Creative Solutions & Data Products – Ep. 333

Here’s the latest episode of the Explicit Measures Podcast—jump in for the news, the main discussion, and what’s coming next.

News & Announcements

• Microsoft Fabric June 2024 Update | Microsoft Fabric Blog | Microsoft Fabric — Welcome to the June 2024 update. Here are a few, select highlights of the many we have for Fabric. Fabric Spark connector for Fabric Data Warehouse in Spark runtime is now available. Query activity, a one-stop view…
• PowerBI.tips Podcast — Subscribe and listen to the Explicit Measures podcast episodes and related content.
• Power BI Theme Generator — Create and download Power BI report themes using the PowerBI.tips theme generator.

Main Discussion

This episode is a tour through community ideas and “creative solutions” that are pushing Power BI/Fabric forward—paired with a recurring theme: treat reusable assets as data products, not one-off reports.

Key points:

• The admin/monitoring experiences are evolving (usage + adoption, tenant-level insights), and teams should actually use that data to guide cleanup and prioritization.
• Templates, themes, and repeatable patterns reduce time-to-value—and keep teams consistent.
• “Data products” means designing models and datasets for reuse: clear ownership, documentation, and a stable contract for consumers.
• The best creativity isn’t novelty—it’s packaging best practices so others can adopt them quickly.
• Small experiments are fine, but production needs a path: versioning, governance, and support.

Looking Forward

Pick one high-traffic dataset/report and elevate it into a true “data product” (owner, description, metrics, and reuse guidance). Once you do it once, you can copy the pattern across the rest of your portfolio.

Episode Transcript

0:32 welcome back to the explicit measures podcast with Tommy Seth and Mike good morning morning everyone good morning Mike happy Tuesday it’s a Tuesday Tommy good morning morning guys Seth is Right AI almost took over my computer but we’re okay so far so good if Cony if tommy’s video cuts out halfway through here you’ll know his computer just took a took a dive and it’s done you’re gonna leave me behind we’re not waiting for you Tommy we’re just going to move

1:03 just going to move on main topic for today just jumping in here there are some really interesting Creative Solutions that are occurring people have been making some really good blogs there’s some like hack together comments here so Tommy’s got a a collection of Articles here about talking about just being creative in data visualization and how that can help your reporting so we’re we’ll go through a couple of these articles we’ll put them out there’s a couple from data chant we have data mark data mark something something from U Pier I believe on medium anyways

1:36 from U Pier I believe on medium anyways we’ll pick out some of these articles and go through these today but before we go into our main topic let’s talk about the news there’s a couple items here I want to touch on one I think that’s very relevant here is especially for any admins or people who are using the admin portal this may or may not impact you very much Microsoft has announced that the admin portal for the usage metrics is immediately retiring not next month wow not next month this month this month like right now it’s

2:06 month this month like right now it’s done done on it’s a little little little short notice there not like I was going to actually download or use anything off the report anyways so here’s the here’s the link or the the news article from from Microsoft so I’ll put that in the chat did you use this at all Seth honestly honestly like the yeah periodically I I don’t think I ever got we’ve ever gotten around a custom building a solution that would give me everything I wanted so

2:37 would give me everything I wanted so yeah when there was nothing the feature usage and adoption report which is now part of the admin there’s like a workspace now that is admin based it’s a whole and I think all of the main objects that number four item the usage audit by workspace with a colored graph it was like even the report they showed you there was like almost useless yeah like what are you really going to do with it like super high level basic if someone really needed to know how many users are do we have in powerbi

3:08 many users are do we have in powerbi that’s a great place to go if I wanted the most quick and easiest way but it’s never changed since I think they’ve had the admin portal with the tree map with the just count of numbers it’s it hasn’t adjusted at all so that’s probably why they feel so comfortable also going hey just letting you guys know we’re getting rid of it because no one’s relying on it and I think I think to their point to their credit here though there already is a much better report out there and the the new report the one that they’ve pushed inside to the admin

3:40 pushed inside to the admin workspace monitoring piece there’s a lot more data there and you get the all the same information and a lot more directly from this new report which I which I do understand it just is interesting that they just so quickly change it over and immediately make a change on top of that do you think they’re going to export in there they’re going to keep as the app the app itself I think they’re going to continue to invest in the Microsoft admin monitoring apps I think if anything they’re going to just make more

4:10 anything they’re going to just make more apps that are going to be built in there honestly so there’s going to be it makes more sense honestly to not embed a report in the admin portal it makes a lot more sense just to put the reports in a workspace that is for admins that they just get to use that makes it’s easier it’s easier from a security perspective too because the fabric roll right the admin rule enables and disables that workspace now right so you you don’t even see that whole thing so if you’re going to build out the whole administrative piece that can be a

4:41 administrative piece that can be a completely separate space now which actually makes makes a ton of sense and and you can also go into the admin workspace and you can create reports on top of it are you able to share things out of there or is it just is it like you can I know you’ve been what I’ve been able to do in the admin monitoring workspace is I’ve been able to take an existing data set that exists in there and make net new reports on top of it so you can take the feature usage and adoption data set and build copies of the feature and usage

5:12 build copies of the feature and usage adoption report and then you get you adoption report and then you get specific things that you’re looking know specific things that you’re looking for right so I think I think even that is that’s actually so one of the reasons why I think I would want to see this thing the admin portal canceled is because I have no flexibility to change that if it’s a tree map it’s a tree map I don’t get the any no adjustments I can make there at least now in the admin monitoring workspace I can at least go create copies of reports I can adjust them I can make new tables I can find what data points are are more relevant to my

