Discover the Insights Supply Chain Framework

Business Intelligence (BI) has been the standard for decades, but it's time for a change. In this insightful podcast episode, Dr. Duru Ahanotu introduces the Insights Supply Chain Framework, a comprehensive new model for modern data strategy.

Learn the origin story of this groundbreaking framework and how it provides a structured, non-capricious way for executives to make critical decisions about their data organization. Discover the key roles, how to structure your team (centralized vs. decentralized), and how to create clear, defined career paths for your data professionals. 

Dr. Ahanotu also tackles the hot topic of how this framework accommodates rapid change and the rise of AI—and why "magic solutions" like conversational analytics won't replace the need for skilled data experts.

Tune in now to understand how to move beyond BI and build a data ecosystem that drives real business value and action.

[00:08:38] And then I grew the concepts from there.

[00:08:47] Speaker 2: Yes. So that's kind of my bias. I was, uh, uh, manufacturing, um, sorry, not a me mechanical engineer as an undergrad, and I'd always wanted to, uh, be part of the revitalization of American [00:09:00] manufacturing. So the insight, the, the supply chain metaphor just came naturally to me because of that.

[00:09:26] So in the inside supply chain, you start with the raw data. Just like you know, in a manufacturing system, there's the raw data. Data engineers doing data mining, you know, it just seemed to fit that whole metaphor. And then they are the ones responsible for supplying that data over to the next step, which is the, uh, data analyst who organizes, who looks for structure, who does queries on the data.

[00:10:18] Um, so that's the sequential. You know, metaphor that I developed. And then of course from there, there are all sorts of flavors in terms of how you mix and match or organize those different modes.

[00:10:40] Yeah. Um, why not just like, work within the existing framework, you know, why, why create a new framework? Why did you want to distance yourself from the concepts and, and frameworks behind BI.

[00:11:05] Um, and to me in particular, the term Business Intelligence does not offer a clear career path for data professionals. Not only does it, and it also doesn't describe a whole ecosystem, right? So I wanted to come up with a framework that speaks to both an ecosystem where data is moving from one form into insights that drives business value and decisions and action, but also a place where people can see themselves moving as well in this system and progressing in their careers.

[00:11:53] almost is not always even in charge of the data ecosystem. It's, it could be a whole data [00:12:00] organization in of itself.

[00:12:33] Michael Meyers: It seems like a very holistic picture, how information, you know, yeah. Uh, goes across the organization, how you capitalize on that information and turn it into something you can execute on and results career paths of individuals. You know how you should and shouldn't, you know, decentralize, which is something that's frustrated me on the engineering side.

[00:13:00] Dr. Duru Ahanotu: Sorry, and I can't speak to that one. Sorry.

[00:13:09] Um mm-hmm. Can you, you know, let's dig in more to the, the framework itself, right? Like, you know, gimme a better sense of, you know, what it is and how it works and all these different components.

[00:13:32] Particularly, you know, management consultants always have a two by two matrix. Um, and so if I were to show that matrix, and, sorry, I'm gonna be waving my hands a little bit on the axes of my two by two matrixes. On one side there's the specialization versus generalization. Um. Spectrum. And then the others other axis, let's just call it the X axis, is the centralized versus decentralized question.

[00:14:18] Data team, but I had very specialized data professionals. So in that quadrant, what the strategy of the company is, is that, uh, people are gonna grow their careers as specialists. So I'm gonna have those specific roles that I talked about earlier, and we are gonna be a service shop across business units.

[00:14:59] That's a [00:15:00] very different orientation. It's a greater challenge to offer the data professionals, uh, career paths because they're going into organizations that. They may be the only data professional, maybe they're one of several. I just happened to be fortunate that when I went to marketing, there were lots of, uh, data professionals.

[00:15:39] Because for the first time, you know, I was managing engineers at that time. You know, data engineering wasn't really a thing. There were people who were doing data engineering, were. We're typically computer, uh, sorry, software developers who just had a shine for data and decided to go down that path. And so I was actually managing engineers and doing the whole, um, you know, the whole thing, [00:16:00] uh, learning how to, how to do that management side.

[00:16:27] Maybe you just need to, you need a generalized set of skills. You can decentralize folks and so on. So that's, uh, that's the beginning of, of the framework, and it allows a company to decide right up front, make a very conscious decision. Okay, what is. My orientation, my company's orientation towards data.

[00:16:53] Michael Meyers: And do organizations move across the spectrum with some fluidity, like, you know, is, you know, [00:17:00] um, is there a reason why an organization would switch from decentralized to centralized? Is it really up to the executive? Like, how, how should that work?

[00:17:32] Structure allows you to make very rational moves. It allows you to explain to your people why it makes sense to go from this state to another state. So for instance, let's just take a company's lifecycle or its trajectory. Where data is just a byproduct of what they're doing, um, at some point they may realize, oh my goodness, actually this data stuff is actually a strategic advantage for us in a competitive advantage for us in the [00:18:00] marketplace.

