4#13 - Juha Korpela - Data Consulting and the Role of Data Modeling (Eng)

«You bring in the knowledge of what works in real life and what doesn’t. That is actually what you are being paid for.»

With a year behind him as a solo entrepreneur in his own company, Datakor Consulting, Juha Korpela takes us on a journey through fact-finding-missions at what he calls "the middle layer" of organizations — the strategic area between high-level business strategy and tactical project execution. It is here, he believes, that data consultants can create the most significant and lasting value.

We discuss the pitfalls of standardized frameworks and "blueprint" approaches offered by many consulting firms, and why tailored solutions based on a deep understanding of organizational culture always yield better results. Juha shares his methods for knowledge transfer that ensure organizations can continue succeeding with their data work long after the consultant has left the project.

Here are Winfried´s key takeaways:

Skills

• The key skill as a data consultant, no matter if on a strategic or solution, project level is to understand «what the customer really needs.»
• The key skills are:
   
  • Listening. Active listening is the key to understanding.
  • Create mental models: when talking to stakeholder you need to be able to put the information you capture together in a mental model.
  • Understanding.
  • Tech comes after.
• Working with data modeling is about listening to stories about how business work. Understanding business processes are key.
• Understanding stories about business and what is relevant for data modeling is a skill that everyone can profit from, but that is seldom taught.
• Data Modeling is a fact-finding-mission.
• It is about understanding what the organization does, how it does things, and where this could be improved.

Impact

• A data consultants impact is dependent on the organization, the structure, and the level of maturity.
• If there is a CDO or CIO to connect to it can be a good way to create results and visibility.
• Also as a data consultant it is important find a place in the organization where you have shared views and understanding.
• If you begin bottom-up you need to be ready to sell this upwards in the organization.

Limits

• Consultants can help with the initial projects to get you started.
• Consultants can help figuring out processes and operating model and design what is needed.
• Organizations need to create long-term ownership in house.
• Running and maintaining needs to fit with the organizations culture, its s structure, needs, maturity, etc.
• Models, blueprints, frameworks that you get from the outside can get you started, but do not work in the long run.

Patterns

• Data Consultants can see certain patterns emerging across an industry.
• That knowledge on patterns, lessons learnt, experiences is valuable to apply.
• That knowledge you bring in is what defines your value, more than specific skills.
• It is easy for people in organizations to get stuck. Consultants can help as a fresh wind.

Knowledge transfer

• As a consultant you bring in new knowledge, and you need to account for that organizations want to transfer that knowledge to internals.
• Find ways to create custom training packages to facilitate knowledge sharing.
• You aim for the organization to succeed with their work, also after the consultants are gone.

Consultant aaS

• Do we move from being consultants to becoming a service offering?
• Service models can crate a distance between consultants and clients.
• You need to have a clear understanding the impact of models that include ownership and responsibility transfer as eg. Outsourcing operational tasks.

Transcript

Intro

Speaker 1 0:00

Dette er Metadema en holistisk syn på datamanasjon i Nordisk. Velkommen, jeg heter Winfried og takk for at du har lyttet til meg i dette episodeet av Metadema. Vår visjon er å promisse datamanasning som en profesjon i Nordisk Vise de kompetensene vi har, og det er grunnen til at jeg inviterer nordiske eksperter i dat og informasjonsmanagning til å snakke. Velkommen til Metadema, og vi er tilbake i Finland. Dette er fantastisk. Jeg har vært i Metadema for what four years now, and we've been around the Nordics, but Finland has been one of those countries where, at least the first seasons, there was just, i think, one or two episodes from Finland. Now this season is packed with a lot of Finnish professionals, and today I'm talking to Juha Kortopela. Jeg håper jeg har nevnt ditt navn rett.

Speaker 2 1:09

Det er rett, takk.

Speaker 1 1:12

Juha arbeider som data-konsultant. Vi ville snakke om data-konsultantjobb. Hvordan gjør du konsultans i data? Hvordan får du den rette gode effekten på organisasjoner? Hvordan sprøver du din kom data? how do you get the right good impact on organizations? how do you spread your competency in a way that organizations profit from it but at the same time, you can also profit yourself as a data consultant, by recognizing patterns across industries, by seeing how data evolves throughout. I think this is really interesting because, as a data consultant, you are in a quite unique position. So let's talk more about that. But before we move into the main topic, juha, it's all yours. Please introduce yourself.

Speaker 2 1:58

Great to be here. Thanks for the invite and indeed great to have some more Finnish voices in this podcast as well. Like Joe Rice, our common friend said some time ago, he said that Finland has this great concentration of data people but then no one really knows them. So it's interesting that we have lots of folks working on data here and not that much visibility I'd say, say, as sort of data-focused nation of people. But great to be here.