5:42 to my organization yeah we’ve actually done that in the past but it was just the usage reports in a given workspace and it was only for that team where the stakeholders wanted to see usage of everybody so this is much better that it’s globally works the same way yeah I’d agree with that I’m trying to shut it down shut it down I would imagine usage has got to be pretty low on it probably right why wait why wait no one’s using it cool all

6:14 why wait no one’s using it cool all right moving on to the next topic Kurt buer and sqlbi so Kurt I believe works for sqlbi now does some articles and blogging for them really interesting status I think Seth you saw this one catch on LinkedIn I saw it catch on Twitter but there’s C has a status here around there’s this what micro what I guess SQL b bi is calling what are they call it it’s it’s the a library of reusable objects patterns and templates that’s what the the Twitter post calls it if you go to the micro the

6:45 post calls it if you go to the micro the sqlbi website I think they’re talking about using Atomic design report and model model development Atomic design I think Atomic design this is like the book I’ve been reading which is called atomic abits very small things impact very big things so I think the the Twitter the link the Twitter post here I’ll put that here from from Kurt it’s very scientific it feels it that way their their graph is like the building blocks

7:16 their graph is like the building blocks of atoms molecules organism templates Pages wait what transforms it into U the different objects patterns visuals page templates report templates and it’s much obviously from a developer bent right like how do you how do you structure the back end to be a repeatable like rebuildable thing so yeah can’t wait to dive into that one I think that’ll probably be a podcast episode for us at some point I think so

7:46 episode for us at some point I think so I think so I did I did very much like to I want to do point out here is you to I want to do point out here is objects like Dax measures build know objects like Dax measures build patterns like Dax patterns which build visuals which build a template for a page which eventually builds a report template and you can have all these kind template and you can have all these things that are stacking of things that are stacking these are these are hierarchical things that you build I’m very proud to say that Seth and I have been building templates since what 2018 almost like six years now long time long time so we’ve been we’ve been on this bandwagon of like

8:17 been on this bandwagon of like it’s important to get templates created it’s nice to see some the community kind it’s nice to see some the community jumping in on some of this as well so of jumping in on some of this as well so not that we need to lay claim to this first but we did it first we did it first we were we were there it’s a different it’s a different area yeah this this is To Nuts well I don’t think this was really possible until we actually had the ability to create Dax measures and measures in Fabric and a a semantic model there because this doesn’t really make as much sense if this is three years ago when we’re

8:47 this is three years ago when we’re talking about powerbi because you’re still dealing with Dax measures in a single model and desktop now you’re actually at talking about different components that in a sense part of it being structurally laid on top of each other you’re going to have Dax measures and tables that are going to be living in a lake house and a default semantic model where it’s gonna in a sense it’s GNA be set there it’s gonna stay there so I really like it I’m really intrigued and I’m going to dive in this too in terms of

9:18 in this too in terms of organization right because it makes sense but you can’t organize it that way well let me I’m I’m gonna I don’t I’m going to pull back here just mean I’m going to pull back here just maybe a different thought here a little bit mean sqlbi has been the forth front before I’ve even started powerbi anything sqlbi had Dax patterns out for a long time so they’ve already been like so if you think about what they’ve been doing they measures by

9:44 doing they measures by themselves they’re really good at doing documentation around what measures are existing and this is I I say this because Dax stock gued it’s amazing it’s really it’s the best documentation place for Dax and then if you look up a level then there is the the patterns Pace piece of this right and that’s and that’s exactly what Dax patterns is it’s a whole book around these are common patterns you’re going to see when you build things in data use these patterns to build rolling totals these are the most efficient ways to do these things so that makes total sense

10:15 these things so that makes total sense where I’m where I’m like laying claim to some of the part of this is you claim to some of the part of this is a lot of the template pieces Seth know a lot of the template pieces Seth and I we were really very early on and like templates are the thing to be doing we should be thinking about how to build templates and one of the things these are all things that Microsoft really hasn’t been able to embrace they built the tool they built the measures but there’s really no good place to collect the patterns collect page templates and collect report templates there’s nothing there that supplies capability and the community I

10:45 supplies capability and the community I think is building these things for you and so Bravo comes out you can now add templates of tables date tables and measures the theme G Gallery at parbi tips themes. parbi tips has report templates full design templates things for the reports and I just don’t see any no one’s really making page level templates yet and you see what I’m saying want to go with this Tommy like these things have been out for a while but Microsoft’s not building this into their tools right and I and I have that one completely right but

11:15 have that one completely right but they’ve always been separate anything page template report design and anything that’s going to be semantic model I think the idea here is that you can actually bind together what you’re doing from the visual side with a measure you still have to do all the drag and Dot yes yes well yeah I I don’t disagree but I I think we talk about this at length in a different episode but I I do and and the reason is and the reason is is this approach seems to be very much building from the back and then can you

11:46 building from the back and then can you throw on page templates when you’ve got this perfectly organized model sure again but guess what that’s not how the world works though typically the world Works build from the front forward so tools tools like we build are designed so that as you’re figuring out what a business user needs from the front end you’re cutting you’re streamlining the whole visualization experience right so I think I think this requires us to dive deeper into the article what the offering is in total I like it and then

12:16 offering is in total I like it and then and then we can embellish and talk a little bit more about it that’s a good good thought it’s definitely gonna be a topic it’s definitely going to be an episode all right last one here is we have a lot of news e articles here Tommy which one do you think is best here we have three left fabric. net SDK automate realtime intelligence using Powershell or the powerbi refresh performance analyzer I think the performance is what I want I think that’s the best one talk about

12:46 think that’s the best one talk about let’s do that first because I’m gonna totally dive into donet because dude no one’s gonna listen to that one I’m like that’s such a nerd Yourself by yourself somewhere on a different on your other podcast okay what we talking about performance analyzer okay let’s go let’s do the performance analyzer first we’ll go through that one all right so this is a an article from Andre talking about the powerbi refresh performance analyzer let’s put this out here this is another is this an

