Nested Value Streams (i.e. scaling)

Welcome to Definitely Maybe Agile, the podcast where Peter Maddison and Dave Sharrock discuss the complexities of adopting new ways of working at scale. Hello and welcome to another exciting episode of Definitely Maybe Agile with your hosts, Peter Maddison and Dave Sharrock.

How are you doing, Dave? Brilliantly well, Peter. Excellent uh introduction, as always. Thanks very much. Now, I wanted to kind of pick your brains a little bit. I'm diving straight in on this one. But last uh week or a couple of weeks back, we were talking about value streams, and I kind of dropped off that call, scratching my head and thinking, how do we take a value stream concept that if I'm a startup and I've got a single product or service that's delivering value is fantastic and easy for us to build around, and uh scale that up to the point where I'm working with you know a large organization with hundreds of people working on tens of different products and services? What are we supposed to do?

So, and this is an interesting topic, and you and I have have discussed it uh quite a bit in the past. It's uh the one of the interesting things that this concept of um of alignment across the organization comes in. And and this is an aspect of value streams. Are we aligned to the value streams that uh we use to deliver value to our customers? If we look at how then these value streams might be related to each other in order to deliver a particular service to a customer, uh then we can start to see that there are relationships between these. And we might also have internal value streams, and sometimes we describe this as like there's multiple types, like operational value streams versus product value streams. And so there's when we start to think about that and then think about our how we scale up across the organization. We might have like a line of business that could be something like retail banking, and under retail banking, we might have our value stream that delivers mortgages and one that delivers credit cards and so on and so forth, which become the product value streams. We might have multiple products in value streams, which is another interesting conversation to have. So I'm gonna I'm gonna stop there because I I can talk about this for quite a while.

And that's our time for today. So I as you were describing that, and it was a great description, Peter. I just wanted to pick up. Um, as you're describing it, what's going through my head is this is analogous, as we said last time, to how do I scale from a single delivery team to you know a delivery team of 50 people? And I don't want a team of 50 people, that's unmanageable. We're gonna have many teams of five, six, seven people. So then the coordination between those becomes the headache. This is exactly the problem that we're seeing describing here with value streams. If I think of that value stream, the single value stream, we can kind of get behind that straightforward. But when we have multiple value streams, there's a couple of things that we need to worry about. One of them is what are the dependencies between those value streams? So as a value stream, I want that container to be pretty unique so that that um if if I'm in the credit card value stream, I can get a credit card product out of the door. There's value is the credit card product going out of the door, and we need all the skill sets in there as minimize the dependencies so they can get it out of the door, number one. The second thing, we need a really clear understanding of what the value is because that's our alignment. It's like the magnetic field that comes underneath. If I've got a sheet of paper and I've sprinkled iron filings, I'm thinking back to school and science experiments, right? You bring that magnet underneath it. That's that prioritization driving to align all of the activity. So typically in a value stream, we talk about value. It can be profit and loss, it can be some other metric, some other way of doing it, but that's going to guide our conversations as we go through it. And then the third thing that I'm going to start thinking about here, and this is where I think value streams often fall short in many organizations, is understanding the customers' needs. So again, if I look at the credit card, we all, if everybody dipped in their hands in their pockets, pulled out the wallet, and looked at the credit cards, the likelihood is each of us has a different credit card, which does has different characteristics. And the reason is because we have different needs, each of us individually. So that's often not reflected in a straightforward value stream. So how do we bring that element into it?

Yeah, the the creation of that feedback loop. And this this is an important part because it ties into a lot of a lot of elements. And uh the one of the problems, one of the reasons that looking at value streams becomes quite critical is understanding at the slowest level, like, can I actually, with my value stream, actually put something of value in front of a customer within the team? Can the team actually do that? Am I able to uh manage that? So that that means I've got to look at not just the the process and the steps I have to go, I've got to look at the underlying systems and the architecture and the overall domain that that's operating within, because this is all interrelated in order to be able to do that. If I end up with value streams that cannot independently deliver, or I need multiple teams all coordinated within each other, who cannot independently um deliver anything of value out into the structure environments, then I I run into it, I run into this kind of ball of mud problem where it becomes very difficult to coordinate anything. And this is actually what you see in a lot of large organizations is they they run into this this problem because they're they're often thinking about things in very much this kind of functional silo. And for me to deliver a credit card, it's not a case of I have a team aligned to a value stream that can talk to the customer and understand what the customer's needs are and the that delivery of that value. I've got a lot of independent pieces that I now need to provide handoffs between each of those areas in order to get to the point where I can even understand uh what the customer got from what I delivered to them.

