Hiring the Report Developer – Ep. 503

Mike and Tommy unpack what a report developer should actually know in 2026. With Fabric reshaping the data platform and SSRS migrations in full swing, the role is evolving — and paginated reports might be the most underrated skill in the hiring conversation.

News & Announcements

• Enrich Power BI Reports with Machine Learning in Microsoft Fabric — Microsoft walks through an end-to-end pattern for adding churn prediction to a Power BI report using Fabric’s semantic link, notebooks, and batch/real-time scoring. Mike questions the architecture diagram’s approach of pulling data through the semantic model via semantic link rather than going directly to the lakehouse silver layer — arguing that data scientists typically want raw features that may not exist in a sanitized semantic model.

• SQL Database in Fabric: Built for SaaS, Ready for AI — SQL database in Fabric has seen rapid adoption since going GA in November, offering a fully managed transactional database with automatic mirroring to OneLake as Delta tables. It now supports native vector data types for AI/RAG scenarios. Mike shares that he’s already building applications on top of SQL database in Fabric and finding the transactional row-level performance far superior to lakehouse for rapid individual writes.

Main Discussion: Hiring the Report Developer in 2026

The core question: if you’re a hiring manager looking for a report developer today, what skills should be on your checklist? Mike and Tommy debate the evolving role, the surprising relevance of paginated reports, and where the line sits between report building and data modeling.

The Case for Paginated Reports

Mike argues that paginated reports are an underrated skill for report developers. While less than 10% of job postings mention “paginated reports,” the SSRS-to-Power BI migration trend tells a different story — organizations aren’t abandoning paged reporting, they’re transitioning it. Tommy confirms via a live Perplexity search that demand for paginated report skills is steadily growing, driven by ongoing SSRS migrations.

The key insight: when business users ask for really wide tables with lots of columns, that’s a decision point. Forcing that into an interactive Power BI report leads to poor performance and a slow, laggy experience. Paginated reports handle unlimited row exports with required parameter filtering — giving users the data they need without beating up the semantic model.

Two Levels of Report Developer

Mike breaks the role into two tiers:

Level 1 — Entry-Level:

• Build reports in the Power BI service (powerbi.com)
• Understand visuals, bookmarks, slicers, and filter pane design
• Build basic paginated reports in the Fabric service
• Know when to use interactive reports vs. export-focused paginated reports

Level 2 — Advanced:

• Power BI Desktop mastery (still the gold standard for full-featured reporting)
• Paginated Report Builder for complex export scenarios
• Embed paginated reports inside Power BI reports with parameter passing
• Manage the full SSRS-to-Fabric migration workflow

Report Developer vs. Data Modeler

Mike draws a clear line: report developers and data modelers are different roles (even if the same person fills both). If you’re writing user-defined functions, building custom measures with complex business logic, or doing heavy Power Query transformations — that’s data modeling, not report development.

Report developers should handle:

• Report UX, layout, and interactivity
• Understanding what data elements belong on the page vs. the filter pane
• Choosing between Power BI reports and paginated reports
• Working directly with business users to understand their needs
• Building the “data application” — drill-throughs, navigation menus, and the full interactive experience

Measures in the thin report layer? Only if they’re stylistic (formatting, conditional colors). Business logic belongs in the semantic model.

The Trust Factor and Exporting Data

Why do users still demand data exports? Mike argues it comes down to trust. Users need to export data, match it against their operational systems, and verify the numbers before they’ll trust the reports. Once that mental model clicks — “this column in the semantic model matches this field in my source system” — they start relying on the interactive reports more and export less.

Paginated reports serve this trust-building phase well: required parameters force users to filter first, reducing load on the semantic model while giving them the exact data they need to verify.

Looking Forward: What to Hire For

Tommy pushes the conversation toward the future: don’t just hire for today’s needs, hire for where the role is headed. His advice — get acquainted with Fabric and the lakehouse. The report developer who understands the broader data platform can solve problems upstream, not just at the report layer.

Mike’s top three skills for a report developer hire in 2026:

1. Power BI Desktop — still the most capable tool for building reports
2. Paginated Reports — for export-heavy, wide-table, parameter-driven scenarios
3. Fabric awareness — understanding how SQL database, lakehouse, and mirroring fit together

The bottom line: the report developer role is no longer just “make dashboards.” It’s about picking the right reporting tool for the job, bridging SSRS legacy with modern Fabric, and being the liaison between business needs and the data platform.

Episode Transcript

Full verbatim transcript — click any timestamp to jump to that moment:

0:01 Lighting up the [music] sky. Dance to the day to laugh in the mix. Fabric and A. I get your fix. Explicit measures. Drop [music] the beat now. H feel the crowd. [music] Explicit measures. [music] [snorts] Good morning. Welcome back to the Explicit Measures podcast with Tommy and Mike. Hello everyone and welcome back to the show. I don’t think I’m ever gonna get over that. The song. The song. Yeah. I By the way, the expensive song, I think that’s your hit.

0:36 It’s catchy, right? So, last week, so for those of you who didn’t catch up on this last week on Thursday, about halfway through the episode, you have you’ll have to dig through halfway through, we released another song. I’ve been cranking out songs left and right using AI and AI generated things, which has been super fun. I’m I’m loving it. Anytime someone says something funny in my family now, it’s like, “Oh, that could be a song. Oh, that could be a song. So, yesterday we were done eating dinner. My son did not want to go to the bank, but everyone else wanted to go to the bank and I’m like, “Ooh, I don’t want to go to the bank. That sounds like a song.” So, we’re making songs up in the car as

1:08 We’re going over to the bank that no one wants to go to the bank and we have to go to the bank to get the money for the bank. It was just funny. Anyways, I’m having a blast with it. Super fun. Last week, in honor of getting over to FabCon Vienna or sorry, Fabcon Atlanta, wrong event. Fab Atlanta, we threw out this fun song called Expense It. And so this is my, the boss don’t care. You can expense whatever you want. They’ll approve it. , stack up your TPS reports or whatever and it’ll sign your receipts and you’ll get it approved. So, fun little thing last

1:40 Week. It was cool. Tommy, you tell me that you can’t even go anywhere now. Like, you have to bribe the children with Tommy’s song, right? Dad’s song. Oh, no. It’s It’s terrible. all the kids every time we get in the car is we have to play the Tommy Puly song and [laughter] no matter what I got to play it. So that’s amazing. I love that. Well, it’s great that your kids think so highly of you since dad’s got his own song, his own theme song now and it’s I know. I know. Awesome. [snorts] Well, that being said, let’s talk about some news items. But before we get to news, let’s quickly

2:12 Talk about what’s coming up in the episode today. Okay. So today I just want to unpack some some con concept around what are the report developers. Tommy things have changed so much. We’ve got fabric, we’ve got semantic models, we’ve got data engineering, we’ve got all these different roles of people they’re doing different things inside PowerBI and fabric. Have we lost touch with the report developer? What should they do? Is there any would we if we just step back and let let’s ignore fabric for a moment. Let’s ignore all the agentic space things. There’ll probably be some conversation here