13:17 this out here this is another is this an external tool what what is this Tommy yeah so this is an external tool that will actually look at when you’re model refreshes and loads basically really break it down in terms of what elements are taking place with a really nice visual to actually see the different elements in a n like in natat style oh I’m at yeah so it’s actually the perfect visual for this but it’s I think a really great way rather than just looking at the numbers and we I don’t think we’ve really had anything

13:47 don’t think we’ve really had anything this easily accessible to look at what’s causing something to take a long time to refresh and reload I like it so I’m trying to I’m trying to thinking here there’s two point out articles here that I think are very relevant inside this this topic here you can connect SQL you can connecting SQL Server profiler to powerbi premium so you can go look at refreshes that are happening inside the powerbi. com service as well as you can also connect the SQL Server the

14:19 also connect the SQL Server the ssms SQL profiler directly to desktop as well and then there’s actually a Json file that or a an external tool that adds this as well so I’ll have to go look at that I’m not sure if we have that in our business Ops tool at this point I’ll double check that one maybe we can go add that in there as well if we don’t already have it this would make sense to add into business Ops Ops excellent I in in the past Tommy have you used the performance analyzer going through SS SMS to look at the refresh

14:49 through SS SMS to look at the refresh what’s happening there’s been a few times but nothing where it became part of a process or workload because anytime you jump into the profiling tools yep if that was not your background and then trying you have how much you have to filter out just to see only just what’s actually the execution of a model it’s it is for those who’ve never done this if you weren’t sure if you wanted to be a DBA this is your this is your test well it’s coming out of ss SMS

15:20 test well it’s coming out of ss SMS which is like the DBA tool yeah have you played around with this at all have you gone through performance analyzer and read some stuff off desktop or the service not specifically okay like like a while ago but like not like recently on like in the house refreshing been a long time this may this may turn into a little bit of beat from the street here a little bit about like real world experience here that I’ve done so I’ve played around with this actually I spent a good amount of time I was having issues

15:51 amount of time I was having issues refreshing data models with incremental refresh because stuff was failing and I didn’t know why this is the ex L the pattern I used to go figure out an incremental refresh was failing or was falling back I didn’t understand what was taking longer was it the SQL server or was it the backend side I don’t remember if it SQL server or not actually don’t know was it the back end not actually I think I was actually think it was SQL Server now that I’m bringing my memory B back about this the SQL server was returning queries that were extremely slow and so there is a

16:23 were extremely slow and so there is a amount of time that takes to refresh the incremental refresh patterns and on your your first load like when you create a new model and first publish it to the service it’s going to run a whole bunch of queries to fill all the partitions of your historical data pattern and if it doesn’t have the data then it it goes quickly through them so there are occasions where if you don’t write your your views or SQL queries correctly on the back end when it’s query folding it will take a long time to do stuff and so I used this and I actually went really

16:54 I used this and I actually went really deep I used Phil C marks analysis there’s a that I think Phil C Mark also put out around this where he was showing you they would give you the timing and they would tell you which partition was refreshing and what was happening in sequence in time really interesting stuff anyways I I use this and I actually spent a lot of time learning how the engine cashes up a second model while it’s doing the refresh because when it’s doing refreshing elements it won’t delete the existing model when you

17:24 won’t delete the existing model when you do a refresh it it spins up a brand new model loads the data and then swaps the data at the very end so you have your new data in your report I think this makes a ton of sense to me this this using this tool or something like this one was very helpful and solidifying how is the engine automatically building the partitions reloading the partitions incrementally refreshing things and then I was able to go into the weeds and figure out which partitions were failing and what was going wrong and I could go back Upstream to the SQL Server

17:54 could go back Upstream to the SQL Server to go find what was the problem anyways I thought this was very relevant but it was super super technical to your point Tommy like this is this is not a normal business user level thing this is a DBA solely or data engineering solely kind of exercise yeah the performance analyzer there’s so many filters and Views you have to do this is just a great visual of basic what you were trying to do before yeah excellent all right I think that’s enough from the the news let’s I guess we could transition I had

18:24 let’s I guess we could transition I had I did a little bit of a a Beat from the street here Tommy do you have any other topics I think you’ve got some pipeline and items here you want to discuss I think we’ve been talking about this for a couple of podcasts now yeah I know I’m not gonna go did you hit the pain I hit the pain and the article is actually great timing by Microsoft with the net SDK coming out okay I don’t what’s What’s your deal about this dude are you loving net SDK all of a sudden I don’t know any I want to stay away from this so I I’ve tried creating my own applications because they were a few

18:54 applications because they were a few like you like start your own project but I’m like why so I I’ve noticed this with all the fabric API documentation it’s all written inet no they just came out with something in Powershell but really until now the only resources available for anything that had to do with the fabric API was net because it could use a service principle there’s no python there’s even no HTTP like examples by Microsoft which was really strange that it was allnet

19:26 was really strange that it was allnet people like who’s using this now so I

19:30 people like who’s using this now so I was trying to basically set up a pipeline to run like basically click a button in powerbi that would trigger a pipeline pretty pretty simple right if you were doing this in data Factory because you have plethora resources well one I realized that there’s no standard Connection in power automate yet for fabric which I thought think is a little strange that they haven’t come out with that yet scorecards are plentiful as actions in Power a M and then two

20:01 actions in Power a M and then two it’s not nearly straightforward in terms of if you want to trigger a job run for a pipeline again one there’s there’s no the only way you can you can do it with an API but pipelines cannot run without a user attached to it so there’s no Ser you can’t attach a service principle to run a pipeline right be an API and I’m actually I’m quite frustrated by this because this seems to work everywhere else like anywhere else if I’m in data bricks there’s an API to go run a pipeline there’s data bricks to go reload a table or run a job or like

