Symbiotic Relationship of Data – Ep. 232

You can build a beautiful report, but if the underlying data is inconsistent—or nobody can explain where it came from—the story falls apart the moment someone asks a hard question. Governance isn’t paperwork for its own sake; it’s the set of processes and ownership that make your insights defensible.

In Episode 232, the crew uses a Medium article on governance + storytelling to unpack what “symbiotic” really means: governance enables trustworthy narratives, and strong storytelling can create pressure and clarity that improves governance. They talk through data quality vs. governance, why not all datasets deserve the same rigor, and how catalogs/lineage (and the direction of tools like Fabric domains) help teams discover and trust the right artifacts.

News & Announcements

• The symbiotic relationship between data governance and data storytelling — The article that sparked this episode’s discussion on trust, quality, and the feedback loop between governance and narrative.

• Submit a topic idea — Send a question or scenario and we’ll turn it into a future episode.

• Subscribe to the Explicit Measures podcast — Follow the show and browse every episode in one place.

• Tips+ Theme Generator — Create consistent report themes fast (and stop hand-editing JSON).

• Mike Carlo on LinkedIn — New episodes, Fabric learnings, and practical Power BI tips.

• Seth Bauer on LinkedIn — Model engineering, performance, and analytics delivery.

• Tommy Puglia on LinkedIn — BI leadership, adoption, and real-world program advice.

Main Discussion

The article’s opener is blunt: you can’t tell a good story on bad data. The team takes that idea further—bad governance doesn’t just create incorrect numbers, it creates fragile insights that collapse under scrutiny, eroding trust in the model and the people presenting it.

They also flip the relationship around: if storytelling is what the wider business sees, then a well-framed story can drive governance forward by making gaps obvious—missing definitions, unclear ownership, duplicate metrics across models, and the absence of end-to-end lineage.

Key takeaways:

• Treat governance as a trust engine: definitions, ownership, and repeatable checks are what let stakeholders rely on the story.
• Separate data quality (accuracy/completeness/consistency) from data governance (the process + accountability that makes quality sustainable).
• Not all data deserves the same rigor—invest the strongest governance in certified, business-critical datasets and measures.
• Avoid “hacking” or reshaping data to match a preconceived narrative; let the data challenge the story before the story goes public.
• A data catalog only matters if it helps discovery and connects artifacts to owners, intended use, and certification signals.
• Lineage is the missing bridge: users want to trace a measure from a visual back through models and curated layers to the source system.
• Good storytelling can improve governance by making gaps visible to the broader audience and creating urgency to standardize metrics.

Looking Forward

If your governance program feels like a side quest, tell a clearer story about the business risk of mistrust and the payoff of certified, discoverable data.

Episode Transcript

0:00 [Music] foreign foreign [Music] good morning everyone welcome back to the explicit measures podcast with Tommy

0:31 the explicit measures podcast with Tommy Seth and Mike hello everybody there it is good morning I keep thinking to myself at some point in time we need to go on a explicit measures World Tour and I can’t wait till we hit Italy and and Tommy shows me all the things I’m eating wrong about Italian food do we have followers in Italy

1:02 have followers in Italy do we have followers your mom follows us this is a wow but she’s not from Italy though though yeah I think we need first generation does your mom still listen to the podcast podcast probably the last few weeks since I haven’t looked this is this is how she connects with her son that’s funny well if you’re if you’re in the chat this morning like feel free to say hello to Tommy yes Mom when you start a new job and it’s also

1:33 when you start a new job and it’s also your own company and the HR department’s terrible terrible [Laughter] not wrong I’ve gone through that pain and it definitely you overwork I think it was it Thomas you were telling me something like this it was business has failed in the first six months but businesses are success after two years was that is that you Tommy yeah it was basically from a a book it’s the guy who’s very successful he’s like things fail it takes six months for things to fail two years things to

2:03 things to fail two years things to succeed so you’ve made it six months you got to keep going to make it to two years do you guys do you guys get to play you do you guys do you guys get to play games around that well know games around that well being the CEO of this company I’ll take into consideration your request for human resources response I will send them a note no note noted HR we’ll talk to the legal team and we’ll come back with a response yeah right

2:34 come back with a response yeah right amazing awesome so a couple articles came out well actually one article that we want to talk about is on the Microsoft blog there is a new released an article around data sets are now in power bi mobile apps and now you can refresh your data sets while also on the go so when you’re going to the bathroom and you’re hanging out for time which scanning for YouTube Facebook whatever you can now refresh your data sets so what do you guys think about this new feature feature I think I think the well my first

3:05 I think I think the well my first reaction was oh great now I can do my job more for myself my second was well along with this comes the push notification right so you through the app it’ll push notify you if a data set fails now that’s cool because yes there’s a lot of Plumbing you got to set up behind the scenes if you get that working which you can right now but ease of use feature enablements like that I think are fantastic so the two scenarios that pop in my head I guess right would be hey

3:36 in my head I guess right would be hey somebody wants a the freshest data and you clicking refresh is gonna work yeah but if something’s failing you’re not gonna be able to solve that from your phone anyway like hitting refresh is just gonna end up in the same phase we’re waiting for the blog behind this one now open open data models on your phone and model your entire model and write measures on your phone now yeah I completely agree with that the ability for the push notifications for like very business critical data sets where you really need to know at any

4:07 where you really need to know at any point and you it may be your responsibility the push data sets work seamlessly though I was really surprised how quickly or just how easy it was just to refresh the data set for mobile a TV it’s easier on mobile than it is on with the web the web I I like the idea of I think to your point Seth I think I think this idea of knowing on your phone when refreshes fail fail golly every everyone has a phone stuck to them somewhere right you you leave your house without your phone you’re like I feel like I’m

4:38 you’re like I feel like I’m losing an arm thing people are so attached to their phones and maybe that’s just maybe that’s showing you a bit of my heart how much I’m attached to my phone all the time but I I think when there are data sets that are extremely important to know what what’s happening with them what’s occurring with them I’ve built reports for companies that that track every time a data set refresh happens and making sure that it’s working and for those Mission critical things yeah you 100 want to know when it fails and you