[00:18:22] And generalists might be, Hey, my finance people are doing data stuff. My marketing people are doing data stuff. My salespeople are doing data stuff. And instead I realize, you know what? I want to hire a data engineer, a data analyst, a data scientist, and I want this team of specialists. To create my data products that I go to market or I accompany my services or products, uh, with this, uh, specialized data, um, application.

[00:19:03] Michael Meyers: Is it how organizations are like predominantly organized. Um, and, and what I mean is. We sort of live in a world where everybody is, or should be a data analyst to some degree, right?

[00:19:37] Dr. Duru Ahanotu: So there's a way you can do a hybrid model, which is, um, let's again just go back to the centralized model where you have a team of data experts and they are servicing different bus business units. Well, within those different business units, there may be data people, right? Um, and those data people don't have the full set of expertise in a particular area.

[00:20:55] Um, but I will always contend, uh, that the, the [00:21:00] specialist, the specialized knowledge of the data expert or the data professional is not gonna go away, uh, anytime soon.

[00:21:17] Uh, they, they, they might need them, but they don't have them yet. You know, if, if you're an organization that you know, is, is looking at your framework and model, you know, what, what's the best place to get started, right? Like, I, you know, these are great ideas. This framework makes a lot of sense. Um, how do I start to put it into practice?

[00:22:04] Right. So that's the discovery phase, like what is your current state, and then let's compare that current state to where you think your business should be. You know, I mentioned again, is data going to be a byproduct of what you do as a business or is it your competitive advantage? You have to make that call, um, uh, where you are.

[00:22:41] If there's a mismatch, then yes, a reorientation needs to be, uh, done. Um, again, according to the. The principles of the inside supply chain. So it's just like almost every other project, right? It's the gap analysis, um, framework that you would go, [00:23:00] uh, that you would use to figure out how to proceed.

[00:23:30] Where is this going? And, and can this avoid the fate of, you know, BI.

[00:23:52] The key components. Now whether, you know, I talked a bit again about data engineering, data analytics, and insights, analytics as [00:24:00] key roles, but there is nothing so strict in this framework that says each of those roles has to be a specific person or a specific set of people. You can have one person that covers all of those roles.

[00:24:39] You would still see that artisan. figuring out how to get the data, figuring out whether they need to craft it, their own specialized pipeline or a standardized pipeline. And then they will do their data analysis. They'll do sql. They might decide there's a particular model or machine learning has to be applied, and then they're gonna do the insights, uh, [00:25:00] um, analysis, which is then to convince the business to act on the role on the results that they've come up with.

[00:25:24] But that basic fundamental flow, I think is going to be pretty sticky, uh, for some time. And that's gonna be my claim. For this podcast, we, we can check this, check in a few years and see how that's going. So you mentioned AI, so let me just specifically address this and I have a very, I, I'm developing more and more pet peeves.

[00:26:13] Organization where there are all these fraught data professionals and they don't have time to do anything. It's like, Hey, we've got a solution. Your data professionals, and I think they call them data analysts, they never have time to get done everything that you executives want to do, well, we have a magic solution, which is conversational analytics.

[00:26:50] You just type out your request and boom, there it is. So major pet peeve for me because one, I have been through this, um, many [00:27:00] times in past data teams, um, you know, leading data teams, which is this tension. There's always gonna be this tension between the, um, executive or somebody who needs some data answer right now versus the data team's interest, which is in creating standardized products like dashboards, like pipelines that can answer these questions.

[00:27:39] So when I see that meme that shows, oh, the data professionals, they're so, you know, they have no time. One, I say, well, why not think about reorganizing the way you work? Maybe it's because you are bombarding them with all these nitpicky questions and not giving them the time to build out the standardized tools that would answer your [00:28:00] questions.

[00:28:06] But, um, here's the, here's the problem. I look at conversational analytics as an attempt to just try to shortcut the entire inside supply chain, right? That somehow the, the, the data will just magically be there available in a condition that will answer every question that you have. Magically and with ease.

[00:29:16] That whose answer will lead you astray in ways that you won't even appreciate until somewhere down the line. Some businesses decision goes awry. When all it would've taken is if you just sat down with your data professional, talked to them about what you're trying to do. You know, you could have, you could have shortcut a major error.

[00:30:04] Sorry. Um. Sorry for you. Sorry. For the folks who are still trying to, um, market, um, things like conversational analytics as a way to replace the need, um, for data professionals.

sits, um, upstream but not all the way upstream to data engineering. Um, and so the way I see the roles is I am always oriented insights first. So. I think there's been a greater awareness, and I see this with the various YouTubers and whatnot, that a data analyst can't just be all about crunching numbers.

Yeah.

[00:35:11] That was great. Uh, we're gonna do a series of follow up, uh, episodes on the insight supply chain. We wanna talk about tactical, we'll talk about strategic.

[00:36:42] Michael Meyers: Awesome. Thank you so much. Uh so fun as always. Uh, thank you to all our listeners. Uh, you guys can send us, uh, feedback at info@tag1.com. Uh, please check out our past episodes at tag1.com/podcasts and, and remember subscribe so you, don't miss any future conversations.

[00:37:13]