20:31 go reload a table or run a job or like there’s a whole bunch of other apis and other tools that are just out why is this not a thing yet yeah yeah exactly right I agree with you on this one Tommy so yeah I’m just finding very strange that they’re putting all that time net even though all of us are python Powershell and just good old apis what I I would to your point Tommy I do agree the the the apis I think is the way to go I and I don’t know net my guess is the net thing is not using anything else other than the

21:01 not using anything else other than the actual apis there’s there’s nothing special about what they’re doing here oh it’s special did you see how much it takes just to do authentication with the net you have to write down but it’s not just net though I that’s just like them how just how they’re doing authentic my my point is the net library is just a library of all the API wrapped in things that are like nsh I think again I could be talking out of I don’t do a lot of development so I’m I’m talking naively here a little bit but to your

21:31 naively here a little bit but to your point Tommy I’d rather have just the apis and whenever I build applications I build applications solely on the apis I don’t use the commandlets I don’t use anything else because I know those things are fundamentally built on top of the apis so if the API doesn’t work the commandlet won’t work like plain and simple and there may be some couple little tricky things where the commandlets that Microsoft is built for powerbi or I don’t even know if they have fabric commandlets yet I haven’t even really heard anything about that yet but if they commandlets sometimes they use hidden apis that you don’t

22:01 they use hidden apis that you don’t actually see normally but they’re inside the commandlet sometimes but that’s I think more of a rarity than a than a common practice at this point anymore anymore anyways I’m with you on that one Tommy I’m not a big fan of what going on here maybe there’s a bigger Vision here of Microsoft building other tools integrate with net doesn’t this leverage or open more doors to custom applications not using embedded yes it does I think I think this

22:32 yes it does I think I think this is doing yeah go ahead no I’m just saying like the some of the like we we view Fabric and the entirety of this you view Fabric and the entirety of this analytics platform in the through know analytics platform in the through the lens of like embedded or calling and retrieving data Etc but in some cases like doesn’t this make it easier for custom applications to integrate with the data on the back end side of this which is extremely important if you’re custom building something and you need access to the data sets or you or potentially the

23:02 data sets or you or potentially the models or whatever pieces behind the scenes here like you can’t just assume that every aspect of the way we build things or put an eye frame into something is a good user experience because it’s not so I Seth I think you’re speaking truth to power here that actually makes a ton of sense with their custom workloads that they’re coming out with right well I think I think to your point Seth there is is I think you’re right and I guess I’m when you said Tommy about the workloads

23:33 when you said Tommy about the workloads maybe this is a play for hey we’re going to let more companies build workloads on top of the back end of fabric and this is a way most there’s a they’re seeing in an experience of there’s a lot of net things that exist and maybe that’s there other honestly Microsoft is very self-serving in these ways maybe this is a way for them to get into the back end of RBI for Excel or other things that are already written on I don’t know Excel is written on maybe it’s net I don’t know but there’s other things that Microsoft has that maybe they’re trying to like hey we need to get back into the back end we need these libraries to

24:04 back end we need these libraries to exist for us to even talk to the back end of powerbi as well so there may be more items in there additionally it’s funny they’re quick start for the the official fabric documentation for the API say hey it’s quick starts create an entra app here’s a CP script to get you started I’m like who’s using that but I think set that’s it really does make perfect sense if they want to do custom workloads it’s just up until now all Microsoft documentation and the products that we’ve been using have been here’s a

24:35 we’ve been using have been here’s a quick start with power shell Powershell and a little python would agree that yeah the vast majority of this Pipeline and everything we’re building is analytics for analytics tools for for what we’re using in powerbi or data dumping or trans transforming and shifting and moving data around but there’s this whole ecosystem of applications that build visualizations right it’s not just it it opens the doors to building those custom experiences if

25:05 those custom experiences if organizations want to right because there is a very specific look and feel interaction between Pages something has to go from here to there in this way and they’re not willing to do that through powerbi embedded experience or something to that regard good point I didn’t think that way you guys are opening my mind this is good this is helping me expand expand my experience here so thank you very much for for going deeper on that one I think that’s also really good all right I

25:36 that’s also really good all right I think we’re at we’re good at a good point let’s transition over to our main topic of the day I think it’s time to jump into just let’s talk about the the main item here we leave the Beat from the street behind main topic today is articles around being creative and not just in data visualization Tom you want to kick us off here and give us some intro on this one and we’ll take it from there yeah so this is one of our special topics where I think it just came from our previous discussions but do as some research and I think Mike

26:07 but do as some research and I think Mike you actually suggested this this topic and okay when let’s think about that word creative and we’ve done episodes on it but the default way to think or the I think the normal way people think when you think creative powerbi business intelligence is are designs it’s going to be your templates it’s going to be how the report look and feels but I going through some of my old articles that I’ve saved in from Community Jam another shout out boom boom just typed in creative and and

26:38 boom just typed in creative and and really there were some really interesting ones like the hackathon that Microsoft did with just did last year for fabric our data Mark did a great one on how do you actually really transform a solution that you might have been working on with a small team and scale it out which the approach to that is I thought I was oh that’s pretty pretty interesting agile approaches how to become a data driven organization and really just data storytelling and I think when you think of yourself in our

27:11 think when you think of yourself in our space data basically is Clos as you can get to it without being it there is a I think a big need one for the talent or the skill to have be creative and I think two to also have the time the time and to part of what being creative is to try things out I think a lot of times just as you look at our intros oh I can’t believe net and we’re talking net and c and python which is always usually

