Types of Data Roles – Ep. 252

Data teams keep getting new titles, but the work still comes down to the same fundamentals: getting data in, shaping it into something trustworthy, and enabling decisions.

In Ep. 252, Mike, Tommy, and Seth cut through the ambiguity by talking about roles in terms of outcomes—who owns ingestion, who owns business logic and metrics, and who owns the last-mile reporting experience. If your projects keep stalling because “someone else thought they had it,” this episode is for you.

News & Announcements

• Types of Data Science Roles Explained — A quick reference for the most common data roles and how responsibilities typically split across analysis, engineering, and science.
• Tips+ Theme Generator — Standardize colors, fonts, and visuals so your Power BI reports look cohesive without rebuilding every page by hand.
• Subscribe to the podcast — New episodes, show links, and the full archive in one place.
• Suggest a podcast topic — Send a mailbag question or situation you want Mike, Tommy, and Seth to break down.

Main Discussion

This episode is a reminder that most analytics pain isn’t caused by DAX, SQL, or refresh speed—it’s caused by unclear ownership. When everyone is ‘doing data,’ critical work like modeling, security, and operations becomes accidental.

Takeaways you can apply:

• Assign owners to deliverables: ingestion, storage, transformation, semantic modeling, reporting, and operations all need a name next to them.
• Separate data movement from data meaning: pipelines move rows; models define logic, measures, and business definitions.
• Make the semantic model the center of gravity: a governed model enables many reports and reduces one-off dashboard chaos.
• Expect overlap—but keep responsibilities explicit: on small teams one person wears multiple hats, but the handoffs must still be clear.
• Treat security as design work: RLS patterns, sensitivity, and access workflows are part of the role, not an afterthought.
• Hire for the bottleneck you actually have: if refresh/pipelines are fragile, you need engineering ownership more than ‘another report builder.’
• Use titles to communicate, not to impress: pick role names that help stakeholders find the right person for the right problem.

Looking Forward

Pick one current initiative and write down who owns each deliverable (data, model, security, and report)—you’ll uncover the next most valuable hire or upskilling target immediately.

Episode Transcript

0:29 good morning and welcome back to the exposit meure podcast with Tommy Seth and Mike hello everyone good morning and a good morning to you Mike how are you doing well interested about our topic today should be good to go neat conversation for today any other announcements things going on all right I’ll just keep talking if you guys want to say anything so when I stop you guys can say something what we just talked about said and in between the are offline and online do you think there’s how much data do you think’s in the bus tracks or

0:59 data do you think’s in the bus tracks or or for the bus routes your guy apparently is on the dot every every minute they’re on if it’s they’re showing up at 637 do you think they record to track this and how important is that oh that would be interesting if I were a bus company I would want to have something tracking where the buses are do they do that I don’t don’t know but yes our normal female driver is

1:30 know but yes our normal female driver is bot on man like they are traveling they are traveling with the most the world’s most valuable cargo they are likeed our future is basically on those buses I was like I don’t think buses are that expensive but the bus is not expensive but the cargo that they carry is extremely expensive this is very true so you think there would be a lot of Statistics there because my bus driver was very late today no problem they’ve been great but we were a little late today

2:01 we were a little late today yeah I don’t know I I haven’t had the I do know that there’s been mean I do know that there’s been instances where there are cameras definitely on installed on buses now and so if people must behave they can be retroactively looked at and figure out what’s going on on the bus so there is some tack on there whether or not those camera add-on things actually contain like a GPS that ping something or tracks stuff I don’t know it I feel like it would be very lightweight be able to have a Tracker like that is someone actually analyzing the data and doing anything of value with it no no

2:33 anything of value with it no no if anything it’s just they’re capturing it it’s one of these yeah someone’s not really and then and then you have to think about like how how accurate is that GPS because are they buying a little bit here because this this this brings me a different thought right where if if the GPS in these vehicles is so accurate then it shouldn’t take me like three to four times to call FedEx and tell them that they’re delivering packages across the way not even on the right street but

3:03 the way not even on the right street but somehow they think they’re delivering to the right address to the same house that’s not my address it’s not on the same street but it’s nearby me if my free application will always get me where I need to be yeah that’s a good point that’s a great point which which which begs the question do delivery drivers I love to see this in the com do delivery drivers have the GPS and that’s where it tells them to go or know the layout so well of oh I’m I’m for if it’s an Amazon

3:34 well of oh I’m I’m for if it’s an Amazon driver I’m sure they’ve got a map or some iPad somewhere or something on your phone hey I 100% Landing you Amazon’s technology stack must be amazing because there’s never an issue there never UPS no never issues all day long in London you have issues with UPS I still think I still think they’re writing them down on handn notes giving them a piece of paper and say Here’s the address drop it off here wow we’re going to get ourselves in trouble here we will get ourselves right all drivers are great I I’ll side note this and we

4:06 great I I’ll side note this and we get to the meet of this only because you’re going to really appreciate this and it’s going to be the most Tommy Thing Ever After College I did not want to I didn’t know what I wanted to do and I wanted to find myself or whatever just figure out something someone told me you can buy buses like school buses for incredibly cheap like when they’re gonna go travel the world aren’t you but here’s here’s the kicker obviously I had no money because but what apparently they said you could do is you can convert the ethanol the gas fuel to like peanut oil or it’s a diesel