5:08 100 want to know when it fails and you need to know what’s happening all the time when that data set is having issues now granted again I may be away from my computer at least there’s I’ll be notified but I would say say I’m not sure so I think this to me will be a feature that I’ll be using when I’m physically away from my computer I don’t think I’ll be using this in conjunction with okay I’ll have my phone up on my right hand side or my iPad up and then my computer it’ll probably still be something that I use when I’m not near an actual PC yeah and and the other thing I would caution is

5:38 thing I would caution is you you can accidentally go the other way like having push notifications is great I would not recommend putting on every single data set unless all of them are mission critical yeah right which is usually not the case yes otherwise what happens is it’ll take a day and it’s like ding ding ding like the The Notebook stop meaning stop meaning something yes yeah let’s let’s hope your data set refreshes just a little bit better than that over notification of of failures as well depends is your data source

6:09 depends is your data source Excel then yeah bing bing bing bing is going to notify you Tommy adds a column and all of a sudden all my data sets fail now because we got one one new column inside our data great lake is doing just fine right now good thank you I will agree yes but luckily for us we’ve been pulling everything from CSV files so it’s much harder to add a column in a CSV file at this point very true very good article really like this this I I think this comes I don’t know if they’re still doing this but I believe the the team out of Israel is doing a

6:39 the the team out of Israel is doing a lot of development here in the mobile app and so they’re they do pretty cool stuff I like I like what they design they do a good job on the design work for all the mobile application pieces so nice job team looks good looking forward to trying out that feature all right all right let’s go over to talk about some of our main topic today so today there’s another article we’re bringing out it’s actually in the description we’ll also add it here as well into the chat window the the topic for today is the symbiotic relationship between data governance and

7:09 relationship between data governance and data storytelling I think data storytelling is awesome I really like it I think there’s a lot to do there do there but data governance I feel like is something some organizations have their hands around I’ve walked into a number of organizations who have data governance in place but don’t really understand what data governance looks like with the context of power bi and what is the surface area of how data can be moved and shared around I think the the

7:39 and shared around I think the the challenges I see here from from my perspective RBI really changes or adds the ease of sharing to a whole new level and so data governance becomes a whole new story so initial thoughts on the article gentlemen anything you want to pick out here and start talking to man if there’s an article Meant For Us like the two probably the two topics we probably talk about the most data governance and probably storytelling and framing you’re probably right yeah yes

8:10 framing you’re probably right yeah yes like could they have picked the most the two topics we talk about every single week yeah week yeah no matter what they come up at some point point it’s a very interesting take and I’ll I’ll be honest going through it do I agree with everything or does it make perfect sense I’m still digesting that but the idea for those listening is the idea here is there’s a very mutual and beneficial relationship between data governance and storytelling where they in a sense complement or they can be

8:41 in a sense complement or they can be to the determinant of each other and I I like I like how they open the article so my the opener here resonates very much with me right you can’t tell a good story on bad data you get a weird story right well did that you story right well did that if we all if if we only hired know if we all if if we only hired another 100 sales people in the month of December we could always have the highest sales every year well that’s because that’s the end of the year and that’s when we’ve had like that’s what we’re like something like you could have weird stuff going on on your data that

9:11 weird stuff going on on your data that really makes this just like not helpful but but when when you talk about even particularly for like machine learning things right if you’re talking about a machine learning algorithm the algorithm is always performing at its Peak Performance again there’s things you can tune add features or take away features but when you start looking at like what level when you look at the research when you say how can you improve an AI model give it better data whenever you give it better data it produces better results this is like really speaking to that artifact of

9:41 really speaking to that artifact of better governance better control better understanding what that means really could impact what that the output of the story looks like yeah I it’s an is at one I like the word symbiotic symbiotic so it’s cool the the combination or his linking really the foundational aspects of the a lot of back-end data grunt work

10:10 lot of back-end data grunt work being being the groundwork for like the story right a narrative that that is being told I I like I like the thought and I agree with it from the standpoint that to your points Mike you can you can build pretty visualizations and you can build you visualizations and you can build a a story but if your data is know a a story but if your data is garbage like at the end of the day one of two things is happening you’re either going to start hacking out data to make it fit the story which is absolutely not

10:40 it fit the story which is absolutely not what you want to do because then you’re modifying the actual truth right it’s not just cleansing it’s it doesn’t look the way I want it to yep but I think it also highlights in where I’m going to spend I think most of the time today is talking about that trust Factor like laying the groundwork and when we throw out buzzwords of like data quality or data governance there’s a lot that goes into that right and I I like the the value prop that he’s he’s pushing and pulling

11:11 prop that he’s he’s pushing and pulling here because the the shiny fun is the storytelling it’s the narratives it’s talking about the insights and the reports themselves but by him in inextricably linking these together in a symbiotic relationship right I I think is a is a really good analogy if that’s an analogy or a good use of the word or a combination of thoughts maybe that’s a better description which one of you two was very sorry sorry not I think you’re actually

11:43 sorry sorry not I think you’re actually right who would made a big distinction between data quality and data governance do you remember a previous podcast okay oh data quality data governance yes yes yeah it would be I think it would I definitely think there is a difference between the two yeah I do think they’re very very interrelated though as well yep so like it’s it’s definitely yeah I and I think I I think I would say good call out yeah

12:13 good stick you’re right I’ll I’ll stay in my Lane data governance sorry well I I also again I’ll keep harping on this one as as I do with all these episodes I feel like everything’s very applicable again not all data is created equal not all data needs the same rigor that you’re talking about I think I think the impact the decisions and the and the the seriousness of the story story really decides a bit more rigor around where did that information come from right anyone can just build pull some

12:43 right anyone can just build pull some data from some system start talking about stuff and look we’re our sales are up 30 we’re good no problem let’s move on in reality costs have increased more than 30 right so there could be other things that are and that a factor here that the story is not comprehensive so you I would say in some cases right the where that source of data comes from needs to have other sets of eyes or other analytical team leaders to really review