27:41 and c and python which is always usually pretty straightforward it’s get this data pull it here I think in our processes I think in our conversations communication with our clients and how we actually grow our relationships with our organizations and our approaches to connecting extracting transforming and then communicating data there’s a lot of opportunity that I think is one of those known unknowns I don’t think we’ve explored it a ton because we don’t think about it in that space usually

28:11 think about it in that space usually we’re very closed when we’re building so I think that’s where I just wanted to start start here you wanted to start with everything yeah everything just a nice in well let me I’ll give you maybe a little bit of context of some so to your point Tommy I like where we’re going with the topic I think there’s a lot to unpack there so let me start with a little space where I’ve experienced some let’s maybe call it push back from from the business I guess often I’ll be in I’ll be in the context

28:42 often I’ll be in I’ll be in the context of reports and we’ll be talking with the business we’ll be building some reports and usually we get good examples initial examples of what the business wants for reporting most often when I look at that reporting it feels like we’re talking about data dumps I need this table I need the the data shaped this way because I have other processes that rely upon upon that sometimes when I step into these situations there’s this deeper level of thinking that comes with looking at the data and what are the outcomes I want to produce from that visual I was

29:13 produce from that visual I was working on a project recently they are in the the manufacturing or the shipping space and there was this idea of we have a limited number of trucks and we need to make sure Each truck is filled up it’s full and so anytime we’re not shipping a full truck of inventory or whatever it is we were we’re we’re not fully maximizing our revenue and so I was like well if that’s our driver if that’s the information we’re looking to get get towards what visuals can be put in place

29:44 towards what visuals can be put in place to say okay this truck is this truck will be Liv leaving on this date what is the opportunity for us to fill up that truck and the idea being the closer that truck is to its leaving date the more I should pay attention to to make sure that it’s full of inventory to make sure that when it leaves it’s it’s maximally using its space so it was kind maximally using its space so it was like that thinking it was a question of like that thinking it was a question of like a deeper level thinking and we came up with a visual that showed this it showed hey here’s it’s it was a bubble chart and we had some other things like how full is it how many days

30:15 things like how full is it how many days away is it leaving so we had quadrants of if things are in this quadrant it’s the the truck is very filled up and it’s about to leave great we’re good but then when we had these other bubbles in the chart well this truck is not very filled up and it’s about to leave in 2 days okay we should really focus on what can we do to sell more space on that truck to get it to work to to sell right so it was that meal mental thinking of we did this pattern with them I produced the visual and when we presented back to the business they weren’t able to grasp the understand the

30:46 weren’t able to grasp the understand the concept of the visual the visual just did not make sense and so while some people were able to understand it there was a data culture gap between what the rest of the business could do with it versus what was actually being shown to them so I think this is a challenge that we Face a lot with being creative to your point Tommy and where where I’m going this I’m going to trying this back this back around in some situations we’re being creative to build visuals rep pages and reports to to meet a specific need but sometimes if our data culture isn’t

31:16 sometimes if our data culture isn’t there even though we’re being creative we just fall back to I just need to get a table I just need to get a bar chart we’re going to throw down a pie chart on a on a page because people aren’t ready yet with their data culture to really digest these more advanced creative experiences and reports so that’s kind experiences and reports so that’s my point there I that was a long way of my point there I that was a long way around to that point I really like Adrian n Adrian in our chat says users think in data dumps because that’s

31:47 users think in data dumps because that’s the closest they get to data sometimes I would say so true it’s like resonates right with your point Mike right from the standpoint that it’s data culture thing right like people aren’t aren’t used to seeing certain visualizations and or what you’re describing is a wellth thought out bubble chart with quadrants like yeah how often are they interacting with those on a daily basis exactly right well it t it takes a little training and learning for them to say Okay I I find the

32:19 for them to say Okay I I find the meaning in that so I think to me there’s two parts of potentially this conversation that resonate with me that I want to dive into one is what you’re talking about is this solving the problem right there there is a problem that you are going about trying to solve and what you’re doing is trying to build the underlying story The Narrative underneath this visual right

32:51 Narrative underneath this visual right and that’s where and one of the Articles was talking about this too is like the storytelling aspect of how he doesn’t think AI is going to take over in in totality because there’s this human element and if you think about the best reports it isn’t the best looking right like that helps AIDS AIDS people in digesting the the information what resonates is the meaning behind it right and I think that’s one of the things that is lost sometimes in this

33:22 that is lost sometimes in this translation between multiple different groups working on a project not being close enough to the the people who actually need the the report or we don’t understand the needs completely where you build something where you’re not getting enough value in the data that’s being presented and it doesn’t have to be like this a very specific use case like you talk about there are reports that are daily need reports I need to see this information in this way and it tells me something that I can

33:52 and it tells me something that I can make a quick decision on that’s a very valuable report right so I would I would almost argue like The Narrative behind that is hey we’re serving this pivotal business need where people need to know what’s going on every morning to check in with our customers and this report shows them the that information so I think it’s it’s that like in and I’m just lumping that into storytelling right but it it’s the the value of the data that’s being presented in in visualizations and I’ll pause there if

34:25 visualizations and I’ll pause there if you guys have any comments but then there’s another followup I want to I want to talk through I think that’s a great place to to hang on here because man it is all about not just the data but I I’ll give you a brief story so I thought I was the most creative person when I started with powerbi and I always wanted to do more because I thought the bar chart wasn’t enough but it turns out people didn’t care about

34:55 it turns out people didn’t care about doing multi-line visuals if they didn’t again know what it meant and again the first example that you gave about what they need to look at every day well that’s almost builds itself right because the people know what they’re looking for and that’s the big distinction there where this is a number that’s integral to me I understand its context and I know what I can what will happen if it goes up or goes down so in a sense that in a sense almost