4:39 fuel to like peanut oil or it’s a diesel truck or normal like food oil basically oil you can go to McDonald’s and say hey give me your old oil they’re like what I was gonna do is go to Chinese buffets around the nation because they have to pay for the oil to some for some they do so I will take your oil give me a meal I will take her oil well I’m paying for Chinese no no no no no they’re going to give me a meal and I’m going to take their oil that’s what I’m saying they’re still paying they’re now paying the cost of a meal as opposed it’s not really like it’s not really at the end of the day how how much leftovers are there so just

5:10 how much leftovers are there so just takey give me your leftovers that’s F yeah I didn’t obviously I didn’t plan this out completely when you are you g you gonna try hitting them right as they’re closing all the time was that the plan or yeah are we reading too much no no no it’s gonna be yeah it’s always gonna be at night when they’re about to close go don’t don’t over there I’ll I’ll take it for you as a good samaritan I can I can take all your leftovers can you imagine though converting a bus and living in that that was a dream that was the dream that was your dream not gonna

5:40 the dream that was your dream not gonna lie I’ve seen I’ve seen some Renovations of like Vans and buses that pretty amazing F have done online yeah it’s crazy it it would be slightly annoying because I believe for Tommy I think you and I would be fine because we’re short enough that it we could walk down the aisle of the bus without hitting her head on things which I think would be fine but I’m not going to lie though I I definitely did find when I was a kid growing up I found a Dunkin Donuts that would throw away Donuts at

6:10 Donuts that would throw away Donuts at the end of the night and I knew to go in five minutes before the end and be like I’ll have a dozen donuts and any other ones that you want to throw away I’ll take those too and I’d come home with like three dozen donuts would they for the price of one oh they just give them to me they’re like we’re going to throw them out anyways I’m like I’ll eat them anyways you’re extending that good gesture right we’re like I’d like a dozen Don so you’re going to pay for a dozen right so that I think that like curbs the threshold of like you have any free donuts you want to give away yeah

6:40 free donuts you want to give away yeah you don’t just walk in be like hey yeah take your trash out for youday I know I know yeah that’s what I was doing basically here’s the trash bag yes exactly anyways all that wrapped up let’s get into today’s topic good good topics for discussion it it’s funny I’ll this is let we get going here it’s funny when you are a data person and you look at the world and you see data everywhere and you’re like oh that’s that’s a cell phone interacting with a

7:10 that’s a cell phone interacting with a register cash there’s got to be data being generated where’s that data going there’s some there’s there’s a you’re checking out at the store and they’re checking every scanning every item yeah I know there’s that thing’s creating data somewhere I wonder where that’s going wonder what they’re doing that and as I’ve interacted with a lot of companies odds are there’s a lot of data being generated but it’s probably not being utilized very well and it’s just sitting there or getting collected or even just being generated and staying around for a little bit of time and then disappearing there’s missed

7:41 disappearing there’s missed opportunities with that data anyways I just see that all over the place I fig it’s the there was a movie that I see I see data and it’s really I see data people yeah what movie is that from six cents the six cents yeah I have the I have the seventh cents that comes from data I see that I see I honestly think there there’s a link that I completely agree because that’s what happened I’m literally walking back on my walk with my daughter from the bus I’m like well why would it be 15 minutes late is it when they left is it they have to

8:11 it when they left is it they have to have some tracking do they get in penalty for that and do they have a rate the same way Jerry Seinfeld says he always does thinking jokes and like you know he and he’s and I think we’re always thinking that why like well why are you inputting that are you even doing anything with that how cool would it be if you did something with that it’s like a curse you always think in data now you’re you’re stuck with it it’s a beautiful thing some may say yes I might say no speaking of data this is actually

8:40 say no speaking of data this is actually going to be very well and open us up right to our topic very well here so the topic for today is 365 data science it’s talking about career advice different types of data science roles explained so this whole data this is their perspective on where they see data science and I would argue there’s a lot of language in here that’s not just data science it’s data science and a lot of other data roles that go along with data activities let’s

9:10 that go along with data activities let’s call that so the article link will be in the description of the video so if you want to follow the link go ahead and hit the description of this video that’s where the link will be I think it would be probably best just to maybe rattle through all the names of the different data people and and give maybe a brief definition based on the article says and then we can discuss what I’m interested in seeing is these are the roles that people are identifying around data where do these roles fit inside fabric which of these roles do we think would come into this

9:41 roles do we think would come into this fabric ecosystem at this point so let’s we’ll go through the list first on the list here is a data strategist the data architect a data engineer a data analyst the business intelligent intelligence analyst the data science scientist the machine learning or the ml Ops Ops engineer the data product manager and that’s those those are the main key areas here let me start with anything stand