13:13 team leaders to really review that story and the the message in that story is going to potentially weigh very heavily on the business well and that’s that’s what we’ve always pounded on too is you cannot have proper dative story framing or storytelling or have the empathy if you’re not speaking the same language or if people don’t even trust the person doing it too it’s the user It’s the final result it’s what they’re actually asking for all that goes into exactly what you say

13:46 I would be curious though because we always talk about that side of the relationship we say data governance to data storytelling but if this is a symbiotic relationship then that means it affects it the other way too and I don’t think that’s something we’ve touched on a lot is how does data storytelling affect data governance ooh ooh I don’t think I’ve ever taken that kind I don’t think I’ve ever taken that angle on this before I of angle on this before I assume because this is symbiotic so they affect each other I think whether whether it’s an implied meaning or I caught it in his closing I

14:18 meaning or I caught it in his closing I can’t tell because when I when I think about that I think the the storytelling is what is up front and center to the the wider audience right right and if it’s not good data good data if it’s not accurate like we’ve all we’ve all been part of like the report that is somebody pokes a hole in and the whole thing starts leaking right like that is where there’s a real

14:48 like that is where there’s a real negative push into hey why why is this data wrong this is somebody’s fault but in reality it pushes towards well if you guys don’t have a governance strategy you can’t expect high quality data coming out of here right so that’s a negative one but in the on the on the reverse side of that you have trust you have like like insights coming forward that are making significant business impacts and then you can you can push back and rely and and point at we have all this because of

15:21 and point at we have all this because of all the efforts that we’re we’re applying to the data and the structures of that before you guys even see it that’s how I saw it I’ll I’ll I want to jump on your story there as well I’ve read the random statistic and I’m trying to Google it here very quickly just to make sure I’m not going so off the deep end here I think there was a research done a research paper done I think this article was written by Robin back in 2017 but it was

15:52 Robin back in 2017 but it was basically the headline is Excel errors How Can Microsoft spreadsheet be hazardous to your health basically what it was for Sheets spreadsheets right okay anyway it’s basically it was saying something like the idea that research had found like 90 of all spreadsheets or sheets of information that was coming out of excel 90 had some errors somewhere inside the Excel sheet right and I can’t tell you there are times I’ve when you build the Excel sheet initially yeah it’s probably spot on but after months of working

16:23 on but after months of working on it and copy pasting cells and moving formulas around sometimes there’s that one cell that’s like whoops I missed the formula here I’ll just go correct that right there’s there’s usually little errors that appear in there and potentially cause issue with that so to your point there Seth though I think that’s potentially the same scenario that we’re seeing here as well with the data storytelling and the government you’re you’re bringing the data together you’re manipulating Excel to start fabricating what is that story and I think this is a skill that can

16:53 and I think this is a skill that can like to some degree I feel like the storytelling element here is an a learned skill to some degree right being able to know what’s what is the most important piece of data that I need to look at where where does the largest amount of sales and I think I come from this from a I was in a category management team for a number of years working on understanding where markets were going and which companies were making which sales and why right so looking across an entire industry trying to figure out who was

17:24 industry trying to figure out who was making moves inside those different that that industry and making sure we understood okay if company a adds a promotion how does that impact everyone else what is the market share gain or loss of people and so even though your whole even though as a company our product sales may have been down we have been may may have been getting market share for whatever reason and in the industry we were in there was a lot of weather-based analysis going on right the weather impacted whether or

17:54 right the weather impacted whether or not we sold a lot or not so when you look at those things there’s a lot of stories that can come out of that ecosystem and you really had to figure out okay where is the most impactful story where did the largest or data impact come from and it’s it’s not easy it takes some time to kind it’s not easy it takes some time to really think through how do I of really think through how do I identify all the datas and figure out which one drove the most return or loss in sales in sales and I think this is speaking to this article summary or that that data story element you can only do the data

18:25 story element you can only do the data story telling after you have some solid data to support it yeah real quick here I don’t I don’t think we attributed this article yet did we like or this oh I don’t limit it either Haro wiersma he’s an independent consultant and data governance evangelist and it has I think I looked at several of his other articles on medium so check them out I like it like it anyway back to your story sorry storytelling spreadsheets fabrication of them well and that and that was

18:55 them well and that and that was that was really my point right I think a lot of times we go at so I think this from my standpoint the storytelling we usually come in with a preconceived notion I think a lot of people come in and again I’m definitely at fault of this too I want the data to say something in a specific way so I wanted I want to jump in there and say okay I I want the data to show this analysis and so you start kind analysis and so you start looking for preconceived notions you of looking for preconceived notions you have to come in with an empty mind sometimes and let the data tell you what’s going on and try not to bias and

19:25 what’s going on and try not to bias and push the data around so you get the story that you wanted versus what’s actually there because you may not like the story that’s there yeah then manipulation of data that’s a great it’s a good point I get the one point you made that earlier that I want to go back to is okay you I think you said that not all data goes through a governance process so if the only parts of only parts of the data organization would would go through the rigor to what what the end

19:55 through the rigor to what what the end goal of that for us would be like certified data sets and and things like that that serve very focused insights in certain reports or the report audiences such that they the business wants to invest in the additional steps yes right and this is where I count one to talk about with like the governance part of things

20:19 like the governance part of things because quality data is part of that like then an outcome of having a a process but when we talk about data it’s accuracy completeness consistency right are the three things that he he he starts with and all of those require somebody whether it’s an analyst whether it’s the like probably an analyst because they’re spending a lot of time in the data and it’s taking that taking the requests

20:49 and it’s taking that taking the requests but also understanding the business are my tables complete are the way is the way that we’re shaping or applying business logic to this data like resulting in this combination of quality and if things are dropping out or we’re filtering them out what is the purpose and the reason behind that so that the entirety of the business understands what is happening in this secure or accurate method by which we’re doing

21:19 accurate method by which we’re doing more due diligence around data than we hadn’t before it and that’s a lot of work especially as data keeps consistently coming in you need to build monitoring processes around capturing errors or issues but I think this also like teases I guess a thought in the lake house versus data mesh conversation right where essentially a data mesh is saying like hey this is this is going to be data controlled or owned by business units and and somehow