Juha Korpela - Data Modeling and Music

Speaker 2 2:30

And yeah, indeed, i'm Juha Korpela. Well pronounced clearly for English speakers I usually need to explain that it's a kind of Y instead of a J in the beginning, so it's Juha instead of Juha, which is occasionally what I get. But yeah, i've been working as a data consultant now, actually in previous parts of my career as well, but currently as a solo entrepreneur in my own one-man company called DataCore Consulting for almost exactly a year. So we're on the last day of January as we are recording this, and I started my first gig last year. So we're on the last day of January as we are recording this and I started my first gig last year. I think it was 25th of January, so that's pretty exactly a year Before that I've been doing all sorts of data things for about 15 years I suppose it is now.

Speaker 2 3:19

I worked at this start-up called Eli Technologies before I joined or formed my own littleup, som heter Eli Technologies, før jeg formet min egen lille konsultasjonskampanje. Eli Technologies gjør en datamodellingskampanje og jeg var ansvarlig for produktutvikling der, for jeg har vært gjennomgående i datamodellingskampanjen i ulike posisjoner I store manufaktører, i ansvar med dataplatformer og datagoverning og gjør konsultasjon i en konsultasjonskompani og på den publiske sektoren og alle slags ting. Men datamodeling har vært noe som jeg har kjøpt med meg fra posisjon til posisjon som en hvordan skal jeg si det metode som jeg alltid bruker, men det kanskje kjører på min overordnende arbeidshistorie. Jeg tror vi kan gå inn i detaljer om hva jeg har gjort i forhold til konsultasjon. Utenfor det er jeg i stedet en hobbymusiker, så jeg spiller bas. Jeg har et par baser der nede.

Speaker 2 4:16

Jeg har gjort jazz, funk og soul, det slags ting. Jeg har også spilt trombot litt I dag litt mindre, for play the trumpet actually a little bit, but nowadays a little less because it takes a lot of practice to get that right and I have no more time to practice trumpet playing that much anymore. Also, i'm very, very fond of cocktails, so cocktail mixology is one of my hobbies as well, a skill that I actually learned from a data modeler called Alec Sharpe from Canada, who's kind of a mentor of mine in this part at least. Yeah, that's it. I'm living here in Helsinki, finland, in data, as I said, for about 15 years, and looking forward to more with this actually quite nice life that this single kind of single person company has provided.

Speaker 1 5:00

Fantastic. I want to dig a bit deeper into that fantastic life that single company har tatt. Men før det, alex Sharp og mixologi. Jeg tror det er en interessant en. Jeg er virkelig inn i koktajler og mixologi også. Jeg gjorde også noen live bartending på en online-konferanse under covid-19, hvilket var gøy, så det er en hobby vi har i kommunen, og jeg tror at musikk er også noe som jeg har utforsket i podcastene. Mange datamenn har musikk som en hobby. De spiller instrumenter. Jeg tror det er mange gitarer som er styrte i datamennet. Jeg tror det er fantastisk. Jeg tror det er noe der som balanserer ditt datavereld.

Speaker 2 5:48

Jeg synes absolutt at det balanserer. Det er et kreativt hobby, det er pure kreativitet. Jeg vet ikke hvilken del av hjernen er ansvarlig for hvilken aktivitet, men jeg tror at ulike deler må være Det, å si. Det er det noe parts must be, although, that being said, there's something logical in music in terms of music theory and how it actually is sort of mathematics in the end, which sort of tickles the same part of the brain as, let's say, data modeling does. So in that sense, i think it's quite natural combo.

Speaker 1 6:18

I think I like to use music as an example to explain how structure and synetization can help to be creative, because it gives you guardrails, Like writing notes. You have to have a common structure to write notes so they are readable for others, But that doesn't limit your creativity. I think that music is an interesting one there.

Speaker 2 6:41

Yeah, and I must say I don't want to distract us too much, but I must say it is also very interesting in terms of templates, how music and data modeling especially are related, because you have these templates in jazz, for example, that you have your one, two, five loops at the end of the phrase or something, and that is just a template that you can utilize and you can identify it in different sorts of songs and you can also have og det er bare et template som du kan utvikle og identifisere i ulike slags musikk Og du kan også ha template i din mentale bibliotek som du kan identifisere og utvikle i ulike slags projekter når du arbeider med data. Jeg synes det er ganske interessant.

Speaker 1 7:16

La oss gå litt dypere inn i det i en sekund, men før det tror jeg det ville være interessant å utvikle hvor din interesse for data kommer fra. Hvor startet?

Speaker 2 7:26

det. Ja, det er en interessant historie. Min dekri er faktisk fra økonomisk, så jeg har en økonomisk masterdekri. Jeg har aldri gjort noe som gjelder økonomisk. Da jeg begynte i min første jobb var det Statistikk Finland, det offisielle Statistikkbureauet i Finland. Jeg It was Statistics Finland, the official statistics bureau of Finland, and I was applying into this job.