2:45 Around that, but what would we expect a report developer to know? What skills should they have? Has this changed in any way? Have we added any more skills to their knowledge set that they need to have in order for them to be effective in this new world of AI, fabric, PowerBI, data space? What does this look like? Okay. Do you want my quick answer or my long answer here? Let’s do that after we do the news and then we’ll come back into it. How about that? So, sounds good. No secrets given early here. You’re gonna have to wait around for for the end of the news. All right,

3:17 Let’s jump into the news items. Tommy, you got two topics here. Yeah, Mike, we we got the first one is on the fabric updates blog. Enrich PowerBI reports with machine learning in PowerBI fabric. And really here looks like this ties into our topic today just about a a little a little a little. So again, nothing here is completely like it’s not a new feature. So we’re not dealing with new features here. Sure. It’s just one of the it’s more of the call it the showcase part and all we’re simply doing here is Microsoft

3:50 Going through exploring semantic model semantic link creating that through a data frame doing data exploration. So your normal data sciency things. Yes. And then feature engineering to get the attributes and basically what you can do with the machine learning add-ons that are traditionally like we’d [snorts] expect logic drift broken refreshes. So optimization there are end to end scenarios looking at predicting bank customer turn. And it’s basically being able to score customers using batch scoring and

4:23 Real-time scoring. And again it’s just all using fabric semantic model semantic link fabric machine learning experience batch scoring is scale or spark real time dataf flow gen two and they just basically go through step by step. So this is interesting. I I want to point out the architecture diagram they pulled out in putting in this article. I think this is interesting Tommy and I’m not sure if you’re thinking this way as well Tommy but what they have here is they have new data coming in. New data is landing with dataf flows gen two. Okay, fine. You can do that. I

4:58 Wouldn’t recommend it. I’d recommend a notebook or something else more efficient than data flows gen 2 just from a cost and CU standpoint, but okay, fine. The data eventually makes itself to the lakehouse. The lakehouse then pulls data into the semantic model for caching and servicing the reports. Great. Love it. Everything there feels just normal. This is interesting. In number two, the number two nugget that here they have in the diagram talking about semantic link directly accessing the model. I find this interesting as opposed to going directly after data tables that were in the lakehouse. So this indicates that

5:32 There is additional measures or columns or shaping of data that’s happening in the semantic model that you want to rele business logic of aggregated measures or calculations. You may want to apply that back through to this analysis. So I I see this as being like the business is adding logic or structure to that clean the data directly from the lakehouse into the semantic model. Now albeit if I’m doing lakehouse and semantic model I’m also looking at there’s a handful of tables at a what I’m calling the gold

6:05 Layer right there’s maybe bronze silver gold there’s maybe like good raw data polish data I I don’t know how you want it staging non-staging data I don’t know how you want to call it in the lakehouse but right there’s probably some tables that exist in both places and I’m not sure if I would in this architecture I’m not sure if I would go directly after the semantic model I’m not sure if there’s enough stuff in the semantic model that makes it worthwhile for me to pull directly from a semantic model with semantic I’m not sure that step is required I don’t know again I think there there’s going to be a shift Mike because again previously people dealing with machine learning

6:38 Data science they didn’t want to hit anything structured because usually didn’t have everything they need but what they’re trying to do is get raw data but we can give them one step better I don’t know if it is I would my argument would be is I think the data that comes through to the amendic model is going to be too sanitized not exactly what the business wants. So I I would almost question that part of the architecture. I’m not hearing a lot of people going directly from semantic model through semantic link to notebooks to then do your data science. Typically in the patterns that I’ve been building has been notebook

7:11 Going to data tables in the raw or silver layer in your lakehouse then producing the churn analysis and then producing gold tables that then get serviced back to the model. So I found the diagram a little bit interesting. I’ve not really seen it communicated this way before. interesting that you can throw semantic link on there. Not saying it’s a wrong design, just not how I would do it. So, anyways, just wanted to wanted to point that out there as as something that I’m looking at. I’d be curious. Let us know in the comments, what are you doing in any space around pulling data out from either semantic models or are you pulling would you rather pull it

7:44 Directly from the lakehouse? Let us know. , I could see this being much more of an effect around this requirement if you were going after like a a semantic model that lives in a pro or premium per user workspace, right? That would make more sense to me, right? The workspace doesn’t have a model that lives in , premium per user and you don’t have a lakehouse. I would argue let’s say you’ve got a semantic model that goes straight to a data source, then I think I would want to use semantic link. That would make more sense to me. pull because the data well think of it.

8:17 The way they drew the architecture is there’s a lakehouse and notebooks already involved in this design. Well, that’s your data engineering. I would rather do the data engineering in notebooks and lakehouses than to involve a semantic model and add more an extra step to everything. If I if I’m so think of it this way, Tommy, let me rephrase my phrase here. If the semantic model is doing the data engineering, if you have the semantic model pulling the data directly from the sources, shaping it, engineering it, and then producing the tables directly for reporting, there is nothing to connect

8:50 To. The only thing you have to connect to is the semantic model. So if you drew redrrew this diagram and said, let’s pull data flows and lakehouse, get rid of them, new data comes in, semantic model is where that is imported from, and then you need an ability to do some AI thing. Okay, that makes a bit more sense to me. Now the question I would also argue is then once you’re done with the data science where do you put the data because you you got to put it somewhere. You can’t put the data science from the model directly into the PowerBI semantic model report. So anyways seems it seems a bit overly I don’t know it feels like they’re pushing

9:25 A little too hard end to end and to end it feels like they’re pushing a little too hard there on the on the architecture piece. the code is solid. the whole rest of the article is just here’s the code you can use to do predictive analysis. this is not a business user [laughter] Tommy no this has got plotly diagrams in it. It’s got multiple lines of data frames. You’re building all kinds of new data frames. It there is a ton of code here that is in this article. I hope there’s a there is there a notebook Tommy where you can download

9:57 This. Oh okay exp explore in this sample repo. Okay. At the very end of the article, there is a sample repo with this in it for you. So you don’t have to go copy paste all this code out of the blog article. Anyways, what are your thoughts, Tommy? Any thoughts? No, Mike, I I this is what I’ve been saying since the beginning since we’ve had semantic models, like if semantic models are best for AI, why wouldn’t they be best for machine learning? And why wouldn’t they be best for data scientists? I know what you’re saying. We need the raw data. But 90 I remember Mike, [snorts] this 90% of what

10:31 Data scientists do is clean data. The rest is the actual analysis. So they’re cleaning data, but they want it at a at a more raw state, right? They don’t they don’t want someone else to clean it before they get it, right? They this is the idea is like they need to clean it their own way. And there’s a lot of things around like building features. Sometimes when you build features, the features that you’re using to build the machine learning don’t are are not carried through to the semantic model. So if all the features that the data scientist wants. So that’s why you go back up to a layer I think around silver which is okay I’ve got 16

