Data Products – NOT Projects – Ep. 269

In Ep. 269, the crew digs into a deceptively simple mindset shift: build data products, not data projects.

The difference isn’t semantics—it’s operational. A project has a finish line. A product has an owner, users, feedback, and a backlog.

Using Umesh Barat’s Data Product Operating Model as the spark, they talk through what to borrow from product/software teams (iteration, prioritization, stakeholder alignment) and where BI teams often need a hybrid approach—because ‘help desk’ reporting demand doesn’t behave like a single, well-scoped engineering build.

News & Announcements

• Azure Diagrams — A free Azure-focused diagramming tool for sharing architecture visuals (with community examples) without hunting for icon packs.
• Time travelling through data: the magic of table clones — Table clones in Fabric are a practical way to test changes safely and compare outcomes without duplicating the entire pipeline.
• Data Product Operating Model — The article behind this episode’s core question: what does ‘data as a product’ look like when you’re trying to run it long-term?
• Submit a topic idea (Explicit Measures Podcast) — Send a real scenario from your org and the crew will break down tradeoffs and patterns on a future episode.
• PowerBI.tips Podcast — Subscribe and browse the full back catalog of Explicit Measures episodes.
• Power BI Theme Generator (Tips+) — Build consistent report theming fast (and keep a single source of truth for colors/fonts).

Main Discussion

Calling something a data product forces you to answer questions most teams skip when they’re heads-down building a report: Who owns this? Who uses it? How do we measure success? How do changes get prioritized?

The crew agrees there’s overlap with software/product delivery (requirements → build → test → feedback → refine), but BI teams usually live in two worlds at once:

• Big, requirement-heavy platform work (new sources, identity unification, migrations, governance changes) where Scrum/Agile structure can prevent expensive misfires.

• High-volume ‘ad hoc’ / help-desk demand (new views, small tweaks, evolving questions) where rigid sprint boxes can create bottlenecks and frustration.

Key takeaways:

• Plan for the post-publish phase. Adoption, training, and iterative improvements are part of delivery—not optional cleanup.

• Make ownership explicit. Without a named owner (business + technical), a report becomes ‘everyone’s problem’ and then ‘no one’s priority.’

• Ship something usable early. The fastest way to uncover misunderstanding is to put a test version in front of real stakeholders.

• Separate intake streams. Run big platform initiatives with disciplined scoping, and handle smaller reporting/metric requests with a transparent queue and clear SLAs.

• Prioritization must be visible to the business. Transparency beats surprise; it reduces thrash and makes tradeoffs real.

• A data product is bigger than a report. The model, definitions, and ongoing change process are the actual asset.

• Governance is part of the product conversation. New capability (like Fabric features or calculation layers) raises the question: can you monitor and manage it at scale?

Looking Forward

Pick one ‘high-traffic’ dataset/report and run it like a product for 30 days: name an owner, set a feedback cadence, and maintain a small backlog you actually revisit.

Episode Transcript

0:25 good morning and welcome back to the explicit measures podcast with Tommy Seth and Mike good morning everyone I’m gonna oh sorry okay you can go ahead do the morning do the morning thing it’s all right problem waiting we are jumping in today with our topic today our topic is around our an article from medium UMES is talking about data product operating model how do you use data in an operational way what does this look like and how will you maintain this over time so that’s our topic for today let’s

0:56 time so that’s our topic for today let’s do the intros good morning Seth and it’s exciting I’m so excited excited to talk I guess happy Tuesday gentlemen there we go okay there it is let’s pick up the news let’s go across the news real quick here so One news bit that came out recently was something that I I think I came across a tweet on Twitter somewhere or I guess you have to say x formerly known as Twitter now is that is that the new I think so so a tweet came out on X that was talking about this thing called Azure diagrams and the link is in the

1:26 Azure diagrams and the link is in the description or the title of this video already so you can see Azure diagrams for yourself it’s a free platform it’s basically for Designing architectures of Microsoft and the really interesting part is you can build these it’s all the Microsoft components here’s how you get a one Lake here’s where a blob storage account make is occurring you can use little arrows and they can link together and show you data flows really interesting stuff great tool and it’s also sounds like they’re trying to make a community of diagrams so you can

1:57 a community of diagrams so you can create a diagram write your language around what it is and and describe it and detail it and then you can share it which is really neat so then other people can take it from there and potentially utilize it and change it tweak it for their needs great way to articulate a architecture diagram very cool tool Tommy have you been playing with this one yeah man there’s so many diagram software first off though now that you said it out loud I think I would prefer you say we found it on Twitter found on X I’m fine changing to

2:27 Twitter found on X I’m fine changing to semantic models that termin yeah too too bad some you like some you don’t man y you got to take it all all nothing you don’t get to choose the name of the company somebody else does yeah I don’t apparent I don’t have the budget to buy X to change it back but anyways anyways just wanted to put that out there there’s so many diagram softwares in so many places you see diagrams around what we do it’s nice that there’s something actually in Azure

2:58 that there’s something actually in Azure about azure from a diagram point of view it’s it’s great because has links to to documentation it’s pretty straightforward it’s just the architecture it’s not some of the different flows I personally would love to see all the end to end scenarios that Microsoft’s doing their documentation building out the diagrams through the Azure diagrams I would imagine they’re going to do more of that that would be my initial thought is they’re just going to keep using this tool and this will be their new tool for

3:28 this will be their new tool for hopefully doing all their architectures and then it’ll be like an easy here’s a one button link share of the architecture that we built it’s got some interesting things to it the the group is for Microsoft but it’s not a Microsoft funded project that’s what I’m that’s my interpretation as well it sounds like it’s three Engineers from Microsoft that said hey we could build this this makes sense we’ll makes a ton of sense to Tommy’s point though like Vio Lucid chart like Vio is a Microsoft product but like if yeah there there are libraries of icons and all the

4:01 are libraries of icons and all the things that are out of date or you have to go grab online and so if this is a way where it’s like hey it’s always going to be maintained and pull the latest stuff out of Microsoft and like you’re always I’ll use it like right A lot of my system is in Microsoft like okay if you’re Microsoft shop this makes sense yep excellent good very good tool would love highly recommended go play with it it’s free to use you can even sign in and when you sign into it you can then save your diagrams you can then hang on

4:31 save your diagrams you can then hang on to them for later I believe you can also get examples or work off of things off the community and then you can continue to modify them or tweak them for your needs as well so a neat idea one this is this tool is using something that I think is a core value that Microsoft should lean into more Microsoft should be focusing on building tools that are Community focused so build the infrastructure for it but think of a way that how can I involve a community around a new feature let me

5:01 community around a new feature let me give you one example have you seen some of the video have you seen the demos around inv visual calculations from Jay Jay’s now talking about this inside user groups have you seen those okay haven’t seen any the demos you got to watch the demos it’s actually pretty interesting what they’re doing with this so I really I’m I’m liking the feature it it it does come with its layer of complications and Tommy to your point around governance and administration in visual CS add another whole ball of wax to can we Monitor and govern these new calculations that are