Speaker 2 7:46

That had something to do with business registers. So I thought, okay, you know, there's numbers involved. I've been working with numbers during my thesis and so on. There's businesses involved. Sounds like economics. And I went there. I got the job.

Strategic vs. operational Data Consulting

Speaker 2 8:02

It was some sort of permanent position, sorry, no, it wasn't a temporary position for some sort of project, whatever planning thing. It was junior designer of something or other. I got the job and pretty much the first day my boss asked me do you happen to know what a data warehouse is? And I said I have never heard of it. And he said, well, you better start googling, because dette er et datavarehousingprojekt. Så jeg endte opp i et datavarehousingprojekt av uansett, uten å vite noe om det, og da hadde jeg veldig snabbt lært noen basisker. Og siden da har jeg vært driftene mer og mer inn i dette dataet faktisk drifte det ganske snabbt og kom til datakonsultansiet et a few years later, and so on and so forth.

Speaker 2 8:42

It's really, really interesting, i think precisely for the same reasons that we kind of were discussing relating to music. There's logic to it, there's patterns to it. It tickles the right parts of my brain to work with these data topics, and one thing that we probably will cover today in a little more detail also is the kind of two-layered nature of it. So there is this technical work of dealing with code, dealing with tables, dealing with data sets, which is, in a way, its own thing, interesting in its own way. But there is also this more strategic level of it, which has to do with how organizations work, how people work, how they cooperate, how the kind of operating model is built around it.

Speaker 2 9:24

Organisasjoner arbeider, hvordan folk arbeider, hvordan de samarbeider, hvordan det operasjonsmodellet er bygget rundt. Det, som er, i en slags business-strategisk måte, interessant. I en annen måte, men det er fortsatt ekstremt interessant i data, hva kommer først For meg? jeg tror at jeg opererer på det mer strategiske nivået. Jeg har vært i pipeline-situasjoner, som alle i noen år siden og så videre. Jeg har, i det hele tatt, driftet bort fra det. Jeg kan ikke huske den siste gang jeg har styrt kode inn i produksjonen. Det har vært noen år.

Speaker 2 10:06

Dette var partisk random. Remember the last time I pushed actual code into production. It's been a few years. This was partly kind of just random drift, but partly also a conscious decision that I find that the actually well important work is being done everywhere, but the biggest impact in the way that organizations utilize data and how they benefit from data is actually happening in the decisions made on that strategic level And whatever tools and technologies utfører data og hvordan de kan få tilbake data er faktisk gjennomgjort i beslutningene som er gjort på det strategiske nivået Og hvilke kjøp og teknologier og hvilke kod vi skriver. Det er impaktfullt på taktisk nivå, men hvis de taktiske nivås-sakene ikke kan bli oppført konsistensivt, så er det alt Ja jeg tror du er rett om denne.

Speaker 1 10:41

Men dette er en interessant en fordi jeg føler at særlig strategi er noe som er ganske integrert Og det er selvfølgelig mange strategisk konsulter som gjør strategisk konsulting. Men til slutt av dagen, må organisasjonen ha sin strategi Og ofte, hvis du ser på hvordan strategiske konsulter fungerer, er det mer av en. If you look at how strategic consultants work, it's more of a certifying an existing strategy or providing elements to a strategy that an organization builds itself And then you bring in one of the big names to put their stamp on it and say this is approved by whoever certified the strategy. So it's easier for you to talk about your strategy internally in your organization, but also externally. But if you work as a consultant and you want to make some strategic change and impact, how do you do that in that field?

Speaker 2 11:37

Very good, Actually, let me take a step back, perhaps, and discuss what do we mean by strategic in this context? Because the way I see it takk for at du så på me between my history hobby and the work I do is that the word strategic and the word operational, the word tactical, mean different things. So if you read, let's say, history of the second world war or something on strategic level, you make a decision that, okay, we're going to be invading Poland. Sorry for any Polish listeners, but that's just an example that popped into my mind. On the operational level you figure out how that is going to be supplied, where's the logistics, what sorts of gear you need, what sorts of replacements you're going to be needing. So operational level in that sort of thinking is kind of the world of logistics, world of sea, world of sea trade lanes and something like that. And how do you make the strategic decision basically turn into actual action? And then that action happens on the tactical level. Now in the business world, for some reason, we have switched the operational and tactical level. So if you read business strategy, yes, you have the strategy at the highest level, but then you talk about tactical decisions and then the action happens on the operational level. I hate this because it's very confusing and I would like to make so many kind of use examples and metaphors and anecdotes from history in this data work. but I can't talk about operational level because everyone thinks that it's the actual activities. But now I'm getting to my point about what is strategic and what is not.

