Pod Insights The Agile Daily Standup - AgileDad Dependencies Destroy Agility and Predictability About this episode Dependencies Destroy Agility and PredictabilityWhy do organizations desire Agile? Two of the most common items I hear organizations and executives identify as reasons for adopting Agile methods and practices are to increase their delivery speed and to become more predicable. When will X be done? There are valid financial reasons for these goals such as: when can we expect to book revenue for product X? How to connect with AgileDad:- [website] https://www.agiledad.com/- [instagram] https://www.instagram.com/agile_coach/- [facebook] https://www.facebook.com/RealAgileDad/- [Linkedin] https://www.linkedin.com/in/leehenson/
Transcript 0:15 Speaker 00 One of our longtime listeners pointed me in the direction of an article written by David Johnson. It talks about dependencies destroy agility and predictability. So I thought I'd give it a quick review. This article looks like it was published in March, not too long ago. Let's do this. So why do organizations desire agile? Two of the most common items I hear organizations and executives identify as reasons for adopting agile methods and practices are to increase their delivery speed and to become more predictable. I just want to let you know those are facts. I hear that all the time as well. Predictability is such a big deal. And I feel like organizations are thriving if they have predictability. But organizations that lack predictability really are struggling and they're tending to blame it on agile,
0:58 Speaker 00 which I think is really, really interesting. So he goes on, when will X be done? There are valid financial reasons for these goals, such as when can we expect to book revenue for product X? In this discussion, we're going to talk about three things dependencies do for larger, for groups larger than a single team. So if you have groups that are larger than a single team, how do dependencies impact them? The first topic is structural dependencies. These dependencies exist because of how the teams, the groups of teams, or how the work is organized and defined. The very nature of team composition can create structural dependencies, such as the creation of a SaaS application with some teams that only work on UI.
1:42 Speaker 00 These will be known as component teams, and they only create code for one component of SaaS application. They would not write the application API or the database code. Similarly, how work is defined in features or user stories can expose structural dependencies. A typical example would be creating three separate features for a single user outcome. One for the UI efforts, another for the application efforts, and a third for the database work. I see this all the time, and this is a crazy thing to do. One of the tips that I can give right away, I'm just gonna throw out a tip, is organizations that are successful using Angela, the ones where the product owner focuses on organizing the work
2:22 Speaker 00 and allows the Scrum Master to help focus on getting the people together and keeping them focused on the goal, eyes on the prize. So in my experience, I'm going back to the article. In my experience, this is the most, this is most often the result of having component teams, but not always. Sometimes the work is broken down by application layer and then sequenced rather than working an entire stack as a whole. This might be the case when the work balance is not allowed to self-level across teams, typically an indication of a push system rather than a pull system.
2:55 Speaker 00 That is also true. These dependencies might not always exist in any given moment, but because of the structure of the group or the work, they represent the potential for an item of work to require the efforts of multiple teams. Okay, so the next one is instantiated dependencies. Instantiated dependencies arise when a potential dependency, structural dependency, becomes a reality for a particular piece of work. The structural dependency is now real. The work must flow between at least two teams to be completed. To carry on with the examples that we've given, instantiated dependencies would be the creation of an object of type structural dependency.
3:38 Speaker 00 Sounds like development terms to me, but I'm gonna run with it for now. Blockers or blocked work.
3:45 Speaker 00 As an item of workflows through a system, and a team cannot proceed until another team does some other form of work that becomes blocked. They cannot proceed until someone else finishes dependent work, so they will typically add a blocker to their workflow. Teams that are being honest and transparent will keep the clock running while the item is blocked. Other teams will stop the clock, which could help hide the negative effects of the item that's being blocked. The clock should keep running, adding the, adding the cycles and lead times while you're blocked. This will shine a light on how dependencies as a problem should be resolved and the impact that dependencies have. That is absolutely 100% true. So far, I'm really enjoying this article. Use techniques like blocker clustering to gather information, which could then later be used for addressing dependencies.
4:28 Speaker 00 Yeah, this is where I talk about cumulative flow diagrams to see where things get stuck, and it's important to have those. Okay. Dependencies are self inflicted wounds. Many managers and executives don't like to hear about self inflicted wounds, but structural dependencies are created as a result of past choices. Choices about team and group composition and structure, choices about how work is represented in the system, and still other choices about how management is measured and rewarded. The mindset of empire building and a reluctance or resistance to change. Uh, structure is just one example. Another unintentional consequence of organizing by specialty into component teams is uncovered by examining Conway's law. Conway's law hypo- hypo-
5:12 Speaker 00 hypothesizes that organizations design systems that mirror the communication structure, or more loosely, systems are designed mirroring an HR hierarchy. As that's the most often way the information can flow. This typically results in a highly componentized architecture of the systems and applications and requires multiple teams of specialists along organizational hierarchy lines to create and maintain such system stacks. This results in many structural dependencies. Back to predictability. If one of your goals is to become more predictable, how can dependencies interfere?