5:32 govern these new calculations that are being that are being made inside the context of the visual and it’s supposed to be doing simple things like hey look you’ve made a visual and I want a simple running total right okay that that makes sense but that calculation to make the running total doesn’t live in your data model Now it only lives inside the visual so you it’s not in the semantic model you can’t go get it from analyzing Excel just in the visual so it’s it’s that I don’t know if I love it yet and I

6:04 that I don’t know if I love it yet and I understand why they can’t push those things into the semantic model makes sense because it’s it’s using the semantic bottle but it’s not part of it anyways all that to say is they have this ability of using these quote unquote templates to make calculations in the visual and I I was like guys this is the same thing you made with quick measures quick measures was this potentially amazing thing where people could make their own quick measures and build whatever they want and then you could have hundreds of these things company standards around quick measures

6:34 company standards around quick measures but they never opened it up it never became a community thing and I’m like you guys missed the boat on that one like people would have loved to have pre-built Dack stuff and had they done it I think you wouldn’t be having all these challenges around wise Dax so hard you’d be like oh I’ll just go to this gallery and I pick out the things I like so of course you still have the challenges around Dax we know no no no no it would be a little simpler sure but it’s always about your data model we know this well it’s about learning how to use Dax that’s the problem I feel like and when you watch other people so

7:05 like and when you watch other people so that’s how you learned VBA that’s how I learned like writing macros in Excel like I would record what I was doing and then I would watch the output of the code and then it would then I would start learning with oh okay that’s how I would do that and then with a little bit of Googling I would fill in the gaps of some like that’s what we want Dax to do right build me some generated or other templates across the internet around Dax let me understand how they’re working and as I see examples over and over and over again some of that will start I’ll start educating myself and some people just don’t care and they just keep copying off the things on the internet

7:36 copying off the things on the internet I’m personally finding a bit insulting that you’re equating VBA to Dax but I’ll let that slide just programming languages so you’re so generous today Tommy so many things slide yes yeah that’s fine what else did you find on X Mike well another one we found here yes let’s let’s move on then another article came out Tommy and this think this is that you pointed out here Tommy there’s another article around time traveling through data and Microsoft calls it the magic of table

8:06 Microsoft calls it the magic of table clones wow wow I felt like this felt to me this feature seems like a bunch of marketing to me at this point I’m not sure I didn’t quite understand the use case of this one it’s the idea that you can go take a table from your Lakehouse and you can clone the table and then it it keeps the time travel features said table when you’re cloning it okay I’m not

8:36 okay I’m not sure Delta has this the Delta tables has this concept of versioning of tables I can’t tell you the number of times I’ve actually used going back in time to an earlier version of a table because I screwed something up it it very rarely am I going back in versions to see what the data look like is is it versioning or is it carving out historical data like that doesn’t need to part ofate at a point in time I creating a duplicate

9:09 a point in time I creating a duplicate at a point in time so to me this speaks to if you’re using in the Delta format the Delta format is just creating Bunches of tables already by default every time you edit the table either with like an insert or delete or any modification to the table you’re creating a whole library of new parket files behind the scenes that you’re able to then go look at and then what happens is you get this log file that just says hey I’m on version one okay we made a change version two so to me it looked like what’s what’s the big Point here I

9:40 like what’s what’s the big Point here I didn’t understand hey I I’m just I’m just ripping through real quick like I suppose table clones facilitate development and testing processes by creating copies of tables in lower environments well that I could see a bit more helpful hey I can I can go to production and then I can clone that table get a readon copy of it some somewhere else and then do my data work on it and then I could probably so so here here are like four points of table clones I guess from a value prop perspective I’m just ripping out of the

10:11 perspective I’m just ripping out of the Microsoft go ahead does it say t table clones provide consistent reporting and zeroc copy duplication of data sets for analytical workloads and machine learning modeling and testing okay Full Table clones provide capability of data recovery in the event of corruption but you have to M do this create historical reports that reflect state of data as it existed at a specific point in time I don’t think I would use versions of that in my opinion I think I

10:41 versions of that in my opinion I think I would build a specific table to have all the data in it as opposed to trying to rely on the versioning of tables to see its state over time and you have to yeah when you hear time traveling in your title you thinking some Marty McFly type of feature definitely yeah of course miles an hour yeah but it it really is the the only problem with that I think being more of a production thing is you have to mainly clone each table at the time that you needed to be cloned so you can’t go back like if I want to clone

11:12 can’t go back like if I want to clone something from last week too bad so I have to say Okay this week or this Monday I’m going to clone these seven tables and now I least have that historical feature there’s I think you can still do this in tsql through code yes you can you could probably write the you can write the autom in fabric they’re simply giving you the uiux feature to do the the

11:35 you the uiux feature to do the the wording of the SQL statement is create table as clone of and then you write and you write the name of the table which is available in fabric like the everything in the data warehouse but you can now utilize the UI to do this as well so to me it the only thing I can see that I would think initially immediate value is I need to copy things from production into lower environments so I could read them figure out what my data problem is work on top of them

12:07 data problem is work on top of them build my engineering code around how I’m going to fix this pipeline whatever that thing may be right if we have found a problem and then I can roll the the code or the the deployment through the different environments and not have the table copies and it’s named the exact same thing right so it’s important if you’re having like notebooks that are accessing this database this table because you need to have that table referenced by name in each of the environments you’re not going to want to do like table oneor Dev because then you’d have to physically change your code As you move through the different environments it was interesting I’ll see

12:38 environments it was interesting I’ll see if I can actually find some better use cases with this one this it’s a neat feature I do like the fact that they’re giving the the UI version of the name so that it definitely makes sense that they have a UI that’s going to help you with this a little bit but still not thoroughly convinced this is going to be amazing yet proof is in the pudding I’m going to call it a goals type feature oh we are spicy today we spicing it up all right enough of our news let’s jump into the main topic for today so our main topic is from Umesh Barat

13:12 main topic is from Umesh Barat Barat I think that’s how you’d say his name sorry if I butcher your name but we’re talking about the data product operating model data as a product all right jump in initial thoughts Tommy you got your mouth open you’re thinking you’re thinking there’s something there either you’re yawning or you got something to say ya yawning oh okay you get you get called on anyway yeah you you’re gonna get called on what’s your initial initial thoughts I hopefully don’t disagree with you guys too much today but it’s this

13:42 you guys too much today but it’s this idea of taking this just a brief summary of the article it’s just taking the idea of what we normally consider our data projects we’re going to call them products and put it more into the agile scrup method and what normally goes on to if you’re developing a software developing some a website which is usually known as a product of that not necessarily the dev test prod but how does a project actually get cultivated Who’s involved in that process and including like a project