Um I think one of the things we should recognize is, and this is I I've come to this from those conversations with clients about value streams. We want value streams that serve customers. So we start talking about customer experience, value streams. What we very quickly find out is you can't, in many organizations, it's really difficult to create a value stream around, for example, a customer. And if I look at, let's just touch on financial services since we've been using that as the example. Uh if I look under retail banking, there, you know, I've got a value stream, for example, around young, uh, young families, you know, first child in the household, maybe they're trying to or just have bought their first house or apartment. And so they have very specific financial needs around the products that are in there. But you're not necessarily going to have a credit card product in that value stream, which is a different team to the credit card product, for example, in a you know, somewhere where the children are now leaving home and the parents in that family, that whole thing is now moving in towards downsizing and retirement planning and things like this. They have a different need for those credit card products, but you don't separate out that team that's doing the credit cards for one to the other. So we need to under what you often see as you scale is a focus on product value stream. Um call it product or service because it may be a service like onboarding accounts or something like this. But basically, it's a it's a set of functionality that serves a purpose within a broader context.

Yes, and uh and that's that's where we start to run this point because once you get into this, I've got a service now that I'm delivering, I need to ensure that service is up and it's running, it's reliable, and it needs to be consumed by the value streams that are delivering the the customer experience, uh, so that the customer experience can be uh understood and uh and delivered. So now we we end up with this interaction between them. And actually, this is something that um uh team topology is uh is quite a good example of a model that uh talks to this, where teams where you have this concept of uh platform aligned teams and uh stream aligned teams, but also around interaction models, the the concept of collaboration between between teams, uh, or where one area is provides something to the other as a service, which is so it's purely consumed as a service, there's no direct interaction between them, it's uh there's a service contract between the two areas uh to more of a facilitating relationship where one's facilitating the the activities of the other helping that uh move forward. So that um though those kind of uh those different team interactions, those different uh areas of interaction also apply to to value streams. How do value streams interact for the purposes of creating those customer experiences from the various different product value streams that exist within an organization?

Well, I and I think this is your you know, the idea that you and I have been talking about about nested value streams. This is where I can imagine, and we've just been talking about customer-facing products like credit cards, but there are other internal employee-facing or service-based products, like let's say it's um the banking backbone in a financial services institution which moves the money around, or data warehousing, data processing, analyzing solutions where all of the data sits. So the analysis of the next value stream up is drawing from a central pool of you know, a data lake or some sort of processed uh consistent global data service. So each of these examples are where you've got some internal service or product which is being consumed internally, where, for example, our concept of maximizing profit and loss falls down because these aren't profit and loss value streams. But then you've got the credit card or the secured loans or unsecured lending and things like this, where where this those sorts of product value streams now have a very clear product and loss, uh profit and loss um metric or or dimension to them, but they're probably consuming in their own right these internal value streams that we talked about. And then go ahead.

I was gonna say then on understanding the those interdependencies, this is part of the reason to understand those value streams and how do those value streams then interact to provide that uh service is the critical part there. And I I don't think we spend enough time considering that a lot of people.

Well, and because that that's uh yeah, if this becomes the loudest voice in the room or the most, the largest profitable value stream in the room, if you like, if we're not careful, right? So we need we need some sort of prioritization mechanism, some trade-off mechanism that allows stakeholders to to transparently and and visibly understand how these prioritization decisions are being made so that we don't lose time chasing things that, you know, functionality that is just never going to end up on our plate, when we don't, you know, sub-specialize our service value streams so they only serve one of the consuming value streams. They don't actually serve the needs of others as well, right? So these are some of the tough trade-off conversations to have. And to your point, you're not you you don't just want to visualize them, you want to actually take the time to understand how are these going to operate.

And I I think uh I think that's gonna be a wonderful topic for the next uh time we uh get together. So uh if we if we look at back at this, I think this is a really great conversation. I really enjoyed sort of the areas we're exploring here because it it is a it is a complex topic, um, and there's many, many different angles to it. So uh if you were to sum this up in in three points, how would you kind of sum this up for us?

Well, um I I really liked the the way that we described the interactions between the value streams. That first of all, the sort of metaphor or the the analogy with with how we scale up our development capability from one team to many teams. There's a lot we can learn from that. Some of the similar conversations that we have to have. You touched on some of the boundaries between value streams do not have to be physical handoff boundaries with me talking to you about what's going on. They can be service calls, they can be sort of at the technical level, not at the human interaction level, if that makes sense. I think that's something to really remember about this. These value streams can be technically distinct rather than anything else. And then the third thing is just this idea of spending time to understand how to make priority trade-offs between value streams. Because when they're nested, you're going to have multiple needs coming into some of those lower in the stack value streams where they have to be able to trade off those needs. And whether it's a first come, first serve, or there's some sort of stakeholder prioritization conversation or whatever it is, we need to give attention to how that is managed to optimize value across the organization, not an individual value stream.

I I I completely agree. And I I think uh I I've got some good ideas for what we should talk about next time, which was going to be definitely around some of those, like how do you create that alignment, how do you then create that prioritization as well, around like how do we make those trade-offs uh in between the different uh areas. Well, so until next time, if you if anybody wishes to reach out to us, they can at uh def feedback at definitely maybeagile.com. And uh as always, Dave, thank you very much and enjoy these conversations.

It's a pleasure. Thanks again, Peter.

You've been listening to Definitely Maybe Agile, the podcast where your hosts Peter Maddison and Dave Sharrock focus on the art and science of digital, agile, and DevOps at scale.