5:52 Speaker 00 Everyone knows that dependencies are just going to cause you to offset the flow of work. And when you offset the flow of work, it's going to cause you to never become predictable because the expectation isn't going to be that you're just going to finish the things that are dependency laden, but that you're going to complete the other pieces of work in addition to the pieces that are there. So the more dependencies you have, the less likely you are to be predictable. It's just the way it is, right? So handoffs and dependencies can increase cycle time. As we mentioned before, if teams are honest and transparent about the work, they'll keep the clock running while they're blocked. Cycle time of their effort will increase, uh, to include the amount of time that they're blocked, and that'll give a glaring show of a cost of delay.
6:36 Speaker 00 So a cycle time increases the time, the value also increases. Thus, the raising cost of delay. Hey, I just said that. All these rising measures are counterproductive to the goal of increasing delivery speed and faster time to problem. If you're measuring anything, you should be measuring cost of delay. The negative effects of rising lead and cycle times and cost of delay increase at scale. So it's an exponential thing.
7:02 Speaker 00 Okay, so I think we get the gist here. I think the gist is dependencies are bad, which we all know that, but dependencies are also inevitable. So how do you get to the point where you can take care of dependencies if they exist? And still be successful in your agile implementation? Besides the game playing identified above with timing of feature creation to influence some metrics, a main point to understand is that every handoff involves a backlog or a queue. Every dependency involves work in motion, going into arrest mode, waiting for prioritization and availability of dependent team or person. That other person or team isn't just sitting idle waiting for that work. They're likely already busy working on something else, some other high priority work.
7:46 Speaker 00 The new work waits until I have availability or something already prioritized that's in flight and needs to be stopped. And these orchestration costs add up quickly and the delays can become real problems. So work sitting in a backlog is inventory. All inventory carries a cost. If you're looking at this from a systems thinking perspective, all the time the work is waiting for dependent person or team to have available is considered a non-value add item. All of this contributes to delaying time, delaying time to value.
8:18 Speaker 00 So if you're trying to reduce dependencies, hopefully by now you you see the importance of addressing the dependencies. And if you do, and you want to do something about it, especially to structural dependencies in your organization, you need to realize you're paying a price each and every time a structural dependency is instantiated in terms of time and money. So if we try to break these down into high level categories or techniques to help you deal with dependencies.
8:46 Speaker 00 I'm going to rattle off a few here from the article and just give you my high level observation about these. So communities of practice. A group of people who share interests or skills or expertise, they'll meet regularly with defined purpose. This is great, but I don't know that communities of practice aren't doing other work. So I think you're still, you're still, you're just dealing with that. You're band-aiding the problem and not dealing with it. Cross-functional, more generalized specialist, I agree. So if you get to cross-functional teams and you have specialists who aren't just deep diving one particular thing, a T-shaped or pie-shaped team, that's really going to help you. Because if you get a whole bunch of individuals that are T-shaped or pie-shaped, what's going to happen is you're going to have a breadth of knowledge as well as a depth of knowledge. And that's where you need to be.
9:31 Speaker 00 Feature teams instead of component teams. Try to get teams that can take a vertical slice all the way through. Once again, full-stack teams always do better with dependencies. Focus on team working agreements and make sure that it's in there. How are you going to handle them, what you're going to do? My number one favorite, and I'm so glad he said it here because if he didn't, I was going to. Reduce WIP. If you take back the number of items you have in flight, it's going to make it so much easier for you to finish things without dependencies. It's going to make it more glaring when a dependency does exist. But I think the biggest message I can give to you is don't wait to do something about this. Do something now. Figure it out. Use some of the above methods that I just discussed and try to figure out how you can implement them inside of your organization.
10:13 Speaker 00 Because if you can get to the point where you can limit WIP, or find a way structurally to take care of your dependencies and make it part of your team working agreement, those are the keys, the ingredients to a successful recipe for handling dependencies. I hope you enjoyed this article and thank you for sending that in. That was a good one. That's going to do it for this one. If you have a topic you'd like for us to discuss, learn more at agiledad.com, we'd love to hear from you. As always, we encourage you to stay healthy, stay well, and stay agile, my friends. Until next time, do take care.
Previous 7 Tips To Read Someone’s Personality in 10 Seconds Next Five Most Forgotten Parts From The Scrum Guide Page updated November 7, 2026 · transcript and summary are machine-generated