14:12 process and including like a project manager and how fast it should go and what what is a methodology to actually create data products in this case so talking about moving faster talking about in terms of how like you utilizing stories and epics my initial thoughts from here is I’ve tried this this is some this is not a new not new idea for me at least back in the day when there were no

14:42 least back in the day when there were no projects around bi when bi really or at least powerbi first came out we tried AEL scrum we tried the two we Sprints and for for different reasons and different scenarios we found pros and a but a lot more cons and trying to take business intelligence and then try to fit it into another type of methodology or project methodology which didn’t seem to work but that’s where all these start off oh I’m gonna love this

15:12 off oh I’m gonna love this conversation as as as soon as you said as soon as you said scrum and agile doesn’t work for data projects I think you just red up Seth right there no no no I what what that that’s that’s some of what is being described here in the article as well right like that revisiting the scrum method scrum methodology and umash is like yeah it’s a it’s a well-written article I think I think like from data product for you think like from data product for versus project right I I think I

15:43 know versus project right I I think I agree with that right like the the things that we build are very similar in Focus that engineering groups and development groups in other areas for applications Etc like there’s a lot of crossover like there we have sources of information there’s a bunch of ways we need to transform it there’s a product that we’re producing related to a report a whether that’s powerbi report or a set of data whatever the case may be like there there are a lot of crossovers in there and I think we’ve talked about that at at

16:15 think we’ve talked about that at at nauseum right like there’s there’s a fine balance here between it and business and some of the structures of how we go about building these products is important so what I I don’t I don’t know where like Mike do you want to I got some initial intro okay yeah just I’ll just just a thing while you collect some thoughts here there’s a diagram in the middle of the article there actually there’s two diagrams there’s one talking about a process start here build requirements

16:47 process start here build requirements develop product get a test version out manage that get feedback and go back and add or refine requirements I think his initial product development cycle is a really good diagram to think through and you need to I think this is actually very when I’ve built data products or projects or tables or things this is really very relevant get an initial build out quick get people looking at it and always in every data project I’ve ever been on there’s always some I don’t I don’t understand that

17:17 some I don’t I don’t understand that this doesn’t make sense these numbers don’t seem right and then the it’s not that the requir initially were wrong it was people’s understanding of what the data actually was doing they didn’t really understand what’s in their data and almost every time we come back to huh there’s some data quality problems now we need to add some business rules into this process or some cleaning or something like that or it becomes a series of action items that people Upstream need to go fix their Data

17:48 Upstream need to go fix their Data before it comes into our process because we don’t want to build business rules for the sake of doing that because someone else didn’t do their job and clean their data so this management piece on top of your data engineering product or process whatever you want to call it needs to be there and then my second thought is there’s a second diagram that talks about the tooling that encompasses all the different people and I think he dialed it in right I think there are people that play the role of the data engineer there’s people that play the role of the business analyst and if we’re not

18:19 business analyst and if we’re not thinking about it there should always be some layer of stakeholder capability in this mix as well so I think the mix of those three people while that may be one or two people the roles in which they play is very relevant and you need someone representing all those areas or you need input in all those areas to actually produce a good data product the stakeholders have to stand behind sign off on it right the business analyst has to say well what is the business actually going to do with the data as opposed to just give me a table you say

18:50 opposed to just give me a table you say just give me a table it’s minimal impact on the product and then the data Engineers have to sit down there and listen to I’m going to physically work through the technical challenges of the data and say but you can’t join it this way it doesn’t it doesn’t match there’s no there’s no keys there’s how do you want to join it together oh and then the business analyst figures out okay well we got to work through that that challenge that issue and and redesign the system so it actually works so I like the the idea or the concept all of it I think later on in the article it

19:20 it I think later on in the article it gets a little bit more interesting where he starts talking a bit more about scrum and agile and how that works and I’m I’m not sure I have an hanging on that one per se there’s some thoughts but I’m not sure if I lean one way or the other on that on that side then let’s start on that the first area that you’re talking about because this goes back to our episode 264 the low user adoption rates and right and we had a huge conversation around when does the project really end does it end when I published and it’s in the app sure or

19:50 published and it’s in the app sure or but and we I think we all came to the conclusion really shouldn’t and we really shouldn’t say done move on and so there’s a lot of that crossover here which I I I completely the more since that episode I thought about that the more that needs to be part I think of our own process of of really when we publish a report there’s really a second phase that needs to begin on the adoption and the user feedback so I want to ask you a question around that your point there Tommy I love your point are there data projects that you’ve done

20:21 there data projects that you’ve done that you that you have grabbed some information put it into report and you’ve totally walked away from it and not made any major adjustments just personal stuff like stuff you’re getting data around like we have the podcast we do analytics for Google and things there are there any things that any projects you have in your mind that you’ve done that you build it and you walk away and you you use that report over and over again without many adjustments never never never yeah have you done that yeah and

20:51 never yeah have you done that yeah and so where where I’m where I’m going is maybe what this question is I have a very few few amount of projects that I can say I built something I made a data model and I walked away and I regularly use that report that same report without many changes regularly using it well yeah that’s the point you can build it more away and come back but like it only makes sense if you go back to the report often and say Here’s a report that I built and I and I need the data in it

21:21 built and I and I need the data in it and it’s designed the way I want to use it over and over again this is such a good point like the our actually our podcast where I get the data as a great example because I curated the data you example because I curated the data I was get from all the sources and know I was get from all the sources and built it and then I published it then I’ll take a look at it go I really wish I could see this and I’m like okay I gotta go back and this is for me this is something this is our data and and that’s what that’s my point is my point is I think the needs like when you initially look at your data you have a

21:52 initially look at your data you have a perspective on what that data looks like and I think as you over time look at that same report the reason you go back to that report is you go hm I think I want to see it a slightly different way or I have a net new question that was not previously answered your your needs are evolving talk about it at all the time like the a business even when we develop a report for them especially if it’s something that you’re using on a day-to- DAT basis absolutely the there

22:22 day-to- DAT basis absolutely the there are going to be improvements or enhancements that can be made and and maybe and maybe that’s more of my point here around this the latter half of this article is this concept of like you can build a report you can push it out but to assume that the initial needs of that report are consistent and they stay the same for a long period of time probably a misnomer and the questions will evolve so how do you build for that maybe let me carry on your question real quick because I think this

22:52 question real quick because I think this is going to be a a pretty big distinction too on what type of requirements are we asking on those type of projects that you go back to are you asking yourself man I wish this was a bar chart or are you looking at something like I want to see this in measure this a different way or maybe it’s like a two week rolling which one is it usually right the latter more so

23:10 is it usually right the latter more so than the former yeah and maybe it’s going into a bar chart with whatever you’re going to build but you’re either looking something going what’s the two we rolling or what’s that compared to everything else which is it now going to be a different calculation to me that’s where always the tinkering is you’re gon you’re gonna say something too Seth well yeah your your question was how do you build how do you build towards or or build for the evolution of a