21:49 owned by business units and and somehow combined and collected does that stress governance strategies to you like because now I I have to apply a governance policy to all teams and they should be managing their data in a certain way or do I have to double validate that before I put together a certified data set because it’s easier and like if I in a centralized way if I’m saying hey all data is coming in through a lake house we’re going to apply all these policies to make sure that we’ve cleaned the data and created a certified data set on the other side of it can can I

22:20 set on the other side of it can can I get the same expected output in both those both those Realms and and at what cost obviously the answer is yes but but like yeah I see a real challenge in one versus the other well I wonder if there’s some there’s some value towards I usually think of governance Downstream to storytelling like whatever you do in governance is going to enhance storytelling but as I think about this and think about a more of a colleague the recycle symbol right of Storytelling and also enhancing

22:51 of Storytelling and also enhancing governance think about what storytelling could do to highlight certain areas of the data or certain things where look at the value that is provided or look where we need help and part of Storytelling can also be maybe in a little pilot program of look what storytelling has done where we’re highlighting things that we need to do we’re highlighting the priorities that governance needs to focus on focus on and that could possibly be one way where we can then begin to say we can’t upgrade this yet or we can’t promote

23:21 upgrade this yet or we can’t promote this yet until we get these certain things done but we always think of Storytelling as for like our sales our our marketing but we don’t think a storytelling for governance like what governance programs are we trying to do what governance program should we do part of that goes to data quality too but highlighting things like hey these are the Stepping Stones if we want to have a higher data governance we need to fix these issues or these are the things that we need to work on yeah and I’m

23:52 that we need to work on yeah and I’m guessing there’s a couple of thoughts I’m I’m leaning on both for both your comments there from your your comment Seth right there is there’s going to be some sort there is there’s going to be some cost associated with governing data of cost associated with governing data and again I want to be clear when we’re talking about when I’m talking about governing data I’m talking about these are the guard rails by which we say we’re going to do certain things or not do certain things and so to me what governance means is when there’s a data set or when there’s some data that’s published we know who the owner is people are are actively looking at the data I’m not I’m not saying like

24:22 the data I’m not I’m not saying like literally you can’t access this table or not access table yes that’s probably part of governance as well but all things being equal there is some business designed processes right hey we’ve had we’re going to load this data in from Finance right there’s going to be someone from Finance that’s going to sign off and say that the numbers that we have in our reports are reflective of what we see in the actual system right so there’s someone going through some quality checks like there’s you don’t have to necessarily

24:52 there’s you don’t have to necessarily there’s not like maybe a set rule there but there there’s guard rails that are saying in order for something to have quality we’ve documented a process we put something down on paper that says these things have occurred in order to make sure that the data quality is reviewed and someone has looked at it so to me data governance is speaking to some of those I wouldn’t say they’re intangibles but they’re they’re more of like a process around around how everything

25:22 how everything Works relative to the the space of governing and deploying that governance to an organization now once something gets through that process you’ve spent money and in all cases here what you were saying Seth it’s going to take time effort money people it’s gonna cost something to do the governance so you don’t want to govern everything because why spend money on the Excel sheets that are just hanging around right let’s spend money on stuff that’s important that that I think is

25:52 that’s important that that I think is has been the learnings over time where it’s like hey we’re going to boil the ocean and create the singular data warehouse where everything is in here and and all decisions have to be made by us adding something new to the opera house right like I think we’ve evolved to the point where it’s like no we don’t need everything in there it’s a very business focused very business driven and it’s the highest priorities and objectives of an organization that she’s focused on and and tools have come a long way to to say Hey you may have this centralized

26:22 to say Hey you may have this centralized data source but you have your own data you need to bolt onto there well now you can in many different Realms right like composite models or if you have access to data and it’s unified within the organization there’s actually many different paths for you to like start to merge or write a query around the same cleansed data to add the value to your reporting outside just a specific problems that’s also part of the adoption roadmap too where they talk about the data governance their role is to

26:53 governance their role is to highlight or prioritize what initiatives need to get done yes yeah the the center of excellence is positioned very well to listen to the business whine about where problems are right right across the organization right seriously right about what wine complain voice The Voice their distaste for like whatever there’s problems there’s issues all across organization that team is trying to get people access to things they need to do to do their daily job

27:23 they need to do to do their daily job again I see them as not just access too right it’s yeah they’re an enabler they’re enable yeah they’re the enabler because they’re going through going whatever our biggest hindrance right now well if we want to have this sales quota dashboard or people look at that we need X Y and Z okay is that a priority okay we’re gonna put that initiative out to do that yeah but and but honestly though I’m push I’m I wanna or as I think about this more in relation to the article why couldn’t this data governance team and this has been like our Holy Grail too

27:53 this has been like our Holy Grail too we’ve talked about too is creating a data quality dashboard right or that that idea of having that ability to see what other projects or where are pain points from our data quality we don’t have that and it’s really hard to do it’s not an easy thing to do and where do you start but and I’m thinking of how can data storytelling affect that data governance team like what could they do to actually enhance their ability to make a decision right I right I you keep driving and driving and driving

28:25 you keep driving and driving and driving at the data storytelling driving backwards and I don’t understand why well so according to the if the article is going to say their symbiotic it means that data governance is not Downstream or or storytelling is not Downstream to data governance right or it means one has to affect the other two they’re great but we’ve like we’ve addressed I think anyway I could just be missing like like the direction here sure they are they are absolutely linked because you can’t

28:55 are absolutely linked because you can’t have a story without the data right right and the better and more quality data you have the more accurate and accepted the insights and and things you’re going to be pulling forward from the story I can’t have a story without data data I can’t I can’t do governance like I I don’t know maybe I’m just missing it so it’s probably me it’s probably so tell me so what are you saying then Tommy then you’re saying that from the data

29:25 then you’re saying that from the data story you will now change your governance strategy or add things around governance well I think that’s the question because we’ve always focused on that you cannot have storytelling without the governance right I’m not saying I don’t think it’s mutually exclusive I think I think you can get I think there’s a I think I liked an article he started talking about trust trust I think is a very important element here about what’s coming out of the information right if I can trust the tables or the model or the information

29:55 tables or the model or the information that I’m giving and again one of those things that we can do to instill trust in the business is we can give the business so there’s a let me see a lot of thoughts here hold on let me consolidate these down when we start looking at Power bi there is multiple layers of depth we can start exposing data to the business right I can give them a report and say Here’s here’s your data behind that report is a data set behind the data set is tables right and in either one of those areas you can have localized business logic being applied

30:26 localized business logic being applied and manipulating the data such that it

30:29 and manipulating the data such that it makes sense to the creator of that data set or table Upstream so what you can do to the business is say Okay instead of just giving you a final report let’s expose something a little farther Upstream okay now businesses you’re going to get access to the data set again there’s still another layer or gap between the tables and and the data set now but they’re getting closer to where the data is being originating from and then there’s one step even further in saying okay well hey Mr business user we now have technology that’s going to let you see all the tables hey what we’re

30:59 you see all the tables hey what we’re going to do is we’re just going to go to the reporting system right this stuff that runs Our Daily Business we’re going to copy those tables and we’re going to drop tables into a tool hopefully power bi or lake house or something there hey I’m just gonna I’m gonna just work to get the data out of the source system and I’m just going to land tables of data for you there so there’s literally no Transformations occurring between what is happening in the Enterprise production system and then what the business can see or use with data now there is a skill around taking that raw

31:30 there is a skill around taking that raw data and being able to shape and form it into the story that we want right you into the story that we want right that monthly sales report those know that monthly sales report those things that we were looking for but as you go further back Upstream you’re removing more risk to to what could be interpreted as a mistake or changing the quality of that data does that does that make sense what I’m saying there yeah yeah and like I said I could have been wildly off because I’m in my head I’m thinking storytelling has dependencies on governance but is there that the other

32:00 governance but is there that the other way around and like I said we can parking lot that but but I was trying to explore that and I would I would say my initial thought based on what I’ve read from the article I think it’s to me it’s more around the importance identifying the importance of data governance and knowing that there is a there is a related related terminology there’s a related there’s a relationship between the governance and the and the storytelling I’m not seeing

32:30 the and the storytelling I’m not seeing from the article saying because we have such compelling storytelling we’re now able to push that back up into Data governance I’m what I’m what I feel like it’s saying to me is or how I’m interpreting the article is it sounds like the same these two things are very interrelated right so if you have a very strong storytelling if you have a very good storytelling be considerate of what your data governance looks like as well because you you need both of these things

33:00 things to think about them holistically right right one of the things I found that was the most challenging for me in in the business is just getting people to own data in the business like that is incredibly difficult I feel like for what I was working in and maybe that was because leadership needed to take a little bit of a different stance on what needed to be done in order for people to own portions of data but when you start talking like Master data cost per product who owns what pieces like if you if you think about a product you sell there’s likely

33:31 product you sell there’s likely different teams that own portions of data that relate to that product now that can hit to the heart of something because if you own your data you’re going to Care much more about the story aren’t you I would agree I think I think there’s a skin in the game story here like I I want the business to kind I I want the business to take ownership of things and I think of take ownership of things and I think oh man this is actually a good point you make Tommy right my analogy earlier of going further Upstream from report data set tables the further up I go I’m actually handing

34:02 the further up I go I’m actually handing off to another team there is more responsibility in your hands now you need to do a better job looking at or analyzing the data and and you’re going to have the responsibility of building your own data models so if you build a crummy model that’s super slow and not efficient and calculate the wrong things that’s on YouNow right that’s not my responsibility the the ownership has shifted away from that Central I. T organization potentially and move more into the business and maybe the business wants that maybe the business is going

34:34 wants that maybe the business is going to say we’re going to hire people who know how to do this stuff yeah but you say that so as we’re talking I’m thinking you say maybe it’s shifted away from centralized that t well it already has in many cases because data is stored in many other locations that’s right teams don’t have control over anymore agreed third-party systems are a huge one right yeah we’ve gone through that conversation we’ve gone through that one before as I’m as I’m cycling through like my own question of like data mesh Like Houses I

35:05 question of like data mesh Like Houses I don’t think it matters because I think I think governance actually impacts whatever strategy either it’s the owners of the data source being centralized I. T or the owners of the data source is being business units there there’s a yes they own the like the quality of data that’s going into the source system but it also means that we still have to do the checks and balances on our end right like we’re still we’re still saying identifying that there are problems with data we’re not just going to arbitrarily you

35:35 not just going to arbitrarily you wouldn’t develop a process that would just arbitrarily say oh because it’s coming from this Source we believe in it right like you’re validating that everything coming through meets the criteria by which you would expect it to from the business right so I I guess I’m answering my own question or would would appreciate your Insight is it in reality governance is is more than just one team it’s actually like if we’re talking about Quality data coming through that governance process it can involve and

36:07 governance process it can involve and should involve more more than one team regardless of whether like regardless of the architecture of of how you’re utilizing this I would agree a thousand percent yeah even with the without even with handoffs right like I’m still responsible up until that point of the integration side if I’m responsible for the data coming in well I might be responsible if anything if anything the way I see this is there’s a lot of processes that we are we have that validate data and if you’re failing one of our first inject like absolutely it’s our like we can create monitoring or

36:37 our like we can create monitoring or that checks those checks and balances to let the business know hey there’s some bad data that you guys have now right you may not have known that it was there I’m identifying that for you and you can go quickly fix it and we always forget about the responsibility of the input of the data too we’re always like well you need to do this transformation yeah yeah and I think that’s well and that goes to the story too and I’ve seen that in my own my own career where we’re like oh well the data is bad the transformation is bad it’s like no you guys don’t own your inputting of your opportunities and

37:09 your inputting of your opportunities and because no one’s checking and balancing that and if you want this to be accurate if you want us to tell a story well we need to actually have this these little loud values not the opportunity is I think it should work thing and that ability yeah and I think that’s going to Mike’s Point too also yeah once you start owning that you’re going to care a lot about what you see in there rather than nah the data is wrong it’s like well no you put it in wrong but and this is less like there’s