Speaker 2 13:33

I think that the impact that an external consultant can make is not necessarily on the strategic level. That is the of course you have the big four consultants is working with the top brass in every organization figuring out that. but I think the actual impact you can make happens at the so-called middle level, which I would like to call operational, but it's often called tactical. And this middle level is where you figure out how that organization implements those strategic goals into kind of actions that are repeatable, consistently repeatable, actions that are being done so that the strategic objectives are actually then reached. And I think that is the level where I operate. That is the level where you can make the most impact. A single kind of development-oriented consultant can make a huge impact on the lowest kind of level of action with a single data product, single data solution. they can make a massive impact there. They very rarely can impact the ways that the organization is going to use after that consultant has left, and what I try to do is to impact the ways that organizations are also going to operate after I'm no longer there.

Speaker 2 14:54

Det er litt mer vage. Jeg kan ikke si at fordi av meg har du fått en ny produktrekommendasjonssystem for ditt webshop og dette har kastet deg til å få flere klienter og kunder og du har regnet 20 millioner dollar i år. Det er ikke den slags impakt jeg vil ha som konsultant som jobber i den middelstøyen. Men jeg kan si at fordi jeg var hjelping denne organisasjonen å implementere denne strategien i åpne oppgavepatter, oppgavepatter, operasjonsmodeller, måter å jobbe, så det hjelper dem å faktisk fortsette å succesere i de fremtidige projektene også. Og det er der det er etterhvert hva du kan tilføye. veldig bra.

Speaker 1 15:39

Jeg liker det, jeg har vært forhånden for og det er en artikkel der ute av Martin om at strategi ikke er en plan. Vi har snakket om det for mye, men jeg tror at strategi er mye om å navigere uansettelse. Jeg tror at det ofte er forblitt med langtidig planløsning, is very much about navigating uncertainty, and I think it's often confused with long-term planning, which is not the same. It's just entirely different. It's a different methodology, it's a different technique, it's a different way of working And I'm looking at it as kind of in gaming.

Speaker 1 16:10

You have a fog of war right, there are undiscovered areas on your map that you don't know what's hitting behind them. But even though you don't know what's hitting behind them, but even though you don't know what's in there, you kind of know where your goal is and you can make decisions to reach that goal and tactically moving forward to lift that fog of war. And yes, there could be surprises there, obviously, but then you have to handle those surprises when they occur. And I think that's a kind of a different understanding of strategy than Men. Da må man handle disse overfladene når de skjer Og jeg tror at det er en annen forståelse av strategi enn det jeg er vant til Og når du ser på det du har jobbet på som datagoverning, datamodeling, det er veldig mye metodologi til å løfte denne forholdsvann, til å få mer klarhet på landskapet.

Speaker 2 16:59

Ja, absolutt, jeg er med to lift that fog of war, to get more clarity on your landscape.

Speaker 2 17:03

Yeah, absolutely agree.

The Art of Listening and Creating Mental Models

Speaker 2 17:05

It is very, very interesting to work on data governance and data modeling and kind of well, so we call it information architecture, perhaps on the kind of higher level, in the way that you kind of lift that fog of war on an organizational level, instead of just working on individual solutions. I mean, obviously I do a lot of modeling and work with individual projects to help them figure out their requirements and so on and so forth obviously as well. But there again, i think that sort of middle layer is where data modeling and data governance actually derive or create most value. And in terms of data modeling and data governance actually derive or create most value, and in terms of data modeling especially, i think that the understanding of what sorts of data assets the organization has, which data solutions are utilizing, which data assets do you have assets that you are not using or you should not be using because of some policies or whatever governance comes into play there. These are actually very, very interesting questions in terms of how that organization then again, going forward, will execute those individual projects.

Speaker 1 18:09

The tricky thing here is that both data modeling, data governance are very much dependent on the organization and understanding the business. I think that's the whole point of doing. Data modeling is to understand your business in a better way. And as a consultant, how do you do that? What kind of skills do you need to bring to the table to be able to understand a business that you are not really part of? You come in and have to navigate through that. What do you need to bring to the table?

Speaker 2 18:41

That's an excellent question. I think these actually are the source of skills that I think every consultant should have to a degree, regardless of whether or not you work on individual projects deriving solutions from your requirements that you're getting from the business, or something. You still have to understand what they actually really need. It's not just that I will need to create these three tables with these names. It is actually more about how does this create value? What do they actually need? And the skills that you need on different levels of abstraction, as it were, are pretty much the same. You have to know how to listen, first of all, listening kind of actively what is going on.