23:40 a report and I my my initial answer would be by getting by setting a solid base of a first a first report and that come and in many cases where I think that is skipped without process or too fast of an iteration it causes many more issues down the road because if we don’t like especially for some reports right if you don’t understand what the scope of the report

24:10 understand what the scope of the report is the intent what are the things we’re supposed to be looking at here then the end and I’m not properly communicating that right like that’s one of the things we talk about a lot end user even in that low user adoption rate things like if they don’t know what it’s supposed to be producing then they’re going to have their own expectations there’s a base level scope I can provide in in the time frame that we have to develop this product for you and then we can move on from there

24:41 you and then we can move on from there what I what I find what I’ll find interesting in this conversation Tommy with your different experiences and where you actually land right now is I do have I have seen that a bit of structure in developing these things is absolutely Paramount to have the right thing to start from MH so the yeah we can the the actual structure I by no means disagree with

25:12 structure I by no means disagree with you on that in terms of for the project development yeah it’s not just a okay when a ticket comes in we’re going to work on it because that be automatically starts a lot of bottlenecks and a lot of frustration around I’ll let you guys dictate where do you want to go from here I have I have another Point around Seth’s comment that I really I lik your I likeed your comment Seth and I’m thinking through like okay let’s imagine okay there’s I’m thinking the scientific method here to some degree right how can we take out some of the variables in

25:42 we take out some of the variables in this process what are some variables I can remove like for ex let me speak to some of my experience right I I I look at all of my Consultants on my team and I need to understand what they’re working on when they’re working on things so I have a report that I’ve built for them around like here’s what you’re working on here’s your here’s your status for me I also have a report that says here’s here’s my team here’s what everyone’s doing so I can understand like what we’re working on and then while I’m making adjustments on HM we’re working too much on these projects we have some due dates or

26:13 projects we have some due dates or deadlines we need to shift the focus here a little bit this is a report I look at almost daily now and so because of that I had very I have very clear questions up front and I have very clear actions I take now looking at the report and I know what I want to do and I’ve got I’ve been using the report enough to say I understand what I want at a high level the aggregated form of data and I know what I want to go down into or what details I need to use report so everything there is a

26:45 everything there is a repeatable process or a repeatable question I’m asking over and over again so this data set or this report again I’m I’m trying to be a little bit scientific here this is not going to somebody else this is just me I’m looking looking at this report as only I need it so there’s no education needed on how it works what buttons were there why the visuals are there like there’s no educ that’s a whole another ball of wax I think that is as you build reports for other people there needs to be an education pattern that’s going on there Bar None because

27:15 that’s going on there Bar None because there’s there’s just not enough how do you say it there’s just not enough of that being done today so therefore people are given things and they don’t know how to them therefore they don’t they just don’t use them where I want to go back to here propensity to not use it is there right yeah correct what I want to go back to is the data model that supports this thing so behind this is a data model that I understand I know I I I built I created

27:45 created it I I’m beginning to think a data product is yes partly the report side and I have this single report that I’ve built a thin report it’s got a very defined scope now in my now in my mind the data model’s key and so what I’m doing now I’m finding myself doing now is I’m having new questions asked of the data and so as I’m asking these new questions to your point Tommy what does this look like compared to last period of time like some comparison things I’m thinking about like how do I performance

28:15 thinking about like how do I performance or optimize my process a little bit and in doing that I forced myself to now start to build other thin reports that are not part of the core report I am I’m I’m I’m physically seeing myself resist the imple the the the capability of oh I could just modify my report oh I could just modify no connect another thin report experiment with the questions I’m trying to ask put visuals and things on pages that may or may not make sense and then if I like that then

28:46 make sense and then if I like that then I can migrate them over to the core report and so as we think about this scrum or agile or even a report building process right potentially and again I’m going to introduce a concept here is maybe what we should be doing more often is here’s your standard report we don’t touch it if you have additional questions let’s do a little bit of side work here let’s do some more experimentation around here’s other ways of looking at your data building other visuals making different pages do these things work use them for a week quick

29:19 things work use them for a week quick fast prototypes try it does it add value to you come back and review and if it does add enough value then roll it into your standard thin reports or whatever you’re doing so I’m finding that works well for me I don’t know if that’ll work well in a bigger team where the emphasis to me I would put on this because I do like that and I do agree but I think the context there is a little different because that’s a single person you looking at a data set that you created I’m trying to focus on a very specific scenario but like could

29:51 a very specific scenario but like could scale out to like an entire team and lots of people potentially potentially but you’re also you’re at this point the consumer and the developer I know it’s it’s so it’s different because I have this whole and the stakeholder so you’re all but I am but all but you’re right all those things like you need all those like this is why I think you need more prototype thin reports on what you may want to build because you need all that feedback from all those different people it’s not just me internally

30:21 it’s not just me internally communicating like okay this is what I need this is my stakeholder perspective okay now I’m going to put my business analyst hat on about like that oh I could engine the data like so I can do all the roles myself which is cool right but in reality you’re you’re going to have potentially those roles broken across multiple people and now you got to figure out how do we all talk together get on the same page because you as a stakeholder also know as a developer what’s possible and what I can modify and how long will take the other the one emphasis I do want to put though is we were talking about the apps and

30:51 is we were talking about the apps and like app development website development I I don’t like putting reports or content in in the same category for one main reason when an app’s developed the people who are going to use it have to use that app like if I’m developing a business application for product management okay well that’s what everyone’s going to use there’s no other way around it a report does not have that same in a sense specification or prere right that report’s out there but it’s not required for you to do a task

31:22 it’s not required for you to do a task you may that may be your new sales report but that’s not always the case if I’m building a new application or a new input or form well that’s what the business now has to use in order to input X Y and Z which you’re going to get a lot of feedback because everyone there’s gonna be a lot more eyeballs on it because that’s what they’re using reports there needs to be an emphasis on that initial stage of we’ve just published this and we talked about this about the adoption about the promotion of that new content okay this is really important I

31:53 content okay this is really important I I think I I I don’t okay so you’re you’re saying that it’s not the same thing because a user ha like has to be taught how to use the report or like it’s it’s a thing in the ecosystem it becomes a thing in the ecosystem not now your new way or your new like Road or Avenue where an app is so I I can see where you would say that if they have access to the data in another

32:24 access to the data in another way but if this is the only way that they can get the data that’s certainly not true and then secondarily I would argue that if we’re going to go through deaf efforts we should treat it like an application that everybody should be using and do the appropriate things that we need to to ensure they use it otherwise you just wasted everybody’s time I agree so so that’s where I don’t agree with you like

32:55 that’s where I don’t agree with you like it absolutely should be looked at as something that we push out much in the same way that we would in an application like the only difference here is you don’t have the other teams that are on like uiux and like marketing and all like the product and pre like where they’re developing like the release notes and everything that’s going to be in there and like people have to use it and go to it every day you assume right but if they were doing their jobs they would be that no but in the same way like reports are just as valuable or