37:39 but and this is less like there’s some political track going on here where you can you can start butting heads between teams if I think this is the if if you don’t have good policies right or an agreed upon structure within the organization that hey the governance team the central bi team the Coe like the group of people that manage data control what we do as an organization you follow them right because and I think that’s the importance of having

38:09 think that’s the importance of having this framework because it’s not a finger-pointing game it’s no this report is wrong or this data has a problem in it the responsibility of the owner of that data is this person right go go fix it right it’s it’s just it’s it’s more of the racy of like who who owns what so that we can fix fix what we need to as quickly as possible right it’s not a Blame Game it bring it brings everybody up to the same standards within the organization saying data is a first party citizen we want to ensure that

38:41 party citizen we want to ensure that certain data within that that system is working at Peak proficiency it’s all of our jobs to ensure that that is the case and we all have different parts to play in that that’s a really good point because I’ve seen a lot of places where they there’s so many technology pieces around it like well in order for us to do that we have to input it in three places or do four or five things it’s like okay then let’s take a step back if we really want it then we have to change this process and it’s almost outside of the world of data governance at that

39:12 the world of data governance at that point and just policy and process of that piece yeah yeah policy and process is data governance business for the business on just how they conduct themselves on a day-to-day basis but but you’re right though Tommy I think a lot of times we don’t we don’t so so there is a Miss in people’s minds I’m just going to go I’m going to go use these tools and there’s like I’m gonna go buy this third party tool I’m gonna go buy the Salesforce thing I’m gonna go buy whatever these things

39:42 I’m gonna go buy whatever these things are there’s a whole bunch of tools that are available to you and now with this proliferation of that technology it’s so easy to sign up for a software as a service and get access to stuff and start putting data somewhere so because of the ease of that there’s not this complete mindset of oh by the way I need to get the data out of this thing I need to make sure it links to my other data that’s product master or some other information so to enter in the right product SKU or number or there’s data that probably needs to be added into these other systems so that

40:12 added into these other systems so that way when data comes out of them you can correctly join it back together again and I can’t tell you the number of times that just gets cobbled together with a well we didn’t really think about the data strategy of what data is going out to systems and then what data is returning and therefore we have this super cobbled together like scenario of we don’t know what anything links together there’s no no relationships between data and now he’s have all this area in our world that’s it’s a it’s a

40:38 area in our world that’s it’s a it’s a data has no home right we we needed to build relationships between this old enterprise system and this new system it’s the same product it’s in two different systems it has a different number or SKU or identifier or whatever that is you can’t synchronize them so someone has to go through and Link things together there’s a lot of that activity and again just making sure that there’s a strategy around that that way when you incorporate these systems someone’s asking the question yeah you want to turn that on no problem how are you going to link that data back to what we’re doing already so that way it’s not super difficult to get it joined here

41:09 super difficult to get it joined here one other point you guys were talking about here and I want to just propose an idea here is it sounds to me like at some level there was a little bit of conversation between the chat thread and what you guys were talking about was there is this concept of the data catalog catalog and I think set to your point right when you when you have data coming from any system right so people are doing things with data with data the number of times I find something in an organization where there’s two of the

41:40 an organization where there’s two of the same thing there’s two of the same measures in different models or we’re trying to calculate the same thing with across data models there is this concept of like the data model or the the data catalog of what everything has where all the data lives the tables and information that we get as an organization as Tommy and I were playing around with fabric this last week I felt like Microsoft is making a move here with fabric and and what they’re

42:10 with fabric and and what they’re developing next is this idea of domains and domain specific information this idea of the lake house and one Lake this makes total sense to me I really think this is a great way of thinking about organizations but what now is happening is we’re not getting we’re potentially getting into a place where it’s a proliferation of a lot of data artifacts that are inside an organization that are more easily discovered to users so being able to quantify and catalog what’s in these tables where do I go look for this information what’s the quality like you

42:41 information what’s the quality like you information what’s the quality like there’s almost this concept of know there’s almost this concept of we have things that are certified but how can users feed back to that team and let this data source is missing these columns or I can do 90 of what I want but I’m missing a little bit of information here can we add this and what does that look like to a backlog and so I feel like there’s a story here in this scenario where we’re talking about yeah in order to have a good data story in order to have a good data governance program program part of your strategy should be around

43:11 part of your strategy should be around cataloging the things you’ve already worked hard to build and and how to share that with people so they can discover what is the data they should be running from yeah I love the idea it’s I I think even part of governance right or a strategy of a central bi team is not just to create centralized data sets for specific reporting it’s to enable business users to access data and data catalog is one of those things where you bring the context of those curated data

43:41 bring the context of those curated data sources right to a library or list where people can go and access that that data directly and especially in larger organizations right like you’re not going to have the Excel file that’s like here’s the link to yes exactly here’s the link to my Excel file on my OneDrive that you can go get I I want to agree with you so badly like I I really do but

44:13 [Laughter] but I I really think Seth harped on something that I think is really no matter what the technology or the interface or the the whatever you got in front of you is if people don’t have the skin of the game or direct consequences of or benefits of the their governance or ownership then I don’t think it matters with the catalog or whatever you have because if I don’t if it doesn’t affect me like truly in my role then

44:43 affect me like truly in my role then it’s not going to matter what I see on the screen it’s going to affect bi but it’s not going to affect the change unless it actually like you become you have the hot chair so to speak if something’s wrong and that’s actually going to affect me in my role then the data catalog is just there well again I think it’s an idea of like Discovery right so there may so again I don’t know this is the challenge I think I face in the organizations I don’t know what I don’t know like I don’t know like there’s so there is visibility problems yeah like and I’ve even seen

45:15 problems yeah like and I’ve even seen this same way for things that we built like at Power bi tips right we have this these great theme generators we have these great tools and stuff we’re like Seth and I are like yeah this thing’s amazing all right let’s write our blog post great we did it blog post is out let’s walk away and then we come back and be like well I they didn’t go to the audience and well we do one thing and then we still we talk to people like hey you still we talk to people like hey they’re like well we’re building know they’re like well we’re building this theme it’s very difficult have you seen our theme generator oh no we haven’t seen it yet like so there’s this concept of like there’s a discovery there’s if you’re not in the know you