11:05 Columns of features, what type of person they are, where they shop, what zip code, what store, what time of day, all these features become portions of will this person churn or not for the churn analysis. Those features may not exist in the semantic model, but I know for a fact they’ll probably exist in the in the silver layer because that’s what you do. You bring in the raw data. You just make another clean filtered table of data in the silver layer. Great. That is the I think the silver layer is the ideal place to pull data science from in my opinion. But I would argue data scientists aren’t too happy when you

11:40 Give them a semantic model and say here you go. This is what we’re using to report on. They’re just not great. Now, I will say this, Tommy, let me caveat my phrase here. When we’re talking about AI, there’s different ways to think about AI. This is data science applying to tables of data. Then there’s another layer of AI, which is help me AI, inform me of what’s inside the semantic model. Inform me what DAX has been written. How would I write a DAX formula about this? , how would I build a report that has these pages and these visuals on them? I think AI in the era of creation

12:14 Is actually really good. But and you’re right, Tommy, in that example, if the AI is helping you build new reports, 100% point the AI at a semantic model. You need the column descriptions and the relationships. That’s metadata that’s required for you to understand how the data works. I don’t necessarily with the amount of red, code that’s shown here on this blog, I’m pretty sure they don’t need relationships between tables. I’m pretty sure they’re building whatever the heck they want to build out any churn analysis on their own without any additional support from a semantic model would be my idea.

12:49 Anyways, that’s just that’s just my impression. All right. Well, we got one more article, my friends. Either way, it’s good. they how they like did their showcases. It’s a good showcase. I haven’t seen a lot of feature news come out lately, but there’s another one. SQL database and fabric. Speaking of which, , built for SSAS, ready for AI. And what we’re doing here is there’s been rapid adop adoption with the SQL database, , and fabric since it’s become generally available in November. , and is again a a showcase of

13:25 Other organizations, just getting started with it. But it’s using the CN SQL databases Azure database. So any TSQL skills work provisioning is incredibly simple. Some may say too simple [snorts] because there’s no other settings there. But every time you write to SQL database is mirrored to one lake and in a delta table and could be available in Spark notebooks, PowerBI and even machine learning workflows. There’s role level security, custom

13:57 Integer keys, private endpoints, and their big thing is ready for AI because there’s native vector data types, which I thought was very interesting because usually, and Mike, correct me if I’m wrong, an AI database or usually a vector database is not your normal SQL. It’s not your father’s SQL database. It’s quite Yeah, vector databases are really more for agentic experiences, right? That’s that’s the reason why you put vector databases in. It helps you find common terms, nearest neighbors, things like that. So, it’s definitely an

14:31 AI ready type system that’s really designed to be short, short cycle you to getting to use AI on top of things. That’s that’s how I understand vector databases. And you can Yeah, exactly. So, I’m surprised that they just said, “Yeah, there’s a data type for vector because I’m like, okay.” I would be intrigued because they are right now I always like whenever I’m not sure I go to the community users are you encouraged to try the SQL database SQL con collocated with Fabcon Atlanta is probably going to

15:04 Have a lot of updates on that as well well I think this is going to be an interesting conference Tommy this is the first time we’ve ever had SQL and fabric all together in one place there’s been various forms of SQL conferences around but I think now with SQL coming into the fabric ecosystem system. We need to have another conference that’s really discussing like all of the data platform pieces. Everything is trying to land inside Fabric and make it easy for people and users to build on top of this. I’ll be honest, Tommy, I’m be very impressed with the SQL database inside Fabric. I’m already building SQL databases in Fabric

15:37 Today and I’m building applications that live on top of those databases. It makes it really easy when you build a full database on top of SQL already built into Fabric. you get immediate mirroring to a lakehouse. So if you want to have mirroring from that SQL database into lakehouse that you can just turn that on and you get a lakehouse with copies of data for reporting. So if you want to do reporting or analytics on top of things that already exists and then the SQL endpoint is your transactional database. You can go in build your application directly on top of it and it just works fairly seamlessly. I haven’t

16:09 Had any issues, no hiccups. I’m on a trial capacity. I’m running my SQL database. [snorts] Always got to love running your applications on a trial capacity. Well, I’m I’m testing it out right now. This is like me testing. I’m not putting any production customer stuff on it yet. So, I’m getting a a feel for what is the usage, like what level of fabric skew do I need to have in order for this level to to show up. Anyways, so yeah, Mike, so and this is what we talked about initially when the SQL database came out. Let me ask you this. Is there a role for

16:42 The lakehouse when it comes to your applications at all? or really it’s just we’re going to prefer the SQL database. That’s a great question, Tommy. I I think it’s a combination, right? You have to think about what the data is doing, right? If you’re going to be doing lots of individual row level changes, the SQL data the the way the lakehouse, sorry, is not as efficient. It’s slower. It just takes a bit more time. You have to read the partition files under the hood. you think about it. The SE the wear the lakehouse is a columner store of data. it’s it’s taken all the column data. SQL databases

17:16 Are row level. It’s trying to store the information at the individual row of information. If I want to do an insert or update or merge, I want to go grab the row of data and update that single row of information. That’s what SQL databases are really good at. I found when we were doing lots of very rapid writes and reads or regular read and writes to the lakehouse with very small individual row type of data wasn’t as efficient. wasn’t wasn’t as fast as it was if I just went to like a database. So, for me, there’s like this tipping point for how fast are you going to be

17:50 Doing the data loading? Are you loading large chunks of data like an entire day, a whole year, like what is the chunk of data you’re trying to bring to the lakehouse? And then that helps you determine whether or not you want to stick with the lakehouse architecture or you want to add some transactional-like system in the mix there. So I would argue when you’re doing more transactional individual rowle details, it’s better to get the SQL database to get that high transaction rate, but then you still have the mirroring that then compacts the data down and you get your columner store data in the lakehouse

18:23 With almost near real-time updates from as soon as you update something in the SQL database, the the lakehouse will get the mirroring effect very fairly quickly. So that’s maybe where I’m breaking the line here. I’m exploring more. I think fab SQ SQL has changed how I view fabric a little bit. What are your thinking Tommy? Is that your impression as well? I wanted to because I’m like well right now I’m looking at the lakehouse and you’re always looking at the purpose like what are all the purposes for a given product and right now when I look at the lakehouse

18:55 It is that central connection between my organization’s data in terms of being able is the collection of data and then not so much as the pushing out of data where as very much with SQL back in the day it was obviously like we need to get our data in a SQL database to analyze it. But right now, at least in the fabric architecture, the way we have it, if I am putting data in SQL databases because I have something I want to either write to it, I need it to talk to something or there’s some like some integration with another application or product.