10:13 here let me start with anything stand out to you guys initially anything you want to we’re going to start things off on on here I here I think go ahead Tommy there’s two main things that took out in terms of one the First Data strategist and then two the need for this in in our world and we’ve had we we’ve talked about this in Spades where if you were to say hey I need a powerbi person well there’s not really levels to that we have the only thing and we always refer to it is cess skill

10:44 and we always refer to it is cess skill matric but that’s not necessarily a uni unfortunately not yet a universally adopted approach towards hiring and when we think about I what type of person to I need I think we need to get more and more granular on the types of roles that we’re looking for and those skill levels okay I think first F at first glance one of the things I’m seen in here is like P the per it’s interesting

11:15 here is like P the per it’s interesting to me because I I think I have obviously a perception on the data ecosystem and all of the parts of that data ecosystem work together right the the different roles I think within here over time I’ve seen a role of a bi developer and this this spans like multiple different roles within here right because I I know my team actually like falls directly into that where we’re playing a lot of data engineering roles we’re playing the

11:46 engineering roles we’re playing the analyst role regardless of whether that’s data or business intelligence analyst and and like there’s the the report author like the PE the building sharing reports Etc so I think those roles right off the the top I think are are missing and then two it introduces a different thought I I don’t think I’d had before around like oh okay so from a data science perspective is this how they view the

12:16 perspective is this how they view the world right like there there are it’s very data science specific data strategy data architect data engineering analyst analyst data scientist engineer manager like there’s a lot of independent disperate roles that I I technically don’t know whether or not are always aligned within organizations because I don’t see a lot of these roles so first first thought second

12:46 roles so first first thought second thought this is overall a very structured theoretical approach to business yes versus reality yeah you business yes versus reality yeah and that’s not a criticism but if I know and that’s not a criticism but if I look back at the author Alexandra it it rings true because there’s a lot of Education behind there and a lot of writing Etc where this falls apart I think in my mind with reality is businesses are in the business of getting it done not in the in the mode

13:18 getting it done not in the in the mode of getting it right right so there’s no such thing as I have never seen such thing as a data strategist or a data architect that start with the idea without engaging with the business it’s always a business driven initiative right and a long ago I think we got ourselves in the these situations where the Technologies were not up were not fast enough so you would find yourself in Realms where you had technical people going off and building a data thing and

13:49 going off and building a data thing and coming back to the business and the business going okay like it’s 25% useful but you need to go build this 75% now and they’re like okay well that’s that’s going to take two more years right like because we built this thing and we we apparently built it wrong and I think things have evolved into a a little bit of those different roles where I’m at where I I would say I’d also throw in you have data analysts but those would be your business people right like we’re introducing more

14:19 right like we’re introducing more business people into this mix because of the reporting tools that allow them to do a lot of this stuff namely powerbi so I think that’s that’s the other part I’d like to introduce into this I I think conversation is like there’s a shift here too because it’s not just a a well-planned approach that you go execute with all these roles behind the scenes there’s a lot of people and a lot of process to go along with the technologies that you build would you agree would you agree or

14:49 build would you agree would you agree or disagree that there’s a bit of a blend happening to where I I read the business intelligence analysts and data analysts explanation in the the article where they’re pretty straightforward clean analyze visualized data reporting where I’m I would foresee or I even foresee it now that those can’t survive as individual roles or sole roles by themselves anymore where I I am more of the opinion where if you’re going to be a bi analyst or data analyst you also

15:19 a bi analyst or data analyst you also have to adopt the engineering or the architect role to probably more to a primary part now to just be full-time yeah I don’t I don’t know if I agree with that I think I think a data analyst needs to be able to understand how to use tools use reporting tools but they’re PRI like depending on it depends on the business sure I’ll put put that first in larger organizations you could absolutely be a data analyst Barn just doing data analysis like there is a ton

15:51 doing data analysis like there is a ton of of data out there and it takes a long time to pull insights or look for you time to pull insights or look for Trends or things within the know Trends or things within the business by ripping through the same data sets right so you don’t need heavy report building skills like you would with an a business intelligence an so if you’re going to add in the tech stack part of it I think you are a stronger data analyst but at the same time those are role salary differences as well yeah right like a data analyst

16:22 as well yeah right like a data analyst is here we’re going to provide you everything you need here and you can you can rip this apart as as best you can in Excel in powerbi in these tools that we’re going to tee up for you whereas when you add the engineering skills and the technical skills on top of that where you’re building the reports I think you’re elevating your skill set you’re also elevating your your position so like I I don’t I don’t think those are the same I think it is a

16:52 think those are the same I think it is a graduation or or that you would require it but at the same time like smaller companies May because they’re not going to have positions that are open for an like just a data analyst they would need you to fill multiple positions and that’s where I would say like well then that’s a negotiated up swing because you that’s a negotiated up swing because both of those know both of those skills I would agree Mike I agree with you on this one and I’m trying to what I’m trying to do is I’m trying to take these different roles the description of these roles one of the one of the

17:20 these roles one of the one of the roles that I’ve not ever really talked about or really put a lot of thought into is is the data strategist yeah so this is a role they’re they’re putting here and it’s it’s talking about they’re there to make data driven decisions they to help them produce products and services they use these people to help improve their existing processes right so that’s one of the roles that I’m not sure I quite resonate with I’m not sure where that sits I would actually probably put the strategist and the architect together at the same level of rolling however when I