Speaker 2 19:19

Leveler av abstraksjon, som det var, er ganske samme. Du må vite hvordan du lurer, først og fremst, lurer aktivt på hva som skjer. Du må være kapabel til å skrive mentale modeller for deg selv om hvordan ting fungerer, hva er kursene og hva er eventene som skjer. Det kan være veldig, veldig enkelt. For eksempel, når du gjør datamodeling, er dette litteralt det eneste du bør gjøre. Du går noe og spør hvordan fungerer ditt business Og som et eksempel, mange år siden jobbet jeg med en bank og vi hadde å finne ut noe. jeg kan ikke engang minne hva. det var something relating to how credit cards are being given to customers who request a credit card account. We just had a person there who was some sort of operating manager of that process and they explained that, okay, well, a person walks into a bank or goes onto this website and they fill this form and then that form is being sent to somewhere and a decision is made, and the decision is registered here and what not.

Speaker 2 20:19

And this is now not really kind of a data thing that you are hearing. It has nothing to do with databases, nothing to do with being able to code or data model actually even, or anything. It is just about listening to that story of how the business works, creating kind of a mental model for yourself so that you understand, alright, you know customer, alright, okay, we have to understand how and when and if they are actually given that credit card, and from that then you derive your actual work. So, in terms of data modeling, it is literally picking the object from the story You have a customer, you have a credit card, you have a credit card application and figuring out how they relate. But if you were working on a more tactical level, as it were, building some sort of analytic solution for that, for example, then you would need to understand that at this part of the process they need information about these things and I can get those things from the source, xyz, where the credit card applications are stored and so on and so forth.

Speaker 2 21:21

And understanding that, listening, that business story, understanding where you need to do something, where the data is, how it relates to the processes and the people, that I think is somehow a universal skill that we all need in this business, but something that no one is being taught in any university. I don't know if there are any very good like online courses or anything about that. I heard that Joe's new data engineering courses actually start with kind of this part of listening and understanding, which I think is excellent. It's I don't know. I also find it extremely interesting personally because, having worked with different sorts of enterprises in manufacturing manufacturing company, i learned how pulp is being made from wood and how does that factory work and where the pulp goes, and so on and so forth, and I find this person very, very interesting.

Speaker 2 22:34

The thing is that if you are orienting yourself into this work, so that, okay, i'm going to do technical things at, hvis du orienterer deg selv inn i dette arbeidet sånn at jeg skal gjøre tekniske ting. jeg liker å spille med kode og tabler og prosesser, så kan du ikke virkelig forstå at, spesielt hvor langt du går, den største delen av det arbeidet har å gjøre med å høre og forstå Og det har ingenting å gjøre med det aktuelle tekniske stoffet. Det kommer understanding and it is nothing you know. it has nothing to do with the actual technical stuff. That only comes after.

Speaker 1 23:01

I like that and that creating those mental models after you listened, that's a certain. You bring in a different level of abstraction of what you heard. It feels like we're coming back to what we talked about at the beginning. It's like writing notes by listening to a song. Right, you listen to a song and you can write the notes accordingly.

Speaker 1 23:22

At det begynner er det som å lese noter. Ved å lese en sang, du lurer på en sang og du kan lese noter på en gang, og det er en unik kunnskap. Jeg tror at da jeg gikk til skolen i Tyskland, vi måtte lese klassisk musikk og da måtte vi identifisere de ulike instrumentene som var spilt innen de ulike songene. Og jeg tror det er litt samme, a different instrument that we're playing within the different songs, and I think it's kind of similar, it's kind of going in the same direction. Now, what I think is interesting there is that it's kind of, i think, in our prep call, you call it effect finding mission. You create a mental model and then you have to communicate that model, talk to people about it, get their input to see does it actually apply, does it make sense and can we find, kommunikere denne modellen, snakke med folk om den, få derim til å se om den faktisk anbefaler, om det gjør noe, og kan vi finne sammenheng tilgjengelig til det? Er det det du mener med faktikker misjon?

Speaker 2 24:09

Ja, jeg tror det mener to ting til meg i en måte. En ting er at jeg må finne ut faktene, altså finne ut mentalen eller om jeg gjør datamodell i den aktuelle datamodellen, som er bare en mentalmodell på en pisse papir eller noe annet Figurere ut hva som skjer i en viss område, i en viss prosjekt, i en viss business domain eller noe annet. Det er en slags faktfinding-missjon for meg selv, finding mission for myself. I need to understand how that thing works so that I can give recommendations or create data models or even create code if need be or something like that. But the other thing is that we kind of again get down or get up, basically from this kind of tactical level, where I'm interested in the single project or the single business area, to also then be able to lift yourself one step up to that level of operating models and ways of working and then seeing how does this organization actually do things? and is there something that I can derive from my previous experiences that, for example, if I am working with a single project, do I see in that project some patterns, some anti-patterns maybe, or something that might indicate, for eksempel, om jeg arbeider med et enkelt prosjekt? ser jeg i det prosjektet, noen patterner, noen antipatterner eller noe som kan indikere hva rekommendasjonene er, som jeg bør gi på det litt høyere nivået av abstraktsjon.