19:30 Whereas lakehouse has become the de facto SQL database. I wouldn’t call it de facto. it has like a SQL like endpoint. you can talk to it. But I really do think the lakehouse is better served for like bulk loading, large chunking of data, long-term storage of information, right? You may not want to have all your data historically inside a SQL database. You may want to offload some of that after a period of time and store that down at the lakehouse. I I don’t think your transactional database should have everything in it. You probably at some point need to start lifting data out of it and moving it some other place after

20:03 A certain period of time. That’s I think a better place for where the lakehouse is going to be, right? That’s that’s it’s that bulk loading, lots of information. , yeah, it’s just a different structure. So, I’ve just found better performance by using the transactional side of SQL. There’s other things in there, too. You could use Cosmos database. You could use Custo Kl. There’s other databases in there. I’m just happen to find the SQL database very useful for me in what I’ve been building. All [snorts] right, that that being said, that’s pretty much our introduction. Anything else you want to comment here, Tommy, before we jump into

20:35 The main topic? , I’m trying to think. No, I think we’re ready to go. Okay, so it’s 2026. We are now in a new year. Has the role changed, Tommy? Let’s unpack what does a report developer do? And I’ll just lead with maybe one initial thought when we say report developer. I think in the past I was always thinking about PowerBI reports. That was it. I think there’s another role that we need to add to the report developers tool set. And maybe in fabric has changed this slightly for me as well. I really think the report

21:08 Developer should also be able to build pageionate reports. Now I think there’s two layers of this. We’re talking about PowerBI reports using maybe potentially PowerBI desktop and or powerbi.com the service. I’m okay with you building in either space. If you’re an entry-level PowerBI report user, I would expect you to know how to use the powerbi.com service to build reports on top of a semantic model I give you or is given for you from the IT group. Now there’s this other concept around I need you to understand how to use the pageionate report builder. And I think that to me when you’re an analyst or building reports for an organization, if

21:42 I’m hiring someone, I need to make sure how to do both those things and how to choose between which of the two tools. Are you doing just reports in PowerBI? If you start showing me PowerBI reports with very wide tables, I’m going to throw a fit and I’m going to push you to I’m going to push you to go build pageionate reports and say you need to learn how to build these as well. Let me just pause right there. Tommy, what are your thoughts on this? Is this a is this a controversial topic to you? Is that what you think report developers should do? The only thing controversial that you said was the fact that you brought up pageionated reports here. And why is that controversial?

22:15 That’s part of reporting, isn’t it? technically it is in the umbrella. Yes. But it is becoming more niche or niche and niche and more siloed to see someone who is doing pageionated reports on a large scale. Oh man, Mike and I one I can tell you this from training from people want to Well, yeah. Yeah. Yeah. Now I understand. I want to unpack your comment there. You said you think it’s becoming more niche. Is the niche the actual building of the page report or is the niche we don’t

22:49 Want to export, we just want everything in a PowerBI report? Oh no. It it’s because there’s just other avenues to get their data right now and a lot of times unless you again it’s much more hyperfocused right now. what pageionated reporting is. because back in the day the export feature you can do it in the pageionated report in the service. which you don’t have to know any of the parameters. So really that eliminates a ton of the export data feature because we already have that and that does not take

23:22 Hope if how to use a Windows computer you can build a pageentative report in Microsoft fabric. How does your how does your customers handle the 50,000 row CSV limit and the 150,000 row Excel limit? Mike, most of the my organizations who said they wanted pageionated report training when I started explaining what it was, they said, “Oh, maybe we don’t need this because they had a completely other idea of what pageionated report is.” more and more in organizations are switching to the

23:55 PowerBI reporting world and what I’m at least finding Mike and from my own clients and doing my own when people ask to do training is pageionator report is purely for something that we’re going to be sending over to a client. It’s for POS it is not for an individual anymore to export the data and do their own analysis. That’s where I want to disagree. Like you’re really building PowerBI reports for a purchase order. Is that what you’re saying?

24:29 Yeah. No, what I’m saying unless you’re doing that. Unless you build Yeah. Yeah. Yeah. Yeah. Yeah. Got it. Okay. So, I was just playing around in the powerbay.com service with pageion reports and they’ve enhanced a lot of things since it originally came out. You can add now headers, you can add now footers, you can add images, you can add text, you can even add a table and move the table around on the page. some very basic things. Apache reports doesn’t do everything that the so let me just caveat this right in the same way that PowerBI desktop does more things than PowerBI report building in the service the same thing goes for

25:03 Pageionate reports where pageionate report builder or page report builder does more things than the service does. This makes sense to me that the service is just limited. If you want the full suite of tools you go download the program and you can use it there. Mhm. So so that I think is very interesting having having that already in place is very useful. I want to argue a little bit here, Tommy, that I think people are abusing the PowerBI report side and putting too many

25:35 Tables in it still and instead of treating the PowerBI report as the visual layer, the layer where we have all the visuals, the interactivity, the pages of things, I think we’re missing a part of this that’s well, how do I export a bunch of stuff? And so I think the export story is still incredibly strong. We can use it with export to Excel. That’s very helpful. that that let limits a little bit of what I need to go get access to. I don’t need to export everything there. But I’m working more and more with teams and helping them understand.

26:07 I think there’s a lot of there’s a misunderstanding. This is why maybe when I want to bring this up, right? There’s a large misunderstanding of where should these two tools be used and what’s the appropriate decision tree to go down the I just need a table that has lots of columns in it. I need to go export it using pageion reports. I don’t need that. I’m actually trying to build an interactive report with clicking and buttons and slicers on the filter pane. That’s different and that should be the PowerBar report. And I think there’s this this really interesting world of like you it’s hard to blend the two. Let me just pause there, Tommy. What do you

26:40 Think? So, I think it’s probably good at this point. Give me your definition then of a report developer, right? because I think you are very at least from the definition so far you’re very much focused on the design you’re focused on all the functionality you could provide in a report is a to me actually before I say what I think a report developer and the key word here is developer give me your definition of a report developer yeah and I’m going to I’m going to chat

27:13 On this in lie of like how do we how are we talking about the report developer as it relates to the the hiring experience, right? If I’m if I’m if I’m a hiring manager, right? I’m going to give you my lens of I’m a hiring manager and I want to hire someone to come into my organization to build reports. As the hiring manager, I want to be able to clearly understand we have a world of report needs around full-on interactive reports. I’m going to assume they’re going to be mostly visual. They’re going to be interactive. We’re going to let people find and filter what they need.