17:50 the same level of rolling however when I sit back and look at these different roles I’m trying to think about okay which if we think about our company we have three things we’re trying to affect or provide change to our people in our company the data culture the process on how we do things and the technology those are our three levers we can pull and I’m looking at these different roles and I’m seeing I feel like many of these roles are falling into one of those categories right the strategist and the architect are focusing more on the process and

18:21 are focusing more on the process and potentially investing in the people as a part of your organization the engineer the analyst the business analyst even maybe the data scientist is focusing more on the technology side of things and the ml Ops engineer yeah before you before you jump though like the strategist is doing one other key part here sure which is which is trying to create new revenue streams right and I I think I don’t it’s interesting that I like the name data strategist is something I haven’t come across like I I

18:52 something I haven’t come across like I I would would almost I I also live in a realm where we’re doing multiple different types of reporting and business intelligence yes but I think that the main driver in in this thought is well this is the one role that would be figuring out what would make the company money right Asos to ex sponsor right so I’m looking at if I’m looking at that in the company yeah well is so I I’ll that the data

19:23 yeah well is so I I’ll that the data strategies resonated with me wholly and maybe because I you heard of it I’ve not heard of the exact role but the exact paragraph resonates true not because I I’ve seen it at a company but the book that I I quote all the time and I had the opportunity to actually interview the author of how to measure anything his entire livelihood for the last 50 years has been basically that data strategist is basically going in and understanding what are people trying to measure and how to measure it there’s a whole and I don’t want to say industry

19:53 a whole and I don’t want to say industry for that but there’s a wide Gap I think in the need for that dat strategies for what are we measuring like what what’s really important no not what you’ve been measuring for 50 years but what’s the goals and what what is how can we actually then effectively even to a 90% confidence get to a point that we’re moving in the right direction he’s worked with the government he’s worked with companies in terms of like it security to say like Hey how do we know our it is secure based on our company structure and well not that might not be

20:24 structure and well not that might not be a role I resonate with this because that’s honestly what tried to take that hat on on a lot of projects of the data Tres where they say oh we want you to do this this visuals or we’re going to look at that it’s like well how does that relate towards your your goals they’re like that’s just what the companies want and that conversation we went into on well where’s that threshold we’ve talked about those pain points there’s a need for that I just don’t think that’s really been explored

20:54 don’t think that’s really been explored a lot I think it’s probably exists probably more in firms agencies rather than than internally well I say I think this varies wide I think this VAR varies widely across companies and I think I would agree that statement yes I think us as consultants coming into companies there’s a reason why we’re showing up we’re acting as that data strategist to some level because they’re basically the idea the company is basically saying to some level hey look there’s this new thing called powerbi it’s on new this new kid

21:25 called powerbi it’s on new this new kid on the block called PBI right it’s it’s helping us do some good things but we’re we need some planning we need some organization we need some direction to figure out where we’re going and I think on at at in some levels again this is probably varies very differently between companies but you’re right there need there does there does need to be an alignment on what do we think is valuable as a company where are we where are we going as a company and in this data strategist it’s basically saying hire someone to show up and say what data should you collect that way

21:56 what data should you collect that way you could use the data to add products and services and add value to people which I think all of us will agree at the end of the day that the data is probably more important than the product at in some cases like if you some if you have a if you have a product and you have more data than your competitors right anyone can build your product but if you have more data than them and you can act more intelligently on top of that data inside that product that’s why people will use you that’s why people would buy you

22:26 would buy you essentially so there there’s two things that come to mind I agree with Mike if in quotes a company has a strategy then I think those three bullet points from the strategist are things that are talked about absolutely how like defining metrics the Tommy toar point I think they this is where the business intelligence conversation is it’s it’s about creating value for the business it’s about understanding met the metrics it’s about like making sure

22:57 the metrics it’s about like making sure that the data is is accurate and clean Etc I think the reason you’re seeing in some cases and I don’t know what the ratio is of companies that don’t have those conversations are because there’s a deficiency between the parts of the organization where this there’s a mentality of just getting stuff done yeah get it done get it done get it done and there’s no there’s no strategy there’s no overall strategy of like looking across all

23:28 strategy of like looking across all those organizational units and seeing where the value is that could be sused out to maximize internal processes remove a lot of the barriers like automate things Etc like but most of the focus for business intelligence is driving towards efficiencies right getting better insights to run the business better and that’s a hard metric for data

23:59 that’s a hard metric for data leaders to to make sure that they communicate the value to the business all the time because if if you even look at this article this is saying you need two four six eight yes eight or nine people just to get things like running and that’s why I’m saying like no you don’t that’s not how business run no but the the reason I leaned into that one point in the strategist where it’s like they could create a revenue stream

24:29 like they could create a revenue stream new Revenue stream via data monetization and I think that is something that I don’t see a lot of right yeah like and some companies you’re not going to you need reporting you need business intelligence every company I would argue should invest time in some form of business intelligence where you you are maximizing the use of your data where you’re able to make decisions based on it because it’s going to set you apart from your competitors and it’s going to let you understand how well your business is doing where where I think it’s Unique is