Speaker 2 25:32

Og ofte skjer det at jeg blir kalt inn som konsultant til å arbeide på et enkelt prosjekt, men det blir en konsultasjon in as a consultant to work on a given project, a single thing, but that turns into a consulting assignment that actually is about figuring out the ways of working. It is a fact finding mission, in a way then of understanding what the organization does, how it does those things and where this could be and should be improved. What are the actual goals of the people involved? Occasionally, of course, this is precisely the task that I'm given that they say, okay, we're now doing, you know, renewing our ways of working or figuring out new roles or something. Come and you know, help us and let's figure it out together, and that's perfectly fine. But it also kind of emerges occasionally from the work that is being done on that kind of lower tactical level, fra arbeidet som gjøres på denne slags flere taktiske nivå.

Speaker 1 26:18

Det er veldig interessant. Hvor tror du at du kan ha mest influens? på hvilken nivå har din arbeid?

Speaker 2 26:29

den beste impakten. Det depener på organisasjonen og strukturen og nivået av maturitet. Jeg vil også si Jeg arbeider personlig veldig ofte direkte med toppmanagment, som I personally very rarely work directly with the top management like the C-level, it is simply not something that comes up in my projects usually. What I do think that works very well is if there is a head of data or a CDO or some sort of even chief architect of data or whatever who kind of has this whole middle layer on their responsibility, then those people are the ones that I seem to get the best results with. So they already have that need to optimize and to ensure that not just individual projects but also the future projects will go well, and thus we kind of have the same, we're talking the same language in a way, and we have the same.

Organizational Challenges and Frameworks

Speaker 2 27:20

Ikke bare individuelle prosjekter men også fremtidige prosjekter vil gå bra, og så har vi samme språk og vi har samme syn på ikke bare prosjektet på hånd men også hva som kommer etter. Okasjonalt. Er det også mulig at dette fungerer fra bottom-up, så at det er en individuell prosjekt og ingeniører og arkitekter arbeider på individuelle prosjek bottom up, so that there is an individual project, engineers and architects working on individual projects and together with them then we create something that might become a repeatable process or kind of way of working based on just that single project or a couple of projects that we do together on the bottom level. But that then means that you kind of have to sell it upwards in that organization to actually make that change happen. I think that the most impact that I can see myself doing in an organization happens when I interact with that kind of middle management layer, as it were, or if a CDO or something like that exists with them.

Speaker 1 28:16

You talked about connecting your work with ongoing work in the organization or embedding it, and I think, especially when you til dem hvordan vi går frem, det er ikke noe som du gjør som et prosjekt eller noe som du kan gjøre som en basert på konsultanter, og jeg hadde en chat i går med en partner i en firma og han snakket om at hvis en organisasjon kommer til ham og sier vi vil at du rører vårt data-governance-program, han slipper det, for dette er ingenting som du kan røre. He turns it down because this is nothing that you can run as a consultancy. That's something you have to run internally. So where's the limit? What do you say, especially for data modeling? where do you say, no, you have to own this.

Speaker 2 29:09

Ja, that's an excellent example, i think. And with data governance I fully agree With overall data modeling and information architecture, as I like to call it, including not just the actual data models but let's say taxonomies, ontologies and what not. if you go into that direction, i can help with the individual first projects. That's what I can do and that's what any consultant can do. I can help them figure out the processes and roles and practices, the operating model, as it were. I can help them design that, but I can't run it for them.

Speaker 2 29:42

And this, i think, is the kind of key pattern that I see that organizations really need to have ownership of the kind of system that is being built around data governance or information architecture or anything, or actually any kind of data work platforms or whatever. And that has to be in-house. I can't see I have never seen a successful organization running these purely with kind of consultants. I don't think that is doable. It has to do with the way of working in that particular organization. It has to be a way of working that fits this organization, the individuals in the organization, the culture, denne organisasjonen. Det må være en måte å jobbe som fitter denne organisasjonen. de individuelt i organisasjonen, kulturen, maturiteten. Det er ikke noe du kan gi fra utsiden. Du kan hjelpe dem å designe det, men du kan ikke lage det fra utsiden.

Speaker 1 30:29

Du har brukt opp taxonomier og det er nok det samme for datamål. Jeg har vært i et prosjekt i årsbær hvor en organisasjon kjøpte og du kan gjøre det, ikke sant? Du kan kjøpe reddmålte, industristandardiserte taxonomimodeller og de kjøpte en og de implementerte den som den var og den g as well. That offer kind of that blueprint consultancy. If you engage us, you get the standard model and we know exactly what we need to do. We have a structure, a blueprint, ready that we can implement as soon as we come in. Have you been confronted with that? Is that something that organizations are still looking for?