27:45 Then there’s another whole world of we have a whole bunch of data and in these semantic models I have it maybe a team that’s building these semantic models and I’m going to need to get access for people to export large amounts of data from those semantic models do whatever they got to do right this is the whole if your if your application doesn’t have an export to excel button nobody’s happy right people are not they’re going to complain right so to me if I’m building very wide long tables that’s the export experience that I’m looking for and I think pagion reports serves that better

28:18 Than in the standard PowerBI experience just because of the natural limitations that desktop and powerbi.com gives you. Okay, so let me just I’m going to phrase it out those ways. There is two needs there. We could argue about how much the need is in one or the other. I would argue the exporting experience is very high still, but we’re not necessarily pushing organizations to do it correctly would be my argument here. Okay, that being said, I’m going to say there’s going to be two levels of report developer as a hiring manager. One is how to build reports and you

28:52 Can use like the simple version. Go to powerbay.com, you can make reports, how to build things using the visuals, how to take things from the semantic model and add them to the page. That’s p bookmarks. how to make slicers. You understand design on the page. You understand what data elements should be on the page versus in the filter pane. you that that is an understanding you must have. I’d also argue in that beginner layer, you need to have familiarity with and I don’t think it’s hard. You need to be able to go build a page report. If I ask you here’s a semantic model, go build me a page

29:24 Report and export this data or get this data in this shape. I would expect a report developer to know how to go to the service and build the report the page report in the service. It’s not hard. It’s basically the same stuff as building the private report. So to me that’s basic level. If how to do that, you’ve picked my you you’ve hit level one, right? Level two is now I’m asking you to understand the applications, the desktop applications that are now being built out. Okay? Now you’re using desktop PowerBI desktop. Now you’re using pagion report builder.

29:59 And again I would argue right now to me the build pattern is go to powerbay.com the service start the pagion report in the service download the RDL and then work on enhancing it locally in the local report builder application because it then automatically handles the connection to the semantic model. It adds all the parameters you may or may not need. It builds all the extra data filters. Building a pageant report is a different experience. It’s totally different than what you’ve been doing in PowerBI. lot less interactive on getting data in and so I think starting with the

30:33 Service and then get it done locally but I would expect the person to have both programs download on the computer and can give me a working example or demo of here’s a PowerBI report that I built and here’s a pageionate report and how to either integrate the two you can you can drop a pageant report into a PowerBI report again if we’re talking like large exports of data that seems to make more sense to me so let me just pause right there Tommy that’s what I’m definition for a report developer has no no lap into the modeling DAX semantic models nothing power query

31:09 Nothing with that I don’t think so Tommy I think I think we’re saying report developer I’m I’m usually I I’m I think the skills to do the report building are different than the skills that are going to be doing there’s some lightweight data modeling maybe but there’s a much bigger story here around what does your company’s data look like and I I like to split the role into two p people right if you can if you can again I’m not saying you can’t I’m not saying there aren’t people that have to have both those roles

31:41 Together but if we’re talking pure report developer I think that’s a role in itself I think there’s another role that your organization needs to have and I’m talking about roles because those roles could be handled by the same person there’s another role that I want to specifically talk to which is the data modeler The data modeler has more experience doing data engineering like things. People who build heavy things in Excel, people are using Power Query in Excel will be great data modelers because they already understand how to shape and manipulate data to get it in a place ready for a report or a visual. So I I think that’s a different skill set

32:16 Than the one around just report development. Now maybe you need another term here, Tommy, that’s like the business analyst that’s doing both those, right? you’re building all the report side and you’re building the model side that I would maybe throw more into the analyst role where you’re actually asking them to do more things. So that’s how I break the minds that’s how I break the two roles up maybe a bit. And here’s the thing though, like I think in an ideal world, you’re you’re probably you’re right. Like I know you’re right, but in real world, yeah,

32:47 Hold on. I’m talking roles. I’m not talking people. I I I get I get it. I get it. I’m saying people that part, but one person can have multiple roles. And so you always go down this path of like, I don’t know. I don’t know. It should have every We can’t hire a person that has everything, which is not it’s just not possible. But we can look for like it’s going to you’re going to be one you’re going to pay more money. It’s going to be more expensive person. They’re going to be more experienced, right? And they’re not going to be just purely a report developer. So I’m trying to distinguish by role what the report developer should be doing. Now, if you

33:19 Want to add into that the data modeling experience as well, well then that’s that’s just a person that has two roles on top of them. And again, I’m with you, Tommy. If I’m a business, I want to hire the person with the most skills, right? be my data engineer and report builder and data modeler all in one person. I would love that. Do we have is that do you have enough work to handle that one person? Is there too much work and you’d need two people? Now you’re starting to divide the roles into different in different people. So it really depends on the size of your organization and how much reorting needs you have. Okay. And that I would not disagree with at all. I think that’s this honestly

33:52 It’s a perfect definition. I I am doing a little perplexity search here and looking and just going scouring the Tommy’s literally factecking me every single episode now. He just No, it’s just because it’s just great because he just runs this stuff all the time throughout the So this is why he’s when you’re if you’re watching live, this is what Tommy’s doing. He’s never looking at the camera. He’s always looking at his screen so he can fact check me on these things. [laughter] But especially when you say something like that universal. And I think I it’s it will bug me until the day I die that we’re in an I don’t know why we can’t get a handle in our industry on clear

34:27 Definitions of job titles and roles because you say report developer and yes I agree with you but most LinkedIn job postings are going to say you’re gonna have to do DAX you’re gonna have to do end to end BI solutions because that’s the word developer. But is that but is that a report developer like that a report developer is a port to your point Tommy right the report developer is a portion of that person’s role right hey your role is data modeler your role is mini data engineer on using data flows right your role is report

35:01 Developer so those you’re looking to hire someone with multiple roles I think if you really specify and pull away and say let’s really focus in on what should the report developer know M these are the like and you could again you could tag that group of skills with any other group of skills you want for what you want to hire. I just think it’s very clear that we point out when we say report developer it means all things PowerBI report all things paging report using service to build both using desktop applications to build both like I think that’s the spans of knowledge I

35:35 Would need the report developer skill to be hired for a new hire so that that’s where I’m trying to go with that. Does that make sense? Yeah. And honestly, the only thing I’m going to keep pushing back on is the page native reports. While is definitely would be part of a report developer skill in it’s less than right now 10% of all the job requirements that say report developer say like pageionated report is a skill and that just came thank you perplexity.

36:09 Less than less than 10%. Less than so they’re out of 18 1,800 jobs specifically mentored pageant report. All right. Ask them about so do the same query now add SSRS. So So I like what you’re going here Tommy. I like pageionated report is the new term for SSRS reporting. So go back to that same query. Out of those 1,800 actions how many mention SSRS reporting or reporting services? that would I’d be interested to see if that number changes at SSRS. Let’s find out. This will be interesting

36:42 Because that will tell you how legacy they really are. So, I think you’re right, Tommy. And also, I think the the language while it’s the same program basically SSRS reporting and pageant report basically looks like the same stuff. They’ve maybe skinned it a bit better. They’ve added some new things for semantic models. But my argument here would be is you’re going to be managing SSRS reports. I think that’s a very legacy thing that’s around SQL Server SSRS reporting and now you’re adding to that pageant reports. The name just changed. It’s still the same skill set. So I’d be curious if that number went higher than 10% after you ask for SSRS. Well, here’s the thing though too, Mike.