25:02 doing where where I think it’s Unique is if you’re able to also leverage then the the data artifacts that you’ve built and created in a fashion that’s a sailable thing and you’re making money off of it somehow that’s where I see the value of the strategist come in as opposed be to being something I think is a byproduct of many people as that that are combined having the conversation yeah and the the last thing I’ll because just I really do think this is the most

25:32 I really do think this is the most undervalued or under realized at organizations because that role could be summed as are we measuring the right things and are we measuring them the right way and that’s a constant revaluation by an organization that’s not something that’s every 10 years I think it’s constantly evolving what those metrics and what they’re trying to measure are and the new data coming in and how they’re measuring it I I I really think that because everything else flows from here the AR the engineer

25:59 else flows from here the AR the engineer all that should be based right on what you’re measuring and I’d also argue here too you have you also to look at your competitive landscape I think about you competitive landscape I think about some data stories that I hear from know some data stories that I hear from very large companies right on oneand you look at the cost to acquire the data in some Industries is very expensive right so if I’m thinking like the Auto industry they every car that comes with the Cirrus XM Radio sirusxm radio does radio for Cirus stuff

26:31 sirusxm radio does radio for Cirus stuff but I think much of its profit actually comes from the idea that every vehicle that’s connected can send its data to mother Ford or mother GM or what and you look at now what Tesla’s doing too same thing right their car is literally an iPhone driving on Wheels so they you’re getting over theair updates it’s sending data back and forth they’re improving your product just with the data alone and and so this is so I will also say when you think about the cost

27:01 when you think about the cost to acquire that data depending on your company’s value or proposition to say what is data that is important to us and how can we use that in our product build you find more data is being sent from things that are more expensive right or in the web space right a website is very easy to collect data from because it’s so lightweight and so it doesn’t cost a lot so there’s there’s different costs to acquire the data so I think the strategist comes in here basically says okay let’s think about what we need to

27:31 okay let’s think about what we need to do to make good decisions within our company and how hard is it for us to acquire that information where I find there are some challenges around that data is when we start looking at like manufacturing or or things that are like very we’re producing products but we’re not necessarily connecting the the idea of what does what what are we doing either in our manufacturing process or what are those products doing so that they produce value of the data and so I don’t know I think I think it’s going to

28:01 don’t know I think I think it’s going to vary widely be between organizations but when I’ve seen this occur I think there’s a consideration for here is well how much is it going to cost us to information because at some point you say it’s too expensive to get the data and we just don’t collect it and you figure out other ways of doing your your analysis H question for you too how often do you hear in your daytoday or strategy meeting well we can’t measure that or that’s unmeasurable the does that ever come up at all I think I think the opposite occurs

28:32 all I think I think the opposite occurs I think there’s always things people want to measure but there’s how would we get the data to support that I think that there’s there’s Concepts around like we know we need to get to this level of measure it’d be really nice if we could measure this but then we we dig into it and the the data that’s coming through is not clean we don’t have the right amount of information we can get like half that way the answer but not all the way cuz we need to add some extra data to it or or what like there was no to your point Tommy there was no plan when the data was being generated it just

29:02 the data was being generated it just started getting generated and now we’re looking back at it and going oh it would have been nice if we had some categorization or it would have been nice if we added a customer ID to something or it would have been nice if we had tied this thing to something else we have data about so there’s like this whole like and I think this is part of where you incorporate like the an the analyst you start talking about like the data scientist like you start figuring out okay how do we apply effort to the data so the data becomes more

29:33 becomes more valuable and I and this do Dove tail I think I I hung this up as we were mean I think I I hung this up as we were trying to get through describing each of the roles but the strategist like to some degree I think we may be taking that or not thinking about that role a lot because we’re we’re the people thinking about it I guess and we come in and like to the point in the article like a lot of companies hire this out yep but it is it is 100% a those conversations with business

30:06 100% a those conversations with business in many regards Mike like yes they built things in this way to just service something without without the thought that oh I might need to roll this up by or this might need to integrate with this system or yes and then yes it takes a bunch of time to clean that up data analysts right like rip through everything understand how both like systems work together what are the things that are missing that are needed like there’s there’s core data issues

30:36 like there’s there’s core data issues from The Source systems that I think business intelligence people run into all the time let me give you a real example of that one right now the apis that collect the activity logs for powerbi is a prime example of this because it’s a single API that returns data from many different teams right that we have the the fabric team we have the synapse team we have like all these different teams that are making data in now power. com the data is coming back to us in very weird ways there’s there’s there’s duplicated columns that mean the

31:08 there’s duplicated columns that mean the same thing they’re different names are you are you becoming the data analyst for Microsoft right now no I’m I’m literally going through their well my hair you are you are because you’re gonna go hey Microsoft you’re dumb here and here and here and fix this and do that I’ll see I’ll see that message coming here there there’s a whole there’s a whole bunch but team’s not doing this I get a lot of notifications I I go look at like a powerbi workspace and like the powerbi data set and it has its sets of columns and then you go find okay let’s go find