Speaker 2 31:17

Yeah, in some cases it does come up And from my personal experience, the way and the situation in which I encountered these frameworks and blueprints that have been applied from elsewhere is that I appear there when it's obvious that the old blueprint doesn't work actually. So we have to figure out something with that organization. Now, i might sound a little bit negative on this, så vi må finne ut noe med denne organisasjonen. Nå, jeg kan høre at jeg er litt negativ på dette, men jeg tenker at framganger, eksternal innovasjonsmodeller av hvordan man jobber, eller aktuelle datamodeller eller taxonomier eller noe annet som du får fra utenfor. De kan være et begynnelsepunkt, men jeg ser ikke noen av disse som get from outside. They can act as a starting point, but I don't see any of those ever actually working in real life for a longer period of time. It is good to use existing patterns In data modeling, for example. I always use patterns and reference models and whatnot. It would be insane to not use them as a starting point. But for each individual organization, there are simply too many moving parts in this. Det er ikke sant å ikke bruke dem som startpunkt, men for hver individuelle organisasjon er det bare for mange bevegelser i dette at man kan bruke noe direkte.

Speaker 2 32:26

I know that it is a sort of tempting direction for consultancies, especially larger companies, to have a sort of certified approach that you always do the same way. I do think it's a good idea if it is basically a checklist of things that you have to figure out, that, i think, does work. This is a checklist of things you know. Okay, there's no data governance in place or it's poorly set up. What do we have to take into account, what things we have to figure out with them? That sort of approach is very good, but you can't come with ready-made roles, ready-made responsibilities and policies and even tools and say, okay, let's just put all of these in place and then, ta-da, you have data governance.

Speaker 2 33:07

It is exactly the same with agile. I'm also I don't know if you are a safe guy or not the scaled agile framework. I am not very much not a safe guy, but I think there it also. It's the same thing. You have this pattern, you have this structure that should include everything and then you just apply it into the organization and I haven't seen that actually working in any agile fashion anywhere ever. But let's not go too deeply into that. I know that there are people out there who are very fond of SAFe and so on and so forth. But as an example, i think those sorts of patterns are useful starting points but not really applicable as such. I very much agree with you eksempel. Jeg tror at disse slags patterner er nødvendige startpunkter, men ikke så applikabelt som sånn.

Speaker 1 33:53

Jeg er veldig til å forstå deg. Vi hadde dette diskusjonen i DEMA også om DM-bok som en bodd av kjærlighet og hvordan man skal anbefale det. Min syn er at det er litt som når du maler en fens, du trenger no, du må bruke en slags primer eller en baskode for å sikre at det ikke kommer rundt, for å sikre at basene er lagt opp før du kan bruke din baskode. Denne baskoden kan være noe annet, det er helt opp til organisasjonen. Du må bruke den samme første lager eller barriere eller primer du har i stedet.

Speaker 1 34:26

Det er litt annet layer or barrier or primer that you have in place and that's kind of the frame. But there is a give it a positive twist here. There is something about the unique possibility you have as a consultant to look at different organizations and different industries and see certain patterns emerging, and it doesn't mean that you have to go into a framework, blueprint kind of mindset, men du kan se at noen ting fungerer og andre ting ikke og du kan bringe denne kjærligheten til ditt kjøp. Tror du at det er noe som er verdig oppfattet og det er virkelig noe som du kan?

Speaker 2 35:07

bringe med deg. Ja, absolutt, jeg bring with you. I think that's one of the key things that, especially on this slightly higher level of abstraction, in this middle layer, what is expected of you is that you bring in experiences from other organizations. You bring in the knowledge of what works in real life and what doesn't, and that is actually what you are being paid for. It's not as much, i mean. Obviously, in some cases it is because people think that you have specific skills. They want to loan or borrow your brain in a way to help them figure things out. But actually it is quite often about bringing in experience from elsewhere.

Knowledge Sharing for continuous impact

Speaker 2 35:45

Organizations are well, especially if you go into traditional organizations manufacturing or something like that. They are very, very slowly changing things outside, the kind of big tech where, you know, not every organization is Amazon or something and very few actually operate on that level. Organizations are rather stable, and that is a good thing, of course, but it means that it is rather easy for people within that organization to kind of get stuck inside the box And you need someone to come from outside and say, okay, i've seen things done this way, i've seen things done that way, i think this is what would work for you And then, starting from that, or starting from some existing framework or whatever, as a kind of template or pattern to be followed, then you go on and create something that is, in the end, in any case, unique to that organization.

Speaker 1 36:36

And when you come in with your knowledge and the patterns that you recognize, you bring in an expert knowledge to the organization And a lot of organizations have that need that. they have people inside the organization, they want to train, they want to transfer your knowledge to them. How do you work with that? How do you work with that knowledge transfer?