37:19 Regardless if a report developer is in our heads supposed to know pageionated reports, I think it’s dying. how like I really think pageionated reports and the need for it because let’s face it you asked the question this is your topic hiring the report developer do we have to like what are we actually looking for today in 2025 and I think that for me is where I’m thinking like the PowerBI role is always been shifting

37:53 So dramatically yes the role power the report developer I am not going to say that it’s the requirement that they’re going to know pageionated reports because they can learn it if they need to because like I said it is a very niche thing just like real it in the same sense that real time intelligence is you’re not going to play with it if you don’t have a need for it if you don’t have an actual actual purpose for it it’s hard to yeah I want to I want to go back into your point there I like that point I

38:26 Understand you wouldn’t hire someone with a pageionate report like experience. But Tommy, when you’re working with people or training people on this stuff, how often are we asking for exportation of data? Yeah, exportation of data from the business users not because they’re going to do a 100,000 rows and even if it’s that Mike the pageionated reports in the service the in fabric that doesn’t have a limit. I thought it does not have a limit. You can go as much as you want. So that that’s that’s the advantage. So the limitations of doing exporting from regular reports has

39:01 Limitations and so that’s why you put up the pageionate report experience. It’s a much better experience. One other thing I’ll just point out here. If you let’s just say Tom you I’ve seen this a lot of times when you build large reports inside the service with lots of columns in them it’s actually really not performant. Honestly, you get really slow, laggy reporting in the service with PowerBI reporting when the table gets really wide, lots of columns, and it just it hangs. That visual will just

39:34 Sit there and chew chew chew chew until it gets all the data into it. So, it’s not a very efficient way of getting data out of that report. And I’d argue when you start landing report pages that have really wide long tables, you’re going to hit the performance limits of the report. And maybe customers are willing to wait for that. But my opinion here is you need some better control around tables of data. Let me give you another let me give you another story or use case around this one too. Okay, if I think about when you have that really wide report, one of the best

40:08 Practices is to make sure that when users enter into that page of the report, there’s already some pre-filters, right? You’re filtering the data down by year or by customer or something, right? You don’t want that single table to have tons and tons of queries behind it because it just really slows things down. If that table also has many many dimensions hanging off of the factual data that’s in that table, you also see that it starts slowing down and getting really slow. So for me that prior filtering right and so we have to train people. The same thing happens in analyze and excel right if you go to analyze excel and you drop in year and

40:41 Sum of sales it will give you sum of sales for all years in your data set. And if your data set’s 10 years long, you’re getting sales for all of that. It’s rip it’s ripping a really hard number to do all that. My experience tells me that’s not really what users are wanting. They sometimes want those long 10-year running sales cycles, but sometimes they also want more recently is I want the year’s data, two years of data. Like that’s most of our action comes around one or two years of data from our organization. And so that is much faster to run. The reason I’m making this point is filters should be

41:14 Applied before you start showing the data. And this is where I would say this is where pageant reports makes a lot of sense. In pageion reports, they can require you select these data fields first and then get your data. And so I really like that apply filtering before you see the results because it really pairs down the semantic model and stops beating up your semantic model. You don’t get that always by default in Excel and in PowerBI. You can turn those settings on or off. Changes a little bit how the report works. , it’d be nice if there was a again, Tommy, I’m gonna

41:47 Ask you a question here. I don’t really know the answer to. Maybe your perplexity can answer it for us, but can you turn on or off per page run queries? So, there’s a setting in desktop that lets you say, okay, I don’t want this report to just automatically grab queries. I want to set something and then hit execute query like run go get the data from the the semantic model and bring it back to the report. You can do that at the entire report level. I don’t think you can do that on a per page basis. Correct. I don’t believe you can.

42:21 I don’t think you can either. I I so the reason I’m bringing this up is the pagion report experience forces you to push in parameters and if you pair the pagion report with an embedded let’s say it this way if you embed the pagion report into the powerbi report you now have the ability of passing parameters from the report side into the pagion report and then you can execute the report the pagion report which again filters down your data so I I think to me this is a very big usability thing that I’m pointing at here Tommy me. Okay.

42:52 And maybe that’s where I’m going at this one. I want to pay some homage here also to the chat. Joe is making a comment here. which I like. I think this is a great very fair comment. It depends on what your port your report is going to do, right? Does your report need to have userdefined functions, UDFs? Does your report need to have parameters, custom measures? I’m going to argue these things are modeler experiences. Like if you’re building userdefined functions as a report builder, you’re not a report builder. You are a data modeler. You have clearly pushed yourself into a

43:25 Higher skill set. That is a data modeler exercise. And you will build userdefined functions in the model, not like it’s in the model side, not the report side. Parameters. Okay, maybe maybe that’s report builder side. Custom measures. I’m gonna have an issue with this one because I think custom measures should only be living in the thin report if they’re stylistic in nature, right? Change this color, make this format. So, if measures are doing stylistic things, I think I’m okay with report builders building their own measures on things. But if they’re doing fundamental

43:57 Recalculations on stuff that’s just brand new calculations, now you’re data modeling. And now I say you have a different problem where you now have business logic living in the report layer and not necessarily in the semantic model. And if we’re taking this whole two-part approach, right? It’s not just one semantic model and report. I have a semantic model deployed separately from the reports. And if you’re splitting workspaces between semantic model reports, it’s another problem as well. So I I see this as being potentially a problem where you need some business rules around what is acceptable in these thin reports that you’re hanging off of a semantic model.

44:30 So I like the comment. I would argue most of these things that you’re describing are more data modeling experiences and not what I would put on a burden for the report developer. let’s shift gears a little in terms of because I think we’ve definitely hit the pageionated reports under the underneath it that’s not the term we’ve hit it on the head but let me just talk about the report developer today in the world of PowerBI and or Mike actually let me ask you was your intention here today to really talk

45:02 About the need for pageionated report to be like minded because if you if you were to take a 100 report developers today randomly. Yep. How many of them do you think are doing pageionated reports on a a consistent basis? Very few. I would say less than like 20%. You like spent the last 30 minutes on it. So let’s Yeah. Yeah. Now is there actual workflow using pageant reports? I would say no. Less than 20% is. Should that be the true statement? I think that’s more of my point here, right? My point is I think

45:35 There’s a hidden gem underneath this rock of pageionate reports that we should be using more. I do think it serves a purpose. I do think it gets businesses what they want. And I think what I guess I’m trying to say is too the slight enhancements that have been made to the pageionate report service with the experience that they’ve built there. It’s not bad honestly. Like I feel like it’s it’s as easy as building a powerb report. So things you were traditionally doing in SRS and I I would also be very clear Tommy I’ve worked with a lot of companies recently that

46:07 Are moving migrating they’re moving from other SSRS experiences into now PowerBI reporting and so they’re trying with perplexity. I was going to say that part. Yes. But they’re they’re so they’re going from an old experience right into the new one. And what I’m saying is you can still lift some of that old experience and bring it with you. You don’t have to like straight up throw away all the SSRS stuff and only go to PowerBI reports moving forward. Sorry, side question, Tommy. Did Perplexity come back and give you any results around SS?