35:25 so in a sense that in a sense almost rights itself but I think we get a lot of more situations when we get more granular in the business when either we’re dealing with departments where there are numbers that they need to see but it’s not that same in a sense direct action however it is it’s just not what they’re asking for so I think a great trans like turning point for me in my own career was trying to build a report off a cell it was basically 90

35:55 report off a cell it was basically 90 different values or numbers of volume and we basically changed the way the organizations were looking at their metrics but they still had the volume but the entire report became rates because just a conversation with them is okay if it was if you’re looking at something 7, 000 on a weekly basis how much are you going to do if it’s five more or five less what they actually cared was something actually completely different they just part of it is not

36:26 different they just part of it is not that they don’t know but that’s how they’ve seen it so I think a lot of times too it’s even before we get to the even report itself in that conversation on yeah I could show you total sales but that’s probably not what going to drive you it’s it’s important but the just volume of sales alone for a single person specifically is probably not the whole story or enough of the story for them to really get any meaningful Insight out of it

36:58 get any meaningful Insight out of it I’m want to I want to comment on your so I want to comment on both your you guys are making great points here Tommy I like your point as well Seth I want to go back to your point as well and when and your point is around something that brought up my mind here around there’s the Gartner analytics maturity model I don’t know if you’ve heard of that one I put it in the chat window as well here so that’s also out here as well it talks about these I think it talks specifically around you were addressing Seth there’s these things like descriptive analytics diagnostic analytics predictive and then prescriptive it’s like getting

37:28 prescriptive it’s like getting organizations to start saying what do you recommend a good example of this would be is descriptive would be how much to your point Tommy what you were talking about I think a lot of times businesses argue about what are the descriptive analytics that we care about so we can make decisions what do we look at historically what is important what is actually driving our business and I think as we get further up here when we start talking about predictive and prescriptive analytics prescrip prescriptive would be doing the basket anal is and going to Google I’m going to

37:59 anal is and going to Google I’m going to Amazon and then Amazon’s recommending hey here’s similar products that other people bought that you might like to buy as well and they may not even be anything close to the existing product you you have on the store you may be buying a a a toy for your kid or something like that and there’s other products that are totally different it’s like here’s a beard trimmer that has nothing to do with like your kids toys but it knows like your profile what you’ve built you’ve bought historically and it’s saying here are some other things you might need it’s it’s it’s not even recommended it based on the actual

38:29 even recommended it based on the actual product you selected which I think is amazing but that’s to me a good example of prescriptive analytics it’s trying to say here’s what we think you should buy I don’t know if you can do this in all situations but again I feel like a lot of companies still are struggling with their their existing insights and their hindsight or the descriptive analytics portion of this of this curve yeah I think I think the the the chuckle I get out of that one is the I just I just want people in the first step just get

38:56 want people in the first step just get get out get people off the data dump yes right like not even there there’s be one before this data dump the same the same right like data dump like get it into the pivot table you guys do pivot tables this is the same thing right yeah this we can always go back to the granny line the the other thought I had that I want to lean into and ask you guys question right there Mike when you were describing you discussing this with the the business and they outlined that limited number of trucks we want them

39:26 limited number of trucks we want them all full here’s the business case right here’s what we’re trying to solve yep within this a bi developer is faced with or I think should be and I try to challenge my team to do this too is it this creative process right what what do you guys think well let me let me flush out the whole the whole thing there is this creative process right that a b Dev needs to go through here’s a business need here’s a data set you may get

39:58 need here’s a data set you may get some basic requirements but go build something meaningful like you should have discussions with the stakeholder you need to understand right we talk about communication be empathetic do you understand like what they’re trying to solve in their daily work when you guys are going through this what are what are some questions you ask you ask yourselves around how do you build that report page right like we we have a lot of these pieces you can do with sqlbi and and create this really

40:28 sqlbi and and create this really structured rigid methodology of like how do you go build things whether or not everybody adopts every single time you do that there’s this this point of how do I place things on the page what am what visual am I building how do I I present this this business need in the best way like what are some questions you guys ask yourselves that help like take this raw information and it’s not

40:58 take this raw information and it’s not it’s curated right and and make it make it meaningful to the business I just go right to the themes. powerbi tips and just download the first most pretty Gallery that’s it I’m done I after that because we all assume that we already knew that so yeah right which how do you you use that tool after that obviously but then like which which data points are you putting in in the particular places and so I’ve done this a couple different ways and I I’ll describe a bit of my process here around being creative and trying to get the

41:28 being creative and trying to get the business to come on board with what we’re talking about visually I’ve done it two different ways I’ve built a report page with what I think are their needs so one is there you’re you’re building this this bridge I think between and this is where SQL bi process I think does make a lot of sense you’re building the model from one side of things here’s the measures these are the tables this is how we have to structure the information in some situations you you shape the model differently based on what you want to show inside the visuals so I think for some of these more descriptive

41:58 descriptive analytics I can build tables I can build dimensions and facts and we can talk about the concept of how your company data stitches together and we can start producing tables of information so there’s a couple pages in a report that are here’s what you’re asking for here’s the tables of data you care about a matrix a table whatever then I think I like to I like to spend some additional time thinking about Okay let’s listen to the business tell me what actions and I actually tried to prompt the business for what do you need to do to make your

42:28 for what do you need to do to make your team successful what keeps you and Tommy you had this question I love using it what keeps you up at night what are the things that keep you up at night that that you think about and then I build additional pages that try to say okay here’s the question I’m trying to say I need to increase sales per marketing spend whatever that is right so I need to tie this marketing spend to what what increase in sales and then I think about on that page what visual or what creative visuals can I put on this page that it