Speaker 2 36:58

Yeah, that's a very good point. Personally, i, for example, have these little training packages that I go through with my clients. So if needed, if we see that it's useful to, for example, train people in that organization to do data modeling which is something that we klienter Så hvis det er nødvendig, hvis vi ser at det er nødvendig, for eksempel, å trenere mennesker i denne organisasjonen til å gjøre datamodeling hvilket er noe vi ender på å gjøre i veldig mange situasjoner så kan jeg runde et par klasser om datamodeling. Jeg kan gi dem basisene på hvordan det faktisk fungerer, projects to kind of hold their hands, get them used to the data modeling part of it. Or, let's say, data product management, which has now been big with all of my clients.

Speaker 2 37:35

Actually, i have some basic kind of info packages, kind of basics of product management type of training sessions that I can do with them.

Speaker 2 37:43

So I've tried to go and build these kind of reusable sort of basic trainings, something 101, that I do with my clients. In many cases it also involves documenting these best practices and ways of working. There might be an internal kind of training platform of some sorts. We might even record some videos with the experts in that organization or just write up sort of internal wiki pages or something where you explain that, all right, in this organization, this is how we do information architecture, this is how we do data products and so on and so forth, and I think that is a key part of it because, as I said when we kind of started talking about this, i think that the impact that really lasts and that I can really make is that the organization continues to succeed with their projects, also after the consultants are gone, and if we want to aim for that, we have to ensure that the expertise that we bring in actually stays in.

Speaker 1 38:42

Exactly. And that brings me to another point that I really wanted to discuss with you, and I mean I worked as a consultant before. And I mean I worked as a consultant before and I always enjoyed being in-house, being on client side, close to the people I work with, close to the project. I never like to be sitting at the consultant's location, right, no-transcript. The consultancy has much more flexibility in how they use their consultants to cater to that need. But you create a certain distance to you, don't become a consultant anymore, you're more a supplier of a service, and I think, especially in an operating field, in the technical field, that is a model that a lot of organizations are av en service, og jeg tror spesielt i en operasjonsfjell, i en teknisk fjell. Det er et modell som mange organisasjoner er opptatt av. Hvordan ser du på det?

Speaker 2 39:58

Ja, det er interessant. Jeg jobbet som konsultant i en medium-sized konsultans et par år siden og det var også denne drivn til å bygge sånne oppgavebare servis few years ago and there was also this drive to build sort of repeatable services that would be sold as something as a service kind of approach in there, and I do think I agree with you. I do think that it creates a certain disconnect between the consultant and the organization actually doing the work. The way I do things now, because I'm just one guy in my one-man company, i basically just go there and I'm part of whatever team happens to be there and that's fine And I am very, very close to the actual client organization. in that sense I'm sort of part of it, even though my role might be more advisory often than it is kind of practical hands on.

Speaker 2 40:46

But I do see the again the kind of maybe interest in larger consultancies in building these sorts of because these are easy to sell, these are easy to organize and so on.

Speaker 2 40:56

but I do fear that it does not necessarily always lead to the best possible solution.

Speaker 2 41:01

I don't personally know where the kind of boundary should be, but I do think that this kind of idea of outsourcing parts of your data and analytics processes, be it then operating or support or whatever, all the way up to ownership of platforms even and stuff like that.

Speaker 2 41:20

Outsourcing that to large consultancies of lots of offshore people and so on and so forth is generally, i think, a bad idea, and I am now very happy to see that many organizations are kind of coming back from that, and I have actually seen even an organization called this process insourcing that they are explicitly insourcing these processes that have been outsourced before because they know they need to be more on top of their own data. they have all their AI projects and what not going on. They have to understand themselves what data exists and what they are doing with it. So I think that there's been a pushback now, or there is starting to be a pushback from that maximum outsourcing of all the services. I don't know where the line should be, but I think this is a good direction if we're getting back from whatever it is right now.

Speaker 1 42:12

I definitely agree with you. We are closing into the end of our conversation, but at the end I always have the same question Do you have any key takeaway or any call to action you want to have for people to take away from this conversation? Maybe?

Speaker 2 42:27

yeah, excellent question actually conversation Maybe. Yeah, excellent question actually. Maybe one thing that because I have been now mentioning the middle layer between strategic and the actual projects now many times, what I think many organizations should do, and actually all of us data people should do, is to kind of consciously pay attention to that middle layer a bit more. And I think the key word here is kind of repeatability. How can we make the projects that succeed, how can we turn those successes into something that is repeatable, that also the next project and the project after that and the project in five years time also succeed? What are the kind of commonalities that help us succeed? And I don't see the kind of very technology-focused type work that most of us are doing in data all consultants and internals alike. I don't see too much thinking going into that repeatability of success And I think we should pay a lot more attention to that. It's what I try to do with my consultancy engagements, but I think all of us would do well if we paid a little bit more attention to that layer.

Speaker 1 43:31

That's a fantastic call to action. Thank you so much and thank you for the conversation. Thank you.