This is a transcript of episode 131 of the Troubleshooting Agile podcast with Jeffrey Fredrick and Douglas Squirrel.
Squirrel describes a client of his that is “thrashing”—trying to do so many different things that they aren’t getting anything finished. We reflect on why this happens, including the idea that being busy feels good while increasing focus can mean disrupting comfortable routines. We suggest asking “why can’t we finish this today?” as a forcing mechanism for discovering bottlenecks and alternative approaches
Squirrel: Welcome back to Troubleshