a token it doesn’t really work very well with it like it’s just got a lot of friction it’d be really nice organization if you want this new data please let me B spin up a single Resource Group or a single workpace that would let me experiment with getting that data in using a more fabric based solution so I guess all I’m trying to

47:42 solution so I guess all I’m trying to say is that was that was more of the example maybe I should have stated my my my statement here first I’m looking for use cases where I think fabric would fit better than what I have in data flows and I think if I’m looking at it I’m saying things that are more important for me to to bring them to fabric are API calls things that are talking to third party Data Systems those are good opportunities to build those Solutions inside Fabric and they’re easier to implement than doing what you’ve been doing in data flows gen one and in data

48:13 doing in data flows gen one and in data Mars my opinion you that’s where I would think you would land so listing for business use cases where fabric fits better than what you currently have then when those then when the need right when when the business comes in and says we really need this this is really got to be here we’ve just purchased this amazing software or whatever the thing is and we can’t get data out I’m like great I now need you to untie my hands a little bit it just at least trust me I got you to data Marts trust me that I’ll be able to get you through fabric trust me as the B as my team we’ll figure it out let us work with you or we’re going

48:44 out let us work with you or we’re going to go find someone or whatever the whatever the situation is but do something that helps the user get them push them past that barrier of to me the big blocker is what can I do how do I convince leadership to give me just a little bit of fabric so I can prove the value that’s all I’m going to say and my my closing thought here I’ll really simplify this I think I have the the easiest answer here I’m I’m writing a letter a Christmas card to SAA and say you’re not really supporting data March in the future right if not please send a

49:14 in the future right if not please send a Christmas card to my company and I’m gonna ask I’m gonna ask Saia to write a card and it says season’s greeting your current solution will be deprecated in the next two years that that’s what I’m going to do because I I well and the point I’m trying to make here is really that we know it’s in preview we know where Microsoft’s focusing their time to your point too like I’m starting that conversation like just so we’re our Solutions working great now but but we haven’t seen an update in two years and that’s usually doesn’t like

49:45 and that’s usually doesn’t like based on we know with research that doesn’t bode well for products with Microsoft if we don’t see updates I don’t think I don’t think that’s a I understand what you’re saying I don’t think that’s a compelling enough argument for leadership to make the decision to open up fabric and I would AR I would argue neither as yours because you’re saying that that it’s it’s difficult to do something but it’s not impossible in the current Solutions so sure leading with a technical solution that benefits Andrew in terms of implementation is rarely

50:15 in terms of implementation is rarely going to move the needle or or say yeah we should spend spend some more cash in this direction or we should pivot on infrastructure I I personally would think that while while the developers are technically thinking in those those directions all the time I want to be more efficient I want to hopefully do better things for the business Etc yeah I think the play is like I said engaging with the team or leaders where your your pitch is on

50:47 or leaders where your your pitch is on improving the data solutions for the business and how do you do that right like hey okay here’s what here’s what we’re building in in the fabric ecosystem right I don’t like don’t bring them along for the ride what what can we do there are these things called domains right I’m in a domain I’m in an area of the business there’s this data mesh fabric fabric is part of data mesh the principles behind there are we as the subject matter experts or the builders in our area of the business can

51:19 builders in our area of the business can build build solutions that make our data work more useful for The Wider organization so what I want to do or what I’m proposing being Andrew is I want to build solutions that are accessible to the bi infrastructure or other teams and we have that capability if we invest in this direction so I’ve got all of this stuff that works really really well for the business I’m able to produce a lot of value and turn reports around not suggesting that I don’t do that but from

51:50 suggesting that I don’t do that but from an R& D perspective I think this is the direction that we start to head because all these capabilities live here and the net net of that approach and proposal is you’re saying I’m actually going to take more load off of that be that Central bi team and the organization and I’m going to provide like be the owner of all of this data related to this area of the business so it’s it’s a quid proquo but you’re also

52:22 a quid proquo but you’re also setting yourself apart in in in that not only you’re just building good Solutions and extending and developing reports you’re also opening the door to get the access you need to start playing with these bigger things and long term right Ben taking advantage of all the stuff we talk about all the time and why Fabric’s so cool right but the pitch I think and where businesses go well hot dang like Andrew is presenting something that with minimal cash investment he can prove out

52:53 minimal cash investment he can prove out or we can start to work with the bi teams and now all the data and work that he’s been doing is is going to be in this place where we can start to you this place where we can start to certify or like govern on a on a know certify or like govern on a on a higher level I I wouldn’t if somebody from the business pitched me that idea I’d be like hell yeah like you’re gonna go 100% now you’re part of my conversation line like you do all this work for me and then just give me access

53:23 work for me and then just give me access to the data you’re curating for stuff like dude that’s win-win how much do you need per month what I’m saying and that’s where I that’s where I think I would recommend a lean in because to Tommy’s point right like we don’t know if it goes GA you got no problems right if it stays in preview and it’s not getting updated how how how long do you have on the legs of that no reason why you can’t get out in front of things and start to leverage the stuff that that

53:54 start to leverage the stuff that that would extend and provide a better solution outside of what data Mars does all right so Andrew you need to try all three approaches and let us know which one works the best so I think we have three solid approaches here I would agree with that one we’ll we’ll we’ll figure things out slowly as we like who’s going to win what what’s what’s we want a question in about six months back from Andrew about what the answer is what did you you try and do you shot down shot down Tommy was an idiot yeah by the time I got to

54:25 was an idiot yeah by the time I got to the Third was actually talking to me awesome well with that thank you very much for your time we appreciate your ears on this conversation I thought this was a very good conversation very relevant hopefully this helps you think through your things around data Mars and other elements inside your fabric environment and maybe conversations you should be having with organization to turn on or or try to get access to other features that are not available to you yet that’s a thing to think about Tommy where else can you find the podcast you can find us in apple Spotify wherever you at your podcast make sure to subscribe and leave a rating it helps us out a ton share

54:56 a rating it helps us out a ton share with a friend or a colleague because we do this for free for you guys you have a question idea or a topic that you want us to talk about in a future episode head over to power. tips podcast leave your name and a great question and finally join us live every Tuesday and Thursday a. m. Central en join the conversation on all of powerb tips social media channels thank you all very much and we’ll see you next time [Music]

Thank You

Want to catch us live? Join every Tuesday and Thursday at 7:30 AM Central on YouTube and LinkedIn.

Got a question? Head to powerbi.tips/empodcast and submit your topic ideas.

Listen on Spotify, Apple Podcasts, or wherever you get your podcasts.