33:27 like reports are just as valuable or should be to provide information to people don’t I’d rather I’d rather say they are absolutely like extremely important because otherwise why am I wasting my time doing any of this I would agree but I think the difference is more there’s not accountability around consumers utilizing reports I think we’re talking the difference between desired and actual is what we’re hearing Tommy I think what you’re I feel like what you’re speaking to Tommy is what actually happens Seth I think you’re speaking to what we desire the

33:58 you’re speaking to what we desire the process to look like like we would all agree documented understood sent out and is it cert and I would add my little my little phrase is it certified right if it’s if it meets that certified criteria yes we do want to have those quality checks on things but where I where I’ll go back to Tommy to your point is the reports are just a commodity a thing and not necessarily an app is like a a process by which we have to like force people through to go use data I would

34:28 use data I would argue you’re right but I would also want to use the app the app is my distribution model and I want to have multile reports in there and I may even have like based on audiences or things like that why wouldn’t I publish multiple reports in there where we have

34:46 multiple reports in there where we have some of the standard reporting again it’s it’s hard to get down this is my point though when I said earlier it’s hard to get down to the single action I need need to do by looking at this data all the time what is that there are questions that occur only every so often and there are questions that occur all the all the time identifying what question you have and I’m literally teasing this out as we think here because I’m I’m putting these two Theses together if you’re looking at the questions I have all the time every day that should be like a standard part

35:18 day that should be like a standard part of the report when we have questions that are just hit or miss or random or debugging or those will be so much more ad hoc in nature anyways and those questions I think are best served by having an understood data model and letting people go pick through the data and go build what they want to build or here’s a Podge podge of visuals Andor tables that come out of the model maybe these will help you but they’re not answering that single question over and over and over again these are it’s the it’s this autogen these Rando

35:48 the it’s this autogen these Rando questions that are happening on the side would you agree I I to an extent where where for me it’s still more around a business application requires data input require that’s where everyone’s going Mike in all of your clients and Seth in all of your experience have you seen accountability around users not using a report or someone being penalized a team being penalized for not using a report there’s out that’s out there how how would that be different than application

36:20 would that be different than application I would say my my question would be my answer would be no and it would be usually it’s not penalize it’s more of like you don’t get access it’s EI it’s either like an onoff like it’s onoff like either either you have access through our thing because that’s the only way we’re going to give you access to this data I see that it’s more of like the the stick side of the carrot and stick portion use this because that’s the only we’re going to give it to you we’re gonna not let you going to go build your own stuff somewhere else now I’ve also seen very strong armed organizations around you

36:51 strong armed organizations around you only get this data model you can only read it and you can’t export it we’re only you only get it in our report and nothing’s exportable which could be totally fine and maybe that’s the right scenario for more of these repeatable commonly asked question type things maybe that’s where it goes right and how are you gonna get those questions without eyeballs without that that second stage and I think that’s where trying to get out people already have those questions not every not every report is going to be like the The Daily report that people have to go use but if we’re talking about one of

37:22 we’re talking about one of those in in this context and like you’re equ equ you’re like the way you describe an application is like somebody can’t do their job without it yeah if I’m building a business application there’s a new process that a team has to follow and utilizing that application a team process yes in the same way should be with reports with with specific reports right like where I’m like if this is supposed to be a day-to-day Game Changer

37:53 supposed to be a day-to-day Game Changer if this is supposed to provide insights and somebody should be using it on a daily basis there are things that you would do in the same way that you would roll out changes in an application that you would expect users to use the application in the same way you would have users and an expectation that they’re using your report because without it they wouldn’t be able to do their job or your report isn’t providing any value but let’s make the assumption that it is that means that that person is not doing their job and I completely agree with that I I

38:25 job and I completely agree with that I I think what I’m I’m trying to get again this may be a parking lot thing is if I create that application that’s the required way of input of data you have to input something through the application the reports all consumer there’s the of course there’s the accountability on input because if you input it wrong that’s going to show up the ex the other way where you’re getting the output of the data if I don’t look at that I don’t know if any team who’s been penalized well we need feedback or you need to utilize this report it’s just oh take a look at the

38:56 report it’s just oh take a look at the report I haven’t seen it yet okay and it ends there someone can actually get penalized for not inputting the data correctly that’s process thing that’s a process thing yeah that means when you implement that report if it’s supposed to be as valuable as it is for the business then people need to know about it and they need to be held accountable to use to use it so like like I said that doesn’t in my mind my mind doesn’t put it in the realm of reporting and what we build is not just as important as what you would do in an application you’re you’re describing

39:26 application you’re you’re describing that some cannot do their job without going into the application and in the context of a lot of the reports that we we we build for the business if they’re not using it they’re not getting the value out of it and there should be the same oh you’re in trouble kind same oh you’re in trouble like methods carrot stick whatever if of like methods carrot stick whatever if your job is to enter data into the application and you’re not entering data then you’re not doing your job if your job also includes looking at this report to see how like what your percentage is of enter dat data in in the application

39:58 of enter dat data in in the application and you’re not looking at it then you’re not doing your job but it’s also the organization has to say like hey there’s this new thing we built our expectation is that you’re monitoring yourself or doing these things and using the report in this way not all reports fall into that but I like I I would tend to believe that those are important to the business because they’re spending money building them and if we’re not getting engagement then we need to fix why we’re not getting engagement not think that

40:28 not getting engagement not think that what we’re producing isn’t of value everything you said I completely agree I my only argument is in a data culture and I don’t think that accountability is there for a lot of the teams but again we can agree I think we have a long road to push for those types of of things because without it then yeah you’re wasting time you’re wasting the business’s time right i’ agree with that one too that just cuts off that manag side I do I do want to roll into

40:58 manag side I do I do want to roll into this other part though which is is the scrum and agile model like effective for product for development and Tommy your description right in the beginning was I’ve tried this and I’ve tried that but you didn’t really land on like what it is you do today or what do you think is an effective way to produce a reporting product so yeah we’ll call my intro a hook so to speak on the interest so no we when just bit of context little back around powerbi first came

41:28 little back around powerbi first came out we were early adopters or organization we were lowly D data analysts who become became powerbi users and the need to move from the existing Source there’s a lot of excel a lot of other different products into powerbi was Paramount and the amount of requests that came in and tickets and people realizing the data the influx was huge a small team the entire business it was a central bi team y and we tried at first we tried these two week Sprints

42:01 at first we tried these two week Sprints which did not work for from a development point of view from if from report creation especially we found a little more success with data engineering with a two-e Sprint but for report requests for report creation for that that side of it the two-e Sprint we would find like well we got more feedback and they’re like well no we need to complete this report in this time frame but as you went through it you realize like well we’re getting more feedback they want this change beforehand that same idea of we’re doing