31:39 and then you go find okay let’s go find a different activity that’s related to something that’s like fabric related and now there’s there’s new columns and there’s potentially columns that have in that new in the fabric elements right I’m looking at and going well that’s that’s like a one like I’m literally and that you can’t get to this point until you actually start coming all of the data speaking from experience migrating multiple different ways to do something into a singular way is not so easy not so easy yes so but this is that that

32:09 easy yes so but this is that that is like a real world example like we know we need to track it we know we need to capture datea on things but there’s not like a standard a bucket a way we’re saying there’s not a common thought of the top level again it’s the idea of like it it it this is the quintessential problem right we need to get something out the door so users can start using it well we know we need tracking so just add some metrics okay we do that and then we come back after the fact and say wait a minute none of this stuff’s matching up like we’ve got a lot of data coming in but we’ve got a lot we’ve got so much that data comes in in a very

32:40 so much that data comes in in a very unclean format and so now we need to build a strategy around how are we going to shape that data so that it makes sense and the only reason I can see all this now is because I’m building tools that will get all the data and let me query the data as a SQL table so I can go in and find okay what activity event is is this where did this come from what you what data is this user making what’s happening here so as as we’re interacting with the actual power. com environment I can get real metrics about what’s happening and what’s occurring there so that yeah that’s something that’s just very fresh

33:10 that’s something that’s just very fresh in my mind well and it goes back too you’re just collecting the data but then it’s trying to understand too what are those key items that you’re trying to focus on I just had had this conversation last week I was like well okay well what how many reports you have we don’t know well what are the top reports out there we don’t know okay get that we need to start there like you have this very large extraction of data which you said is not in a really cleaned format yes and not only are this is again this is not just the RPI this is a lot of things we deal with not only

33:40 is a lot of things we deal with not only does it not really tell you that where we need a hyperfocus or where do we want to lean but at the same point it’s like okay how are we going to measure this and then again talking to like we need we can’t actually just proceed until we have this until we know where we stand and that’s a big part where usually we’re going it’s always the other way where it’s like here’s all the data here’s the things we’re going to be doing figure out where you find Value from it right rather than going what are you trying to measure where’s the value now we’re gonna actually create the process correct the DAT is

34:11 create the process correct the DAT is nothing more than people process technology in a nutshell yes yeah what’s interesting about this strategy conversation is Mike this this brings to mind something you said in a prev episode where we were talking you had mentioned technology seems to be moving faster than businesses could react yes I agree with that and I I think we’re

34:37 agree with that and I I think we’re we’re in this world world because the the new reporting to powerbi is the tool that has allowed many more users to engage in understanding and getting visibility across systems where data from those systems is hugely important to them because be for for whatever the reasons right whether it’s an Excel documents whether those systems don’t talk to each other and

35:07 systems don’t talk to each other and they’re not integrated whether there’s no Warehouse like even if you have a centralized repository of data there’s always stuff on the on the peripheral pereral right that’s going to be that’s going to have some phased approach of getting into the main system but isn’t going to be there yet because the business continues to move at the speed to business yes so and the business to that point let me just I agree with you and in addition to this the business is continually buying other third party tools that have not been

35:38 third party tools that have not been integrated with everything else that’s what I’m saying that’s why I’m saying like the business moves at the speed of business yeah exactly if they need to if they need to solve something they’re going to go solve it and it’s a lot of times it’s with a third party tool and last consideration given the data portion of it Andor a lot of times they’re their own business unit they can make their own decisions and they don’t talk to technology they don’t talk to data people y right so you don’t get in the way of them you adapt to the way the business wants to operate and what

36:09 business wants to operate and what that forces continually and you think about our conversations a lot is reworking large systems that interact or reworking how we collect data from those systems to make sense of it make sense of it to the business so in in reality to me as I’m like thinking through the strategy component of it like there’s just there’s a big breakdown between business in the technology Parts because they’re building Technologies into the business without thinking that they’re

36:40 business without thinking that they’re building technology into the business they’re creating problems for the business because there’s no if you don’t have a a structure around these units getting together and making choices around the types of tooling that they bring into the systems they may be creating more of a problem for the business than solving their little thing so it should be part of the same conversation so I maybe I’m wrapping myself around the axle where there should be a data

37:10 should be a data strategist in all organizations but we’ just probably call it something else so I would I would agree with you I think I think a data strategist helps the conversation I think it also provides more direction for the business in general I think my argument here is that executive sponsor either should be hiring out that skill or should be taking on the role of the data strategist to some level because I think the strategy informs a lot of everything else inside this team right the strategy informs how many Architects you need the

37:41 informs how many Architects you need the strategy informs where the data engineering is doing what tools you’re because yes those those are those are things that cost money yeah that that’s where I’m going with this yes and business and the strategy you can have a strategy but if the strategy and road map is a five-year plan right that that’s going to look different from an organizational structure as opposed to the strategy and the plan needing to be condensed into one year and then you have all those like we need a rapid