45:45 there’s if you’re not in the know you don’t know so like what what governance things so instead of training people on or maybe even bringing to them here’s the data set or here’s the data sets you have access to maybe there’s a broader topic here that says okay we’re gonna communicate across our governance strategy here’s what we’re doing we’re going to start centralizing and locating things together and where I where I’m trying to push on this idea a bit is it feels to me like since power bi is trying to be that Central area for data and now that we brought in data engineers and data scientists also into

46:16 engineers and data scientists also into this ecosystem we need a better method for for the one leg data Hub right we need a better method for be able to apply domains to a finance or sales team and collecting all the workspaces together that that data they build comes out of right we need a better process for communicating what is certified and how to discover what is inside a certified data set so to me I’m thinking like Microsoft is going down this path but the data cataloging is so important here

46:46 the data cataloging is so important here it helps you identify who is an owner it helps you identify your descriptions and how data is supposed to be used it a lot of those policies we’ve just been talking about I think become written out and and communicated via the data catalog and so that’s where I’m really excited about the unity catalog thing from databricks I know it’s not a Microsoft product but I think Microsoft is catching up here yeah and our Dream day will be watch me go get data from a source system watch me build a lineage from SQL Server through

47:16 lineage from SQL Server through bronze silver gold medallion architectures into a data set into a report into a visual with a measure attached to it that’s that is that is the story how can I give users full visibility from this measure came from here and walked me back up the story of where that data came from originally because that because that’s incredibly powerful show right and that’s I think one of the things where data catalog falls apart is is it it’s a Documentation Center right that

47:46 it’s a Documentation Center right that has to be maintained in a cut however that would be significantly different than having a framework like fabric provides where yes I have my domains but like Can I toggle that this one’s going to be available to for people to view right yes and then they can see the lineage and then they would be able to exactly exactly maybe there’s a summary or a blurb of like hey this is what this thing does right and this is a part of our production system so I’d like I don’t have to do anything above and beyond that yeah I love the idea and yeah I think there’s just a a whole lot

48:17 yeah I think there’s just a a whole lot of different features and capabilities that that the architecture that they’ve laid out within frame fabric is really appealing appealing before we run out of time I do want to poke on one part of the article that that did get me we we did talk ad nauseum in episode 135 about the buzzword storytelling right like is there like the governance and storytelling like we’re encompassing

48:48 and storytelling like we’re encompassing all Insight reporting visualization like Etc contextualized contextualizing data and providing that to customer end users where storytelling is we were talking about is a very specific to one part I think that he talked about which was infographics right like it’s four presentations it’s very narrowly focused whereas a lot of the reporting that we do right is much more related to like providing

49:20 much more related to like providing insights in reports and it’s not really storytelling or is it well we have the great episode I think that really distinguished that with the story framing framing and I think that because how many organizations internally are doing infographics I don’t know internally internally right or are building a report for a particular purpose maybe four times yeah yeah you’ve done this before yeah but is it really a infographic I I don’t know

49:50 it really a infographic I I don’t know at the same time the board the challenge between but the challenge between this is like if you spend a ton of time and you are analyzing data and you’re trying to find an answer and you br you’re bringing insights to a a high level meeting or you’re trying to convince somebody of something you’re absolutely building a storytelling report that may not tell the same story next week that’s true and that it changes the story may be different there’s a purpose behind storytelling we talked about that at

50:20 storytelling we talked about that at length in 135 which is why not every Point why I’m poking here right I’m not no I’m not saying it’s like a negative thing it’s just when I talk about bringing like data governance and quality to reporting it’s all reporting it’s not just a specific realm yeah storytelling and not every report is storytelling every report should provide insights that are actionable but there’s a lot of

50:48 that are actionable but there’s a lot of analysis that goes along with a lot of these reports understanding the framework like your business area and gleaning insights from them and then telling your own stories from those reports that’s a great point because usually those infographics or we’ll say infographics but are not just like giving me a State of Union it’s usually some initiative like the the e-board or the C boards asking for and it’s a very targeted slide one slide two or page one page two is because each one is a

51:18 page two is because each one is a request that they wanted to see and my gosh competitive the governance better be there right but yes yes exactly yeah and that’s a great Point too for your storytelling to highlight to your C board where they’re where they’re gaps but but again that’s few and far between like that’s not your day-to-day nor is that most of the department stated why I’m saying yes there’s a distinction buzzword of storytime well I I will but

51:52 buzzword of storytime well I I will but to me whenever I hear storytelling in an article now I just think of the article about story framing about like that’s our role people just think stories and that’s what I think he’s saying too like it we’re all using that one yeah yeah there’s a there’s plenty of nuance in what he’s determining or right saying is storytelling I’m right just saying we spent a whole episode oh yeah and I’ve been one of my favorites yeah we’ll be I just think the idea of being able to frame the story because you have

52:23 able to frame the story because you have a report that you’re not changing every week to your to your point it’s not like it’s a report that’s refreshing where it’s not going to tell like this distinct this bar chart or this particular campaign it’s meant for this self-service okay I’ll stop talking Mike so so good transition yeah thanks sorry I’m done Mike your turn that’s the torch I was off the guard rails yeah I I really like this conversation and I again I don’t think there’s a a silver

52:53 again I don’t think there’s a a silver bullet yet I don’t think there’s like a single like do this one thing and everything will be happy with data governance and and your data quality challenges I definitely think this is a an organization by organization stance that organizations need to figure out what time level and effort and I I always draw I draw a graph for people on people on a a wall here when we always talk about like a hundred percent data quality all the time so I would draw basically an asymptotic curve for those who know maths

53:23 who know maths it never reaches the line basically it keeps exponentially growing closer and closer and closer but never really gets there and I said this is the graph of 100 data quality on the x-axis and then money spent on the y-axis right you could just keep spending money and it gets that last threshold of where data is a hundred percent quality becomes incredibly expensive and there’s always effort or time or money being spent in order to keep the data clean and quality