42:32 beforehand that same idea of we’re doing Discovery Well there’s now that iterative feedback approach and now we’re going to do development so that trying to fit things in for every report in that method in work the agile scrum method or at least the like the standing this the standups every day worked fine but you realize that we have more reports coming in than we can handle and just working on things on every like from a we’re going to say like what are these stories doesn’t work because it we realize it was more like a help

43:03 it we realize it was more like a help desk and what I actually came to when as director bi and moveed forward was this help desk system and really like a priority impact Matrix because we realize not every report is the same not every request not every request is the same agree so we had a impact Matrix at least developed of who’s the team who’s asking for it what’s the business value trying to rank that is it coming from an executive because that automatically takes priority is this something that

43:33 takes priority is this something that doesn’t exist that we need that takes priority where’s the value there and then the that went to the basically how we prioritize the projects in the help desk okay it was more important for us to make it transparent with the business rather than having this two week Sprint so they knew where their project was they knew where their request was in the time frame and if it does if it did become from the team more important it worked on like well I can’t wait that

44:04 worked on like well I can’t wait that long I know I’m in the backlog but I can’t wait that long then at least that was a conversation with their executive or their vice president and a conversation with our head our head of data was it certified I’m serious like that’s that are you building was your team focusing on building certified things or were you so there’s two layers to this right there’s there’s the layer of there are you’re addressing what you’re what

44:35 are you’re addressing what you’re what you’re describing to me is why it didn’t work is because you had a mix of known questions or problems that we’re regularly answering every day with common reports what I’m hearing you say was there’s a whole handful of like we’re just trying to help the business answer a lot of these ad hoc questions so that’s where a lot of these other tickety type things came in the door I can’t get this this report doesn’t do that they’re the users are using those reports and they’re physically asking

45:06 reports and they’re physically asking different questions of that report the report probably wasn’t even intended to do exactly all those things or answer all those questions so this is where I would argue like in this case there needs to be a delineation between you needs to be a delineation between sprinting an agile I think work know sprinting an agile I think work better for certified reports because you’re adding process to what that is there’s another world of this that’s more like I would call it like a combon process it’s literally requests come in we prioritize them and we just work from the top down and like you just you just

45:36 the top down and like you just you just do them and get them done so I think if you try to put both of those processes together and do everything the same way it doesn’t work yes and that’s and I think that’s what your point is right you can’t always think of like everything as a Sprint you can’t because in order to do that in order to Sprint sprinting is good when you have a defined amount of work for a certain period of time that’s what the Sprint’s good for so requirements are done on Monday build happens Tuesday through Thursday we release on Friday and

46:06 Thursday we release on Friday and acquire feedback the next week we come back and say do we have new requirements or are we adjusting things what are those new requirements and you continually sand refine tweak and then the report after a couple weeks actually

46:21 the report after a couple weeks actually meets the needs they’re looking for all this other junk is just no the system it’s still important but you’re right the Project based things where like we need to create a new user tracking for all of our tables to create a unified ID that’s perfect for the Sprint method because it’s the we’re getting the requirements is longer based but 80% of at least the quantity or volume of work was requests coming in we do need a new report maybe off an existing model and it

46:53 off an existing model and it may not necessarily be certified it may be coming from a team our sales team wants to see everything in a different way we want to see the first three customers this year thing which still important yeah but doesn’t doesn’t fit in that scrum method or does not fit into the Sprint method and I’m smiling here you can’t see me doesn’t what doesn’t fit okay oh are you’re asking me yeah I in general it’s what what what seems to not fit as much is more things that come

47:24 fit as much is more things that come from the help desk whereas single reports different views of the data new ways to measure it what seems to work or what has worked a lot better because it is more of that I think that approach is more Project based work where we have to integrate a new data source into our system we have to you source into our system we have to we’re creating we’re utilizing now know we’re creating we’re utilizing now this new platform we have to find a way to integrate that with our sales team we have to find a way to integrate that with the customer ID we’re creating

47:54 with the customer ID we’re creating single sign on we need to create a way to do that we’re we’re migrating from table X to table y so to speak in all of our reports I would argue I would argue everything you just described and that works well with scrum and agile those are big projects those are big things yeah you can’t throw big projects in with I need a new report right because new report builds should be and I would my thought here is as I’m hearing you describe this I’m saying to myself what

48:25 describe this I’m saying to myself what you’re describing is we actually we don’t need that what we really need is you need an ability to Big projects go through scrum and agile because there’s large requirement Gathering up front and if you miss some of those major requirements up front you could screw up the whole project and you could take off waste weeks of time building the wrong thing that we don’t need on the other hand if you’re if your if your team is focusing on I’m the the team that’s going to be building all these reports and again I know in your background Tommy like your team was the

48:55 background Tommy like your team was the C data team and the rest of the organization was not equipped enough to go build their own stuff so in my opinion here was if your models are so complex if your models are so hard to work with that you can’t give someone else a model or you can’t start self-servicing a model to a champion in a department to start doing things in their own Department that tells me two things either your company’s not big enough to start delegating work of the report builds to the individual

49:25 of the report builds to the individual user at the end of the day what’s really important and this is where I think the scrum and angile makes sense what are your data tables that you care about what is the relationship to the factual type data what are we going to calculate what are the measures or basic measures that we need that’s your that’s your bare minimum that’s your table Stakes element there the the work should be focused around making that right because then you can hand this over to the business and say look we know we can’t answer all your questions about every little piece of data you have however

49:56 little piece of data you have however when we’re thinking about things analytically here’s where you can bring the calculations from here’s your handful of tables oh and by the way if you want to build any analysis between these two different factual types of information call it promotions and sales that’s usually two different data sorts sets so either you have to merge those two data sets together to get one solid one one single fact table or you’re building Dimension tables in both that link to both of those fact tables to me the the value would be here

50:27 tables to me the the value would be here is you need to teach people this is an education thing or the skills are not right in the team yet there is a there’s a people problem that we’re addressing and we need the people to be able to self-service to some degree so so here’s where I’m going to weigh in because a lot of I I think I I’m drowning over here I think I found a hybrid approach where I I agree with some of this but other aspects I I think do fall

50:59 but other aspects I I think do fall into the the same standard process and I think that’s what’s important in all the things that Tommy just described a lot of is people find themselves overwhelmed and need a way to manage not only expectations but communicating with the business what can be executed on in certain time frames and that’s what’s that’s what this whole thing was designed to do right create some some breathing space and also clearly

51:29 breathing space and also clearly articulate and deliver in time frames that could be acceptable I think it’s important to understand what is a quick data request give me the logic give me the columns it it’s something that we’ll will throw into the 30 minute like thing boom we’re just going to keep throwing these in what I found is like so let let me back up Tommy priority Matrix like that that’s just work item status like how you set those things and how they they come into a workflow like it can be managed within tools like I’m