38:11 have all those like we need a rapid build and then once we build all this stuff we need to maintain and and fine-tune and continue to grow like those are those are two different conversations and well you’re forgetting the other side of the coin there Mike let me lean on to what you just said am I forgetting coin or do I have the whole coin or maybe I have a dollar and you’re looking at coins I don’t know oh oh or maybe I have some Euro thing I don’t maybe I’m notbe I’m not even on the right currency way too far into the we’ done this maybe I’m using crypto and you’re still using dollars I don’t even

38:41 you’re still using dollars I don’t even know I haven’t I CED my wall for 10 years digitized it’s funny you say that Tommy because I just saw I just saw a YouTube short of someone going into Burger King and they’re like people are buying Burgers on credit and they were like it’s a news article like no one’s going to use these credit Cardy things inside the inside our it was anyways it was funny it was just like this whole cash versus credit thing anyways sorry go ahead Tommy had to be there I’m sure so so but yes yes very much so okay we’ll

39:14 but yes yes very much so okay we’ll frame it back we talked about the data strategies going through and understanding where the architect the engineer but from our conversation if we’re framing this data strategies the way we have been where they’re also respon responsible for the measurement and what we’re measuring in that other side of the coin that’s going to then dictate Downstream the policies and the process and the people at the company right because if they’re shape if they’re shifting or responsible too for what the company’s going to try to

39:44 what the company’s going to try to measure what the company’s going to try to achieve that’s going to then go Downstream to who’s your marketing team what technologies that marketing team have and also who those people are so they’re affecting the culture they have to have that authority too where if they’re going to be successful and I think I think this definitely and this is why to me the strategy is typically a team approach or multi- it’s part of the adoption road map of or a data culture road map where

40:14 map of or a data culture road map where to Mike’s Point you have an executive sponsor that at least understands that this isn’t important right like the byproduct of not having these things is very apparent in organizations that don’t have them may not may not know how to fix them so yeah like creating the the the center of excellence or Community Practice or whatever you want to call it the body of hey collectively we need to solve these problems because collectively to your point Tommy like it’s an organizational

40:46 point Tommy like it’s an organizational thing ultimately right and it may be led by a team like a business intelligence team to create these artifacts and ways in which the the organization can leverage tools how they build things to streamline to make sure that everything is is being built in a similar fashion so that they can be used and reused I think is is absolutely part of this conversation and how it it spools up to take this on a different tack though what’s interesting to me and maybe I I make this comment you guys rip

41:16 maybe I I make this comment you guys rip through but this is this this article is like the the roles and responsibilities within an organization but if you think about a green field like a brand new data source if you read through the responsibility section this almost reads like a road map in and of itself yes it does it doesn’t read like it doesn’t read like you need I if I if you would take out the names of like I have to have this person yes this this is the

41:47 have this person yes this this is the the list of responsibilities that are shared across individual like people whatever they are that make a a data solution work so I want I want to tag on that thought and I think sometimes we talk in terms of roles as if they are a person and I think you have to generalize this article a little bit and say okay there are eight roles that are probably being performed within your organization you need to prepare for

42:18 organization you need to prepare for these roles and think about okay you have a team of three or two or one how do these roles overlay on top of your existing team right and so this is where I’m going a lot with with the powerbi skills Matrix right now where I’m if I if I take these roles and say okay there is a skills Matrix people who can consume reports people who build reports people who model data there’s this thing called a release manager and an admin if you just focus on just the power VI

42:48 you just focus on just the power VI ecosystem I think those are the this the main skill roles an organization needs to plan for and they’re just what I would call may be personas right they’re not necessarily a role tied to a single person so you may have an architect Who’s acting in some way a portion of the business analyst a portion of the architect role and a portion of the strategist all rolled into one person right they’re doing a lot of the planning and a lot of the communication I use when I talk to clients is we’re

43:15 I use when I talk to clients is we’re building two Bridges and this I think makes sense a lot with this conversation we’re building one Bridge from the data side what data do I have how do I shape it so it’s ready for the report and on the other side talking a lot about the business requirements and and trying to figure out what are the insights we care about how many pages in the report what does the user need to look at to do their job and so as you build this bridge from both sides you have a lot of technical it space data analyst data engineer business

43:45 analyst data engineer business intelligent analyst trying to figure out or actually let me say that again data engineering and maybe the data scientist or the ml Ops engineer right there’s a lot of engineering work Happ happening at the same time you’re also trying to figure out how do I take that engineering work and make sure that it makes sense to what the business wants on the other side and you join the bridge in the Middle with I now have a cube that is built with measures that I understand with a with a model of my data that I can then build reports on top of and shoot Mike I’m seeing this more and more too where this is not even

44:16 more and more too where this is not even their full-time job these these rules where this is maybe from the dep department and that’s where I was just like theoretical versus reality right because if this is viewed under the lens that each of these is individual positions it’s a non-starter yeah for for like smaller and midsize companies as they grow but if we’re looking at it as for as far as like roles and responsibilities then yes there are people in the organization that H are are doing multiple different things that