46:39 Yeah, it drives what you just said. It the demand for pageant reports has seen a steady growing trajectory driven by the enterprise needs because because what you just said, say it Tommy. let our drivers the ongoing SSRS to PowerBI migration trend still boom organizations aren’t bad abandoning page reporting they’re transitioning to PowerBI’s so this confirms my point no one’s talking about pageionate reports because it’s a new term I think a lot of

47:12 Organizations are talking more about or at least having conversations around SSRS reporting so if you search the perplexity side of like okay if you just search for patchy report it’s going to be like yeah 10% job description are asking for pageant reports. But when you go back to that and say, “Hey, tell me how many jobs are actually saying SSRS in the hiring portion,” you’re going to see a much larger portion of that. And that’s where organizations are coming from. Organizations are coming from SSRS. They’re trying to figure out how to like shoehorn the entire SSRS experience back into PowerBI. I think it’s a miss. That’s what I’m saying.

47:45 You’re the point here is pageionated serves that need. That’s the place where you’re going to bring that SSRS knowledge and lift it over to PowerBI. Now you can use that, right? So I think that’s an undertalked about or underdised item. Right? That’s that’s why I think this skill has to be a part of that report builder because that’s where this is going to be impactful for the business. Interesting. Tommy, I was surprised Perplexity was so good at answering my question. You were right. Yeah. I’m surprised I was Well, not surprised I was right. Yeah, I had like a I wouldn’t have

48:17 Brought up the topic if I had a an inkling as to like what’s going on like what’s the trend here? So the trend this feels affirming to my trend. Let me ask you this, Mike. Just in general, what’s a report developer doing today that they were doing five years ago? Are they doing anything different? I think this is probably a good place to as we look at if even if that’s 10% and how important that is still I need a report. Yeah. The life of a report developer. [music] So so five years ago if I had to go back to like what a report developer was

48:54 Doing. I would argue a report developer five years ago was probably rocking SQL Server Reporting Services SSRS heavily. Five years ago. Five years ago. No. In 2020. maybe they’re playing with PowerBI, but Tommy, a lot of people weren’t PowerBI was still growing. We’re still we’re still trying to figure things out. Five years ago, we we didn’t even have fabric at the time. We started the podcast because we thought it was there’s enough of an audience. True. But I’m I’m just saying like u general business like maybe yes, but it’s still like small at this time. Like I don’t know what user base we’re at

49:28 Right now with PowerBI right now, but back then it was much smaller. 30 million maybe 15 10 million I don’t know 10 million monthly active users not as nearly as big that’s not a very large audience for businesses right so if I go back five years I think a lot of organizations are still struggling with I’ve got a SQL server on prem I’ve got data in servers I’ve got SSRS ripping reports out and I have a team of people because SSRS is complicated it’s a full-blown application like all the bells and whistles pixel perfect reporting like that’s the impression of

50:00 What reporting was reporting was this very heavy-handed way of doing reports. And then you’re starting to get introduced to PowerBI, right? Oh, wow. This is so fast. I can build three visuals on a page and when I click a visual, stuff interacts on the page. This is brand new. Like SSRS has none of this. Totally new world. You could do a little bit of this in Excel, but then it’s locked in the Excel document. Not easy to share. So I think 5 years ago you’re going to see a lot more involvement with organizations doing SSRS reporting on SQL server stuff maybe crystal reports maybe business

50:33 Objects that those are the the patterns you’re building for people and so now moving forward right we’re trying to get off those systems where do we where do we move that workload from what we were doing which was we had a central team building these reports and having everyone email those reports daily right that that as I think of the the mindset. I’d like to think we’ve moved a little bit away from emailing people a bunch of Excel files and exports to okay now we have a new world where the link is just always available just go to the page

51:06 Just go to the interactive report and we’ll have you land there. So I think you’re going to see more absorption of reporting being handled purely by PowerBI but there’s still this underlying theme of I need to export a lot of stuff. There’s stuff that needs to be exported and I still think that’s around. It’s just now trying to migrate into PowerBI only and I think that’s the part that I see as a miss. It should be a conversation around PowerBI and pageionated. I think you’re selling report developers short because really your biggest point for report developers the ability for

51:42 Their ability to get exported data out for users or for them to have access to that. That’s been the biggest thing you’ve talked about when it comes to report developer. Yes. So, okay. Which I think that’s telling them short like if I knew this like my job is to help people get data out of the systems. No, I I would that’s where I would completely disagree. That’s part of it. That’s not all of it. Part of it. No, it’s part of it. It’s not all of it though to your point, right? It’s help me understand what your insights you

52:15 Need are and let me help you deliver a pipeline of data so you can self-service some of that. Right? So this is going back maybe Tommy a little bit here too is I need to understand enough about the business users needs and if we just give them only reports that meets some of the need really I think the power here is the part of self-service I think serving the semantic model to business users and teaching them how to build their own reports is the real move for business here’s the semantic model we built you

52:47 Developer was that do you think that’s the role of a report developer I think if you’re hiring dedicated for a report developer, right? So, let me think of let me say it this way. There’s a handful of reports that central organizations are going to have to deliver bar none. There’s going to be some level of reporting. Then there’s going to be this idea of we want to distribute or give data governance out to business units. So, here’s five reports that we have built. We think this meets your needs. Here you go. Oh, by the way, these reports are built off of this semantic model. here’s access to

53:20 This amendic model. Go build what you want. You can’t touch our five because we’ve said this is what we need to see for data. We can stand and approve and certify that these are the ones we want to handle hand out like this is the certified data sets but you can go build your own things over here. So if our reports don’t answer your questions, build what you want on the side. And I think that’s the combination of here’s some canned reporting paired with the semantic model and handing that to the business. I think that really enables a business to do what they want to do and and build out what they need to build out. I don’t know if you answered my question

53:52 Because you’re still right now telling me something distin u distinct that a report developer does besides help people export data because you just described to me managed self- service. You described to me the business getting started with reports and you described a business analyst doing the endto-end workflow of doing the discovery helping them design. Sure. To Okay, cool. , but I don’t see I don’t see why any of those skills can’t fit under that report developer. You’re you’re the liaison. So the report

54:25 Developer is building, right? so the report developer knows the model that he has. The report developer knows how to build good content, can handle requirements, generate pages, add the interactivity that he wants, build the data application with the PowerBI reports. Also, the report developer should know how to export the data at scale and have a performant report. So, if you already know all that and that’s what you are as the report developer, part of your role may be working directly with the business and helping them build their own reports or helping them educate around what is in