51:59 it can be managed within tools like I’m a Jura user the types of requests that are coming in all the time like you need to groom your own backlog you need to figure these things out Etc yes there are the big projects and I think we all find agreement like it’s great you go through full groomings other teams are probably the product owners or project right like they’re coming in you’re talking about them because they’re big lifts I think we where the hybrid comes in I think is with report builds we are assuming

52:30 with report builds we are assuming responsibility for a lot of those grooming type activities but that’s where I would I would argue it’s more important that we we follow the process or stick with a process that is aligned with these systems that manage execution because otherwise you like especially in larger teams it’s like in the article it describes like oh the whole team needs to get together and execute against this thing it’s like yeah that’s great but that’s not how the world runs like all of my team is spread out doing different

53:01 of my team is spread out doing different activities and tasks and there’s not always this realm of like yep somebody can work on the data set somebody can work on the report like it’s usually the same person so having this process where I have a story but there and a process by which there’s an understanding of how we engage with the business and those are tasks the tasks are would get assigned in the Sprints because I can’t I’m not going to say we’re going to get some we’re going to get this thing out in two weeks what we’re going to do is we’re going to iterate through a lot of

53:32 we’re going to iterate through a lot of the things Discovery this is where we’re doing the grooming this is where we’re Gathering the requirements talking with the business understanding all the nuance and I do think that that’s important right out of the gate as opposed to allowing business to dictate what it is they want because that that challenge comes in with what you described which is oh yeah we started building something and then they were like no we want this up no no about that up no like I’m sorry that’s not how this works up to a point like we’ll adapt

54:02 works up to a point like we’ll adapt we’ll will Define but when we’re defining what the base is supposed to be that’s what we’re shooting for in a collaborative method if all of the sudden as we’re finishing up you’re like oh what I need I need another page and we need like yeah I’m sorry now that we’ve defined the base that goes into our cycle again that gets scheduled that’s getting pushed out we’re not doing that right now and that’s how you control that conversation so I think like what I’ve done is I’m still sticking to this process the and I still

54:32 sticking to this process the and I still have a lot of the framework for scrum methodology where twice a week we’re meeting with the team because it’s important that the team members understand what’s going on with everybody we are doing groomings or at least scheduling those meetings consistently for people that want to show up we are doing technical planning we are estimating things and we we do follow a two weeks Sprint always does that mean we’re not flexible no but it means that we can stick with a a

55:02 means that we can stick with a a methodology in which we want to engage with the business and set ourselves up for success and being able to clearly articulate when we can deliver things and I think that’s Mo the most important thing we get out of sticking with that cycle even though not everything flows through the cycle in the same way and I think that’s the difference here is I’m not saying everybody’s got to come to a grooming you’ve got to formalize thing I’m like no we’re the bi team like we will be doing those things but it’s probably in

55:33 doing those things but it’s probably in conversation or it’s in a meeting but there’s a structure to it it’s just the business people who are making these requests and we know this nine times out of 10 don’t know what they want when they make the request they just know they have a need 100% I have a need and we’re like we will address your need here’s the priority we’re going to affix to it and when we engage with you we have a methodology that we always follow so we can do all these things with the business which is where’s the time te like the team time being spent where are

56:03 like the team time being spent where are we like what are are we dialed into the right priorities how much can we deliver in these these time frames where we at in the phase of the project like all that shouldn’t be this ambiguous like backlog of like oh now we had this thing up now we had this thing because then there’s utter chaos so I I that’s I I guess where I push back in like scrum methodology and whatnot I agree because I think there’s this hybrid area but we are still following that process just albeit in a

56:35 following that process just albeit in a different way yeah I like what you said there Seth and I would agree this works well for me that same pattern works well for me as well in in projects there is certain things and this is where i’ said this is why I’m I’m I’m on the certified Jam right now I don’t know why it’s just things that are certified yeah there’s there’s a there’s a regular release Cadence cyle ccle and you don’t try to push stuff out as fast as it comes in however that’s like I don’t know what percentage 60% of your team’s time there’s another 40% of time that’s

57:06 time there’s another 40% of time that’s just okay two days a week or every morning from this time to this time we’re just doing ticket ticket pushing like and so I think there’s a hybrid approach where you’re you have to carve out there’s there’s only so many people to do all the workes of work types of work yes and then if there the and then the business has to understand like look we’ve got longer strategic things happening as well as your short-term quick wins with building your reports doing these things whatever

57:36 reports doing these things whatever issues issues enhanc little fast reports exactly yep you have to meet both needs at the same time so it helps you can’t do either one you can’t focus solely on one or the other you have to have a hybrid approach of a mix of the both of the workloads the more you guys talk the more I’m I I’ve said this

57:56 talk the more I’m I I’ve said this before but the more I’m on the train not just the certified train but the establishment of the bi team as a bi Department I I really think this is the logical next step with organizations because right now what’s Mike he keep saying this and I know that this is kind saying this and I know that this is the protocol is bi is under Finance of the protocol is bi is under Finance I’m beginning more and more to disagree with that where I don’t say the is under Finance I say Finance lot of organizations I say you delegate well I organizations I say you delegate well okay I will say this obser

58:27 mean okay I will say this obser observationally the people that hold the checkbook who understand the value of data value of data yeah is typically Finance because Finance can’t even do their job unless they’re getting all the financial datas and so usually what this is how I see it right powerbi was born out of excel so this team is already the best Excel team you’re going to have in your company almost Baran inside that team you’re going to have experts who understand the data engineering the business rules all the shaping like they get it and they’re building databases inside Excel which is

58:59 building databases inside Excel which is incredible so awesome but that’s the team that says hey we can now use this parbi thing and then sometimes that spend just gets attached to them just by default then Le the next question who’s the sea level in in charge of bi usually I think it’s the CTO where it typically Falls because we don’t have a very well-defined scope of what the CDO would be doing because well yeah but but is it though like because even in your your I think it’s all over the place right now like sometimes it comes into the CFO right some sometimes it’s

59:31 into the CFO right some sometimes it’s operations some I think it’s just by by hybrid depending on where a group of data people like grow or where the focus of the business is but I agree with Tommy like and and where I’m at like it it is a it’s CTO but it’s an independent body and I think I think the the impact you can make to an organization is larger in that realm because then you truly need to rely like couple reasons one you’re plugged into

60:02 couple reasons one you’re plugged into the Enterprise systems 100% And the teams that are going to help facilitate and support that that realm two when we look at increasing the value of data culture or building relationships into all business units right you’re looking for people finances a perfect one where you have Experts of data right that you leverage them as an organiz right there’s a central team that is providing value in terms of training enhancement data access etc etc but

60:33 enhancement data access etc etc but you’re doing that across the organization and not just within One Singular business unit and you’re looking to skill people up across an organization like I don’t see how you would like succeed in a lot of the adoption and data culture things without that being an independent team yeah and like to your point if you’re trying to create a structure red methodology or process with an unstructured team where well we may live under CF under Finance right chall yeah it’s a huge challenge