44:46 are doing multiple different things that are outlined in this article yeah not even from a data point of view I’m seeing more personally in some of the projects where these people they’re their primary role is not in data they’re either operations or they’re in other roles but they’re do they’re taking on more where they would relate to this ETL and this strategy point of view but they’re not necessarily by by by title a business analyst I don’t know if I’d go that far I’m seeing more of it yes

45:19 far I’m seeing more of it yes within their business unit they’re understanding how they need to shape some data and like model it a little bit but I I think the way I think about this would be much larger organizational efforts not necessarily the business unit I think the problem is well more well understood because more people have a concept that there’s more to there’s more to engineering and putting together data sets than for their reports than they ever have

45:49 for their reports than they ever have been before but I I wouldn’t put them in the bucket of oh they understand like data engineering no and and how no I agree with that I I’ve been putting more of the permissions in these roles trying to bucket it where what are you allowed to do and what are you expected to do rather than a title from a per from when we’re thinking about permissions and these roles where I’ve been breaking into those two areas and that dictates what that person is yeah I like the I like the expectation side of thing I think

46:19 the expectation side of thing I think that’s a good way of using that to be able to define the expectations are this this and this and I think even even in these different categories that broken out here there are probably varying degrees of sophistication in each of those Levens like are you a data engineer level 100 are you a level 300 are you level 500 like what data engineer are you and I think that also informs where you’re at so go ahead I was gon to say I was G I was going to wrap it up here for final thoughts so any final thoughts we probably want

46:49 any final thoughts we probably want to get closer here on on finishing on time here so we’re on time on this one any final thoughts as we go through this list of guess roles around data engineering yeah I guess I guess the a final new thought would be it’s this was a much more thought-provoking article than I thought it would be as far as it relates to a lot of different roles and responsibilities so I awesome I I think it’s also a good litmus for how you would build out a team based on the responsibilities of all of

47:20 responsibilities of all of these different tasks because as a if you think about it if a business thinks these roles as all of them need to be in place and they do right we’re doing all of these things I I don’t think there’s any argument there I think where you focus where you want to focus your time and money from a business perspective would drive deeper into do you need a specific role over another one right depending on where you’re at in your data Journey you may have a team or a

47:51 data Journey you may have a team or a couple people that are doing a lot of these tasks but if you want to ramp up per se and like hey you need to connect to a bunch of systems and you need to do a bunch of different things okay well now you’re looking at specific data Engineers okay we already have our data we like we’ve got the the team that that’s well done we don’t or we don’t have a ton of different independent systems we want to do start doing predictive analysis Etc like well you need a specific data scientist right that you can leverage as an analyst so I think it’s also

48:21 an analyst so I think it’s also interesting for me that you could also look at this as a Liss of like hey across this Blended roles and responsibilities that I have within my my ecosystem if I’m going to stress a particular area that means I’m going to go look for this individual yes yeah my my final thought I think this is making confirming something that I’ve been seeing and then this conversation just a stamp on it where if you’re seeing stop gaps barriers in your analytics at your organization I think

48:51 analytics at your organization I think it comes to it really is coming to strategy what what are you trying to achieve as an organization what are you trying to measure how do you measure it do you understand a clear path for each of those questions on terms of what do you measuring and how are you measuring it everything from here then flows Downstream to what data are we collecting to your point Mike to how are we extracting it then how are we GNA highlight it yes so what I will say is I think there’s a unique opportunity for a lot of people who understand if you understand app development and you

49:22 understand app development and you understand data development you’re you’re a very valuable person because a lot of people build applications and things without technically minding how should the data be stored and collected and reported on later on so I think that so that one is a that’s a valuable type person as we go back here and look at the the roles here I think I would say don’t be scared if you don’t have all of these roles in your organization right now I think this is a good place to look forward to and I would I probably

49:52 forward to and I would I probably echo or or think through more of I think many companies tried to measure too much and if they simplified what they measured if they simplified their definition of success what does that look like to a consumer a customer whatever that is if you if you think about focus on the customer what do they need how can I better serve them and if you focus on that side of things you now have the the right kpis to bring forward to then generate or build the

50:22 forward to then generate or build the right tool program app whatever that thing is so I think it’s I’m very hyperfocused on listening to what the customer needs and trying to anticipate how can I make their life easier with our data and or our strategy anyways that’s my thoughts with this thank you very much for listening to our episode we appreciate your time I hope you found this one valuable there were a couple nuggets here you pulled out of this one our only ask if if you like this episode or if you liked what you were hearing here please share it with somebody else either on social media or in your company if you’re having conversations with people we’d love for

50:53 conversations with people we’d love for you to just let people know that you’re finding value in the podcast with that where else can you find the podcast Tommy you can find us on Apple Spotify or wherever podcast wherever you get your podcast make sure to subscribe leave a rating it helps us out a ton if you have a question an idea or a topic that you want us to talk about on a future episode head over to the powerbi tips. or powerbi. com

51:26 we should probably record that so that you don’t have to like do it every single time we literally make you write it every single time it’s way more entertaining when it’s different it’s always different every time anyways thank you all very much we appreciate your time have a great [Music] [Music] week

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.