54:58 The semantic model. I think that’s fair. I don’t think that’s out of the realm of possibility here. Okay. Does that does that conflict with your thinking on this one, Tommy? I think for me I’ve always also had the role of the business analyst or the data analyst like because you talked about the champions program too. I think that’s great. Yep. Cool. as we get near the end of the time here, I’m looking at what Fab Honestly taking your definition of the need for the pageionated report experience or the need to export data to get that data out with what fabric

55:32 Really offers today with their in the service. Mhm. There’s going to be a very niche role for the one who can do the application side because why else would you use the application now like the report builder, right? Okay. there are a few obviously purposes but and to your point the SSRS migration but if I was hiring a report developer today and I honestly I’m not looking so much right now as what I need them to do today but also what is this person going to do a few years from now

56:06 In the future like are I don’t want someone who’s stuck in SSRS that’s their skill set for where we are with fabric today and I think what the world of report developer is changing. I would agree that’s not their only skill. I think the idea is to migrate some of what SSRS was doing into PowerBI reports because there’s to your point there’s like a I want to call reports the data application. Yeah. Right. It’s it’s the interactivity. It’s drill through pages. It’s what navigation menu do you put in there. Does like there’s a whole bunch of other considerations that

56:38 Are on the report layer. Does it look good? Does it interact well with your data? Is it actually answering business questions you care about? Right. There’s a lot of those things that need to be there as well. So to your point, Tommy, like I agree. I think there’s this this lens of does the PowerBI report builder, right? It’s a mix of this old world that is SSRS that’s coming into this new world which is PowerBI. And I would say I would want someone if I’m hiring someone, I want them to have the skills in both areas so they can betterly better betterly that’s not even

57:10 A word. So you can better better articulate what which way you’re supposed to build this stuff, right? I hear requests from the business user. If you don’t ever know pageion reports exists and you’re ask you’re getting this ask for the really wide table. Can’t you just add these other columns? Can’t you add this other stuff? You have no knowledge of pushing that user to what you’re what I hear you saying is you need a large wide table of data to export. And they’re going to say yes, that’s what I want. If you don’t know about it, you’re going to go back to the PowerBar report. You’re going to build a very poor experience. it’s going to be slow. People aren’t going to love it and you’re going to

57:43 Need to go back to the user and say, “Well, here’s what it is.” Instead, I’d rather have them have knowledge of pageion reports and say, “That’s a decision point for the pageant report side. Let me build you one of those or andor pairing the report and the pageant report together. Select your parameters here, hit run, and now you have a paging experience of pagion reports that you can now export and not have any limitations around data limits.” So I do think that is a legit business case. All right. So if you were a report

58:15 Developer today or if you were hiring report developer stuck in this role for them to for looking out to the future, what would be the first three priorities of skills for that person who’s stuck in the weeds where you say this is where you need to be or this is what you need to learn because this is where the report developer role is going. First and foremost, I’m going to say it’s going to you’re going to need to know most knowledge around PowerBI and PowerBI desktop. That that by far I think is the best desktop. That’s surprising for you to say that.

58:48 There’s things you can do in desktop that you can’t do in the service. There just is. Service is still there’s things you can do. Oh yeah, I know. There’s things you can do in desktop you just can’t do in the service. Now, it would be nice if they all I wasn’t sure if I heard you correctly. I didn’t need AI for that. I did not need an agent for that. [laughter] We’re getting there. Tommy’s going to clip that segment and he’s going to put it all over every single social media. Mike said you got to have desktop. I’m not going to disagree, Tommy. It doesn’t do everything, but if I’m doing lightweight stuff, it gets a it gets you really close. So, you can do a lot of things there. Agreed.

59:22 But that’s where your your knowledge should most live there. You’re going to need the point where you’re going to have to decide, do I only build this user or user experiences only in desktop or only in the report side experience or do I also need to build some pageionated report exportation experience and I I think a lot of this comes from when you need to export data is when the user of the data the business user doesn’t quite understand what to do with it yet doesn’t quite trust the data right until they get their hands around I actually trust the

59:54 Data structure or what information I can I can literally export data and go back to my operational system and see the numbers matching right I can see the transaction is this amount here’s the invoice details the way I expect to see it okay now I have mental model of like this column in the semantic model matches this field in the database in in the operational system now they can start trusting right so I I feel like there’s this trust issue that’s an underlining topic around exporting data there’s once you get through that people are able to relinquish the exporting of

1:00:27 Data a bit more and give responsibility more to the report and the reports that have been built. So maybe that’s my opinion there a little bit. I’m sussing this out Tommy as we talk. What do you think? What your reaction? So yeah, honestly what I would take a look at for the report developer role is get used to fabric, man. And I think the report developer is going to be I don’t think they’re going to take the role of being like when we say developers like front and back end. but I feel like there’s going to be more of a need for them to be they already have the role of

1:01:01 Working with business and they working with the stakeholders and understanding the business. Sure. So to me I’m I I don’t think it’s going to be the universal way, but I would like to see that tiger team approach to get acquainted with fabric, get acquainted with the lakehouse, and then what you can solve for people is what they’re trying to do anyways with reports is get their data in centralized location and help people out through the lakehouse to get their data connected and all the different ways that they want to view their data. Because if I

1:01:33 Have the lakehouse too, Mike, then there’s a lot more I can do with that in terms of why are people exporting? I guarantee you most of them are not just doing pivot tables. They’re probably sending it to a vendor or, some piece of information that way. So trying to find solve those 100% Tommy, we’re just trying to get around problems. Why not build why not go in and just build that page report that has the columns and data you want? instead of going in like right like this is we’re in that place again where like okay I got to get this preview maybe I’m going to send a screenshot or two and I need this large

1:02:06 Export of data I’m going to go give it somewhere like that makes sense yeah I’m with you Tommy awesome awesome well that being said thank you very much for listening to this episode this was a weird conversation around where does the report developer sit in 2026 what skills should they have well apparently we have some lingering SSRS skills that are still hanging around the BI space that we need to kind address here and I think that gets fit met by the pageionate report experience. That being said, thank you so much for listening today. We hope you enjoyed this topic, this conversation and we hope you gleaned some value from what we

1:02:38 Were talking about today. That being said, Tommy, where else can you find the podcast? You can find us Apple, Spotify, wherever your podcast. Make sure to subscribe and leave a rating. It helps us out a ton. You have a question, idea, or topic that you want us to talk about in a future episode, head over powerbi.tipsodcast. Leave your name in a great question. And finally, join us live every Tuesday and Thursday, 7:30 a.m. Central, and join the conversation on all of PowerBI tips social media channels. Thank you all so much, and we’ll see you next time. [music] Tommy lighting up the sky. Dance to the

1:03:15 Laughs in the mix. Fabric and AI, get your [music] explicit measures. Drop the beat now. Dest feel the crowd. Explicit measures.

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.