61:03 right chall yeah it’s a huge challenge because again where does your work get dictated it’s going to be from your manager to their whoever their vice president or sea level is if we’re going to have this hybrid approach which I completely agree with because the other parts of your team is not just those big projects it’s not just helped us but it’s also the adoption it’s also you it’s also the adoption it’s also the promotion and of thech know the promotion and of thech technology and the people the people utilize this so you’re not just doing Project based things one of your projects may be adoption and how are you

61:33 projects may be adoption and how are you promoting and promoting data I I think more and more that the need for business intelligence being a department who has a CA level who’s dedicated to this because we know how much data is getting created we don’t just live Building Technology and we just don’t live under the the numbers right because we’re in charge of so many areas of the business where their requirement for all this to work effectively with the right stakeholder with someone who understands this to me it has to fall under someone

62:05 this to me it has to fall under someone like a CDO or someone who understands the three main areas I wouldn’t put under I don’t think people defined a CDO well enough and organizations are still not defined I don’t think it’s GNA get there there honestly stake we’ll put a now I I’ve tried to look at what cdos were going to be doing I was cuz that was a role I was looking I was seriously interested in looking in doing a CDO role and I don’t there’s too much ambiguity around what that looks like is it a security role is

62:35 that looks like is it a security role is it technology focus is it only on data there’s there’s so many the CD I read an article recently that was like why all CD rols fail is because it’s not well defined because there’s not a scope so what happens is I think this CDO role is likely shared or underneath the realm of the CTO the chief techn officer or potentially there’s some exe some executive I don’t think it matters honestly I don’t think it matters where it sits I think all it matters is what Microsoft speaks to is you need an

63:05 Microsoft speaks to is you need an executive sponsor to be that person to run and own the vision and the leadership of that Central bi team where it sits under Finance under CTO under operations I don’t think we care honestly as long as they have someone who’s responsible for it and understands the vision of where you’re going with it Mike be before I turn 40 I’ll bet you 20 Stakes that that Ai and 20 Stakes before that Ai and data there’s going to

63:36 before that Ai and data there’s going to be a CDO type level because where AI is going that’s going to fall under data that’s not going to just fall under technology and where how much data is getting created and the technology around data bet 20 Stakes how many organizations have even thought about their data being a strategic asset I think just starting to get that point I I I agree with you Mike I I and I think the it is very dependent and you need to

64:06 the it is very dependent and you need to be very specific about the roles and responsibilities of a CDO within an organization yes I I do agree with Tommy in that I think if you are going down that that route route there is there is more than enough responsibility for data within an organization that could be aligned to that role but at the same time like I agree with Mike in current state ITR when we when we’re talking about a group

64:36 when we when we’re talking about a group business intelligent like we’re we’re taking we are part of it we are part of business it doesn’t matter who owns that provided that we’re running that group with both of those things in mind and I think that’s what what we push the structure of it for this group it is 100% a good thing to to to manage the influx of everything that we have more so probably than any developer group because of the way they can control that workstream whereas there are business pressures and there

65:06 there are business pressures and there are things that turn around much faster in business intelligence teams and it is much closer to the business because we are trying to solve those problems in incremental phases even even from a step one step two step three thing right here’s value but our underpinnings for this thing are on like they could break at any time right phase two is we’re GNA we’re GNA we’re going to while you have value we’re going to build up the accuracy the governance the quality of the underpinnings and you don’t need to know about that and then

65:37 don’t need to know about that and then we’ll iterate right so I I do think it’s much more agile in terms of a scrum methodology or whatnot than than developer teams in general can can be will there be an ultimate role that is a CDO maybe I don’t know I think the things that we expect a CDO to do are going to be broken apart existing roles that are already in the organization and they’re just going to pick someone and be like you’re the dude for data you’re doing good with it right now seriously like in in

66:07 now seriously like in in companies I was in the past there there are pockets of teams that do really well with data all right let’s let’s wrap up here let’s do some final thoughts here Tommy give us your final thoughts from this article great discussion today I thought this was really relevant give me a couple of your final thoughts and we’ll wrap it here yeah I I really think the big takeaway for me is the evolution of bi you’re seeing with the evolution of the process as well we started off with trying to fit into different buckets and realizing that bi does exist with its own process I think we’re going to continue to see that Trend in the roles and I think we’re going to

66:37 roles and I think we’re going to continue to see those Trends in leadership as well Seth what’s your final thoughts here yeah I would just say have have some structure but but don’t be bound to it completely in the way that we you it completely in the way that we would would see other examples of know would would see other examples of but in order to effectively operate as a team communicate and deliver in in ways that we are expected that there that hybrid approach that I I talk to is is one that I

67:07 talk to is is one that I found to be successful I’ll give my last final thought here I think it really to me this really conversation sented around there are certain types of questions that are repeatedly asked over and over again there are reports that can be used for long periods of time identifying those things and turning them into projects makes sense because you can solve the problem and you can get value out of it and you can walk away and let it run for a longer period of time there’s a whole another world of AD Hocking reporting and unknown

67:37 AD Hocking reporting and unknown questions and this We’re just trying to get insights on data and if your central bi team is running all that you need a different process than scrum and agile to meet those needs so I see like there’s two worlds there’s project or process driven stuff and then there’s ad hoc and the questions are changing so you’re going to have to manage that work differently some other some other way and you’re going to have to split your time between the two so I thought that was a really good conversation today thank you all very much for listening we appreciate your time and your

68:07 appreciate your time and your listenership if you’re struggling with your data process comment below what are you using are you doing a mix of both are you doing something else one way or the other so we’ really appreciate your comments please make sure you share this with somebody else we we don’t do any promotion we don’t pay for this to get PR anywhere it’s all word of ear mouth word of mouth I don’t know what the it’s all by you sharing and letting other people know you found some value from this so we thank you very much for listening we appreciate your time I know your time is valuable if you would do us a favor share it with somebody else and you found value from

68:38 somebody else and you found value from this conversation we’d love to keep having more of these and continue talking about this stuff Tommy where else can you find the podcast you can find us in apple Spotify or wherever you get your podcast make sure to subscribe and leave a rating helps us out a ton do you have a question an idea or a topic that you want us to talk about in future episodes head over to powerbi tips podcast leave your name and a great question finally join us live every Tuesday and Thursday a. m. Central and join the conversation on all of powerbi tips social media channels and if you want to win or if you want an

69:08 if you want to win or if you want an opportunity to win 20 Stakes go to x go to x formerly known as Twitter and go bet Tommy something doesn’t matter what it is just bet him something and you too could win 20 Stakes directly from Tommy’s pula’s pocket you’ll find me on Twitter not X whatever either one you can go find them anywhere thank you very much we’ll see you next you next [Music]

Thank You

Thanks for listening to the Explicit Measures Podcast. If you have a topic you’d like us to cover, drop it in the suggestion link above, and we’ll add it to the queue.