42:59 visuals can I put on this page that it start drawing correlation between this spend and the impact of that spend towards more visits more sales more something that’s creating action and so those to me there’s like to me there’s like a little bit of table Stakes here right you have to have the data and you have to have some visuals there that people can look at but eventually you have to get to a point where you’re starting to push the boundaries of people’s minds a little bit and being a bit more creative to bring in these bubble charts or extra things and users may not get there initially but at least you’re having the

43:30 initially but at least you’re having the conversation around what is what visuals are really driving action for you so I I’ve done that in a Power report by making one page per visual or a topic or a question per visual where I maybe have three or four visuals on a page and the questions at the top and I explain here’s what I’m thinking with the data and then the other thing I use is I’ll go to figma or some design tool and then I’ll just talk in concept we’ll just talk conceptually what do you want to build what’s the flow of the report Steve and I did a really good YouTube video around this

44:00 really good YouTube video around this one about talking about how to do report design and like okay we want like a narration approach we want an overview we want some details like what are the details just being able to quantify that out from the users I think also helps you design a better report so anyway those are the two approaches that I use Tom yeah so the what keeps you up at night one was my attempt because we’re in a we’re not in the most authorative position usually in these situations because someone’s

44:30 in these situations because someone’s coming to us and they say I want sales I want leads and build a report and it can be very you need to be treading carefully before you go all right well that’s not really what you want what we’re actually going to do tell me like what’s concerning you because they’re like no you just built a report dude so it’s also the relationship here too it would be great if we could go in like if we were consultant to everybody and that everyone was waiting on our every word to see what insights we had but that’s

45:02 to see what insights we had but that’s not always the case so to your point Mike I start off regardless if it’s a client added forever or someone new where that first question like okay what would keep you up at night if we’re not getting anywhere just you get to this point where we begin to and so many words calibrate people they think they just want leads and they want to see that on a monthly basis but that’s not what it would stress them out on a Friday night if something to happen let’s find that and I’m trying to still find a better way to say this but I I

45:33 find a better way to say this but I I want to know if this number were to go up how L like or if the number was going to go up or down at what point would you get hired or promoted at what point would you get fired that’s the name of the game and again there’s a better way to say that but there are and again those pain points on both sides of doing greater doing worse so let’s begin to start there and rather than being all negative one of the best I think things I have in the tool belt is let’s say we were on a

46:04 tool belt is let’s say we were on a building report let’s say and I I’ve used this before on the podcast but say we had $1 million and 100 people working on just this project right it doesn’t have to be a report what would it do and that to me I have found the best gold from people where they weren’t sure how to explain what they wanted for the data and really we find in those situations like okay it may not do all that but I can do pretty good with it because I’ll give you a really

46:35 with it because I’ll give you a really good example was working with a one of a sales team and there are seven sales teams at the company and it’s like okay now we’re on the seventh one and they’re like well we want to see what everyone else does like okay so I asked that question all of a sudden you you started hearing all the rep’s talks like well I know you can’t do it but it would be great if we saw you but it would be great if we saw who to contact based on like if know who to contact based on like if they’re three days out from not being a gold member anymore thing and it

47:06 gold member anymore thing and it was like that’s not a met that’s not a sales number thing but it was just more like almost behaving like an app and I’m like and I’m going through my head I can do that in da like there’s a Dax there’s a Dax way to do that I’m like we can build that and this mindblowing because when people are asking you for data again they’re expecting to get a bar chart they’re not expecting this intimate conversation you’re going to have however there is something that they want and just really that step here I try not to bring up bar charts or the

47:39 try not to bring up bar charts or the any functionality at all at this point if you can do that because again there’s a very ideal of people are also willing to play the game with you too however however however it’s very hard if you don’t have someone on the other end who’s not willing to at least Le open up or have those conversations because to me and I’m not going to do a hot take on this yet but I I’ll pause there but I I think that’s really the preferred

48:09 I think that’s really the preferred approach my take on this is this right like typically regardless of whether or not I build it in these ways I I’m Mo and I think I’m thinking about like a report in total as opposed to a specific case like you were talking about Mike like I like I yeah I I always start with the kind kind yeah I I always start with the kind what we we started with a long time of what we we started with a long time ago there there’s a summary right there’s key metrics am I being held to something and I need to see that

48:40 something and I need to see that progress on on a page right my measure sometimes yes sometimes no right do I need a summary page or a kpis page or something that tells me like boom here’s my numbers for the day I’m moving on then I I think okay great do I need that page or not then I have an analysis page and typically this is the page where I’m building some visualizations maybe some matrixes or some detail but it’s the page where I

49:10 some detail but it’s the page where I can use all of these things to dive into the data a little bit more and then a Details page and more often than not that’s just a table of information and I play off these things right like I as I’m thinking about driving an experience I I have to and I think somebody in the chat said this too like I I think one of the most important things to nail the creative like process part of this to give something valuable is to sit in the

49:41 give something valuable is to sit in the chair as the stakeholder and it does not matter whether or not somebody’s giving you requirements and saying go build this or you have to like think all of this up for yourself always sit in in the in the chair as the stakeholder and think about how am I going to do what I need need to in their role to make decisions based on what it is we’re we’re trying to solve and this this this does two things one is it automatically will generate

50:13 one is it automatically will generate other ideas for different reports or different pages probably right like oh well we’re we’re about this scenario but this I know this user doesn’t have this or this or this and those are conversation points for another time but also something that I’m gleaning as I’m sitting in this role trying to like solve whatever I’m I’m building for them so I’m looking for those visuals that allow them to quickly digest information whether that’s on summary or analysis

50:43 whether that’s on summary or analysis but also use these visualizations in the context that I want them to use them as filters on the page right one of the greatest things about powerbi is like oh that’s an interesting bar like click everything else filters right oh okay well that did this to this visualization like shift click hold that right control click like all the interactions that I can hold and retain the parts of the data that I’m finding as I’m going back and forth between these visualizations