53:53 order to keep the data clean and quality so organizations have to really sit back and say what level of threshold are you willing to accept in data quality in order to get your reporting done right certain parts of the business there are a little bit looser with the numbers right Financial sectors are probably much tighter with the numbers they don’t want to be losing your money oh yeah well we we lost 10 percent of your money but that’s okay it’s not a big deal we’ve got it somewhere we’ll just go shake the couches and hopefully some money will pop out right so there’s

54:23 money will pop out right so there’s definitely industry specific things that but with that higher data quality comes additional cost and so I think this this concept here is very relevant organizations have to decide where they feel like they fit and how much money they want to spend towards this with this I went back to chat GPT and I I this is our chat GPT segment time I went back to chat GPT and I asked them well I asked chat tpg two things I said summarize this article and gave them the link to the article and it did an okay job summarizing the article however right behind that I

54:53 article however right behind that I asked a second question actually this is Bing chat was what I was using and I said how does data governance impact data telling stories and it gave me answers answers and then right behind it it gave me the link for more information link here and when I clicked on the link it went back to the same article that I was just on and highlighted like literally the sentence or two that it used inside the I there there is creepy good things happening inside chat TBT it knew me it

55:24 happening inside chat TBT it knew me it knew it asked me to summarize this article it knew I was referring to this article and then the secondary question it used the context of he just asked me to summarize this article and then a detailed question about it and it went back to the article and found the answer to the question and put it back into the chat GPT as Bono points I don’t know I thought it was going to take it so well to catch up or soar poorly on us or not maybe like part of the article my prompting was so bad so the question really was how does data governance

55:54 really was how does data governance impact the data storytelling and it basically said according to the article data governance ensures that the use the data is using in the storytelling is accurate reliable and consistent and trustworthy we’ve been beating that up the whole time here is yes we definitely agree with that part the data governance helps address those trust issues and then also says that data governance also ensures that the data for used for storytelling is compliant with relevant regulations policies and ethical standards standards interesting I don’t think I don’t think we talked about that point but I would definitely agree those are relevant

56:25 definitely agree those are relevant talking points around why you have data quality in place the data governance also enables the Storyteller to access and use the right data for the right purpose and the Right audience and I think that we were touching a little bit on that as well right I get the right information in front of people so they can do the proper analysis and I do think that makes sense I think that’s also is speaking a bit more towards we’re also speaking a bit more towards the idea of data cataloging and actually

56:55 the idea of data cataloging and actually having a place where people can go find information that is certified ready to go use this information start here kind go use this information start here methodology of methodology then lastly is data governance Fosters a culture of collaboration and communication among data stakeholders I liked that point actually and I think we were talking about ownership and who owns the data and when there’s a problem in the data and we we actually talked in our conversation around well what happens when the source system people are entering in bad data like we get the data from the system

57:26 like we get the data from the system it’s wrong we either clean it ourselves or we go back to the original system and say go fix it it’s actually wrong in the source system it needs to be corrected so by digging deeper into these stories potentially I think maybe just Tommy says what you were talking about as you dig into these stories you find there’s really issues or problems in the data that wasn’t caught which then the story you’re trying to develop needs to be cleaned which means you now funnel that back into the data governance policy that says hey I found

57:56 governance policy that says hey I found these 10 products that have inappropriate or not right descriptions or the data’s literally wrong someone needs to go back in and fix it because I shouldn’t be applying the Band-Aid after the fact we need to go back and apply more quality to our information further Upstream so that way when I build my data story it’s now correct right does that that’s pretty much exactly so what I’m going to do from now on is just communicate to you internally what I’m trying to say just send me yeah yeah just send me the just send me the chat GPT I’ll just prompt that you’re

58:26 chat GPT I’ll just prompt that you’re using so I can just use that instead and I’ll I’ll probably understand this is PT isn’t it Tommy like you you are prompting all of your random ideas this is not even Tommy this is AI generated Tommy he’s been sleeping the whole morning well every time you prompt something it’s Tommy behind the scenes typing up

58:56 it’s Tommy behind the scenes typing up an answer for you very busy yeah well Tommy GPT Tommy GPT yeah maybe maybe one time what we should do here is we should we should prompt chat TPT at the beginning of the episode and we should ask it a question and that’s the only position we’re allowed to have for the entire podcast is whatever and we have to keep prompting chat PT for some more answers to give us more details on how to speak our points and talk our stuff hugely popular episode I’m sure Chachi be Tommy I like

59:26 episode I’m sure Chachi be Tommy I like it I like it it’s good all right well we’ve done it again we’ve wasted a pretty good well you’ve learned no wasted you’ve wasted a pretty good hour of your time hanging out with us at the explicit measures podcast we really appreciate your listenership the chat has been on fire really love it Chad thank you everyone who participates it makes it super fun and really adds a lot to the conversation so we really appreciate you jumping in taking time out of your day or evening or night or wherever you are in the world thank you so much for joining the conversation we really enjoy it with that our only ask is if you like

59:58 it with that our only ask is if you like this conversation if you felt this was a good topic today and maybe pulled some nuggets out of this and liked the article that we pointed out today please go share it with somebody else let somebody else know you like the podcast share it on social media or someone if you go into the office anymore I don’t know if people even do that anymore so if you see people that are not on a virtual screen let them know you found this podcast interesting and there might be some additional nuggets they may want to get out of it thank you all so much and Tommy where else can you find the podcast well it’s Christmas in July and we have three gifts for you one

60:28 July and we have three gifts for you one you can subscribe to the podcast on Apple and Spotify too you can have your question answered by us at the mailbag go to powerbi. tips the podcast and your last gift is watch this live every Tuesday and Thursday 7 30 a. m on all power bi tips social media channels not sure these are great gifts these are like can I re-gift this one I think I need to give it to somebody else I’m gonna I’m gonna bundle this up and give it somewhere else I’m not sure if that’s the right thing but anyways I thoroughly appreciate it thank you all very much appreciate your time have a great day we’ll talk to you later

Thank You

Thanks for listening to the Explicit Measures Podcast. If you enjoyed this episode, share it with a coworker and drop a question or topic suggestion for a future show.