Pod Insights The Agile Daily Standup - AgileDad Agile Failed Us After 18 Months - Here we go... About this episode Agile Failed Us After 18 Months - Here we go...
On month eighteen, our average lead time crossed 27 days. Production defects doubled. A supposedly minor release missed its window by three weeks.
Nothing had “broken.” Velocity charts still looked healthy. Every ceremony was running on time. But releases slowed, confidence eroded, and engineers stopped believing what the board said.
This hurt because customers felt it immediately. Bugs lived longer, features arrived stale, and every delay came with an explanation no one trusted anymore.
How to connect with AgileDad:
- [website] https://www.agiledad.com/ https://www.agiledad.com/
- [instagram] https://www.instagram.com/agile_coach/ https://www.instagram.com/agile_coach/
- [facebook] https://www.facebook.com/RealAgileDad/ https://www.facebook.com/RealAgileDad/
- [Linkedin] https://www.linkedin.com/in/leehenson/ https://www.linkedin.com/in/leehenson/
Transcript 0:16 Speaker 00 So a recent subscriber just sent me a field report of all things and I thought this was kind of interesting. So I'm going to read a little bit from it and we'll talk about it. "On month 18, our average lead time crossed 27 days. Production defects doubled. A supposedly minor release missed its window by three weeks. Nothing had broken. Velocity charts still looked healthy. Every ceremony was running on time, but releases slowed. Confidence eroded and engineers stopped believing what the board said. This hurt because customers felt it immediately. Bugs lived longer, features arrived stale, and every delay came with an explanation no one trusted
0:54 Speaker 00 anymore. Agile felt correct because it optimized for local clarity. Short sprints, predictable rituals, frequent planning, everyone always knew what they were working on. Under real production load, however, that clarity became noise. The system optimized for activity, not flow. Work entered faster than it exited and the process politely hid the queue. The failure mode was subtle. No single sprint was bad. The damage appeared only in aggregate, growing whip, aging tickets, rising cycle times, and decisions made too late to matter."
1:36 Speaker 00 So here's the problem. Sprint commitments hid the queue. We committed to fix Sprintscope while production interruptions arrive randomly. Incidents, urgent fixes, and cross-team dependencies displaced planned work without removing it from the board. Unfinished work rolled forward. Over time, average ticket age climbed past 21 days, even though sprint completion stayed above 90%.
2:01 Speaker 00 Says the change, "We stopped sprint communications for backend teams and we kept a rolling priority queue with explicit whip limits for service. We only started when capacity freed up and the measured result was that lead time dropped from 19 days to eight days and, yes, sprint predictability decreased, but delivery predictability improved." Let's pause.
2:24 Speaker 00 This is something that I affectionately call treating a symptom and not treating the problem. This, while it may seem on the surface, oh, well, everything went well and it was better. This is just a symptom of a much more diabolical thing. Okay? So if we go through and say that this team, so let's just start at the very beginning. This team had predictable work that they were doing and they had interruptions that were coming in. My first question is why so much technical debt, you know? And the answer there is because they weren't having a technical debt sprint, obviously. So then my second question is, what are we doing to address the technical debt? And why is the team following more of a scrum bond format where they say with the scrum points, they have certain allocated work that's predictable?
3:07 Speaker 00 And with the combine points they'll use to handle any roll in tickets that come through.
3:13 Speaker 00 Once again, it's just something that you chose not to do. So that is, that is the second big issue I have with this. And the third one is that you're continuing to allow new work to push out work that's currently in a sprint. And that work's just carrying over from sprint, sprint, sprint. That should never happen. That should never happen. I've been doing agile long enough. I don't carry over work from one sprint to another. It's a rare occasion and it's shocking when it happens. People are like, whoa, what happened? And usually when it does happen, it's because I'm trying to sound an alarm. It makes someone aware that something is happening. This is just poor behavior. This is just symptomatic based on things that are handed to the team late and the way that they're processing the information and pushing information
3:57 Speaker 00 that they shouldn't be pushing out of the sprint. Let's look at the second problem because I'm curious to see where this goes. Story points are rewarded for the wrong behavior or story points rewarded at the wrong behavior. Points optimized for estimation accuracy and not throughput. Engineers learn to split work until it fit the planning, not until it reduced risk. Large architectural work was postponed because it looked expensive on the board. Small changes accumulated hide, accumulated hidden coupling. We stopped tracking points entirely as their solution. We track throughput per service and aging work items. Okay, let's pause again. Once again, we're treating the symptom.
4:38 Speaker 00 Story points here are not the problem. Your lack of understanding of how to use story points the problem. And I think that if you're trying to figure out how to get things stabilized and how to get things reasonable, you need to take two things away from this. One, you need to figure out how to manage your technical debt workload. And two, you need to figure out how to handle in sprint or, you know, during sprint interruptions that come in, so that you can figure out how exactly to address these issues. Because to me, this, this is just, yeah, this is a atrocity so far, atrocity so far. Three ceremonies delayed real decisions, the problem. Critical decisions waited for sprint boundaries, architecture reviews slipped to planning, production feedback waited for retrospectives, latency and decision making became more expensive than bad decisions.
5:24 Speaker 00 So the change was we decouple decisions from the ceremonies, architectural and operational changes happen when the signal appeared, not when the calendar allowed for it. Okay, let's pause.
5:37 Speaker 00 Once again, I think we're addressing the symptom and not the problem. We're back there again. I don't think the ceremonies are what was causing the delayed decision. I think what was causing the delayed decision is the people who are doing the work being overwhelmed with sprints that were too full. You're exceeding their, the team's capacity, you're exceeding their velocity, you're expecting them to play the tradesy game, take one out, put one in, and you know, as long as same story points were good. And I can tell you right now that these things never work. When you're forcing your team into sprint roulette, as I like to call it, you're not solving anything. And the ceremonies are going to make it worse. And that's exactly
6:18 Speaker 00 what's happening. But this is all symptomatic of you not having control of your whip limit, and not having control of the items in the backlog, and not having control over the inbound flow of work that's coming in, and not having control of your defects. It's all the same.
6:34 Speaker 00 I think the key here is you need to check yourself and ask yourself the question, what is our current technical debt load? Why is that that load so big? And what are we doing collectively as a team to solve the problem? Not one individual, not 20% of their time or 10% of their time. As a team, are we blocking out a whole sprint to address the top 10 debt related items? Then at that point, are we addressing the work that's coming in? So that the debt queue doesn't continue to grow. Are we leveraging known but not resolved statuses? Are we looking at things like cumulative flow diagrams to prevent things from getting stuck in certain states? What I can assure you is that in this situation, none of these things are happening. Because this particular team in question is only looking systematically at
7:17 Speaker 00 the things that are symptoms instead of the things that would actually solve the problem or resolve the issue long term. So my challenge for you is if you feel like you're going through some of these things, don't just let the data drive your decision. This is one of those cases where you need to figure out the why behind the what. And once you do, you'll realize that making some pivots and some adjustments is definitely the right thing to do. So that way you're not addressing the symptom, you're addressing a real problem. That's going to do it. I hope you enjoyed this episode. If you had a topic you want us to discuss, learn more at agiledad.com. I'd love to hear from you. And as always, we encourage you to stay healthy, stay well and stay agile, my friends. Until next time, do take care. Bye now.
Previous Ernest Shackleton and Leadership When Everything Falls Apart Next What Is Scope Churn? Page updated November 6, 2026 · transcript and summary are machine-generated