51:13 and forth between these visualizations and that’s why sometimes it’s really like is it necessary to have three bar charts and they’re producing information in different ways well no but if I click on one of those bars and the other ones change in an interesting way now I can start to segment data so I’m I’m looking for opportunities to make visualizations play off each other to produce something I otherwise wouldn’t potentially see in the data set and that’s why visualizations are important because they they key you into things that

51:44 they they key you into things that you’re innately not just going to going to understand or look at from a table so the and then the last point is i i as I’m building I as the report Creator I have no limits right I have access to the whole data model but I have to think in terms of what I’m limiting the end user in doing because I’m only producing so much information from the data set

52:14 so much information from the data set typically so I always think about how am I how do I have to interact with this data what other things am I going to want to slice and dice by because that would drive what additional what additional columns and things I put in filters right or slicers just as add-ons because it’s part of the business like as I’m interacting of course I want to know who the owner of this is it’s just not relevant completely to this page but I it’s it’s in my org right like I need to know territories and that’s like over here oh

52:45 territories and that’s like over here oh hey it would be really interesting maybe we should talk about another page that’s all territory driven and blah blah blah blah right so I think keying in on like this empathetic role sit as the stakeholder but I’m also trying to take what the the tools of powerbi allow me to do to create this experience that I know is valuable right because I have my own data sets where I interact in this way and by taking that

53:16 interact in this way and by taking that I think you you drive a lot more of that business value and that narrative into that production light for the end user I’m I’m literally taking multiple thoughts down here and writing them down to paper here it’s probably about time we’re right around time right now on so let’s maybe do some final thoughts here and we’ll wrap it a little bit but Seth would you would you quantify that your final thoughts you want to jump back in for some my final I can quantify my final thoughts

53:46 final I can quantify my final thoughts good good good topic on that one so I want to respond to your comments there Seth I think you do a great job of summarizing a lot of what we’re talking about here for creativity it feels like creativity in report design flies in the face of what I’ll use there’s this there’s this term called self-surface are you doing exploratory analysis of your data sets or are you doing explanatory analysis of your reports an explanatory report seems to be a bit more designed for a specific purpose the exploratory seems to fall

54:18 purpose the exploratory seems to fall more in this self-service space so for me I feel like explanatory reports are a little bit less creative you have a little less Liberty to get through the creative portion inside the explor reports not say it doesn’t happen but it feels like there’s better requirements around what do I need to do to run my business in a daily way and it doesn’t need to be it probably isn’t a fancy report a lot of times I use bar charts and that seems to do a lot of the needs of what I need met so I think in that respect there’s there’s this balance balance between using self-service not

54:50 between using self-service not using self-service getting handed reports and having someone design them and give them to you with with clear requirements the last one I like I really like your your point Tommy your last point and I really want to emphasize this it’s going to be very important for you I think as a report developer to step back and give the business freedom to explore what they really think they need I I really like that idea of like okay imagine no budget imagine no holb I think sometimes we focus too much on the details in the weeds and to your point set the example of well in this one

55:21 set the example of well in this one example of this customer at this point in time we had these and you go down this long Rabbit Trail of like this very potentially narrow thing or narrow thinking I think stepping back or at least giving the business the flexibility to step back and say big picture unlimited money unlimited budget what do we do that’s that maybe is something we can do faster now with these new tools and so I think I think the light the idea of that or at least allowing the business to think that way you may may not be able to deliver that day one but at least you can start working towards it and that should be your bigger Vision on where you

55:52 your bigger Vision on where you go with reporting so those are my two final thoughts maybe take a step look back at the big picture and really evaluate what your reports are doing for exploratory versus explanatory reporting and maybe those are going to allow you to have more or less creativity based on those two patterns in report building Tommy what what else what your final thoughts yeah so I’m not gonna end on a complete hot take but a slightly hot take so we we’ll see how it goes medium take medium take but honestly if you’re doing it

56:23 take but honestly if you’re doing it right and you’re in the right place with the right people and you have the ability to also have the skill of Discovery 80% of the requests that you get should be drastically different at the end than when the ticket came in I agree got it report that’s a constant issue completely different than what they asked for the requirements change faster than we can get them done in the first time the first place so I

56:53 in the first time the first place so I like like it just reminds me of something that just just happened yesterday was like where one of my devs was like what what do I what do I do cuz obviously my estimate blew out of the water because of this exact scenario it’s like just had a comment man like we gotta get in this case we got to get this one done requirements change like we got sh ship something and then we’ll come back and fix it later that’s a phase two I do a lot of I do a lot of that’s a phase two requirement let’s let’s let’s move that to phase two as we

57:23 let’s let’s move that to phase two as we slaughtered the end of the show yeah is it phase two or is it’s something that’s core and we need to get it out anyway yeah I think I’m pretty sure we need to do a phase two on this this talking conversation all right awesome with that thank you very much for listening we appreciate your ears today thank you so much for joining us on the podcast I hope some of this Insight around being creative what you can and cannot do pushing on the business getting more thought around where and how you build reports for your team this help you with your conversation with that Tommy where else can you find the podcast you can find us in apple and

57:54 podcast you can find us in apple and Spotify or wherever you podcast make sure to subscribe and leave a rating it helps out a ton do you have a question an idea or topic or even maybe a awesome story that you had that you want us to talk about in a future episode head over to power. tisp podcast leave your name and your great question guys you’ve been submitting a ton of mailbags keep it up and finally join us live every Tuesday and Thursday a. Central and join us on all of power bi. tips social media channels thank you so much we’ll see you next time appreciate it

Thank You

Thanks for listening to the Explicit Measures Podcast. If you enjoyed this episode, please share it with a friend or colleague, leave a rating/review, and subscribe so you don’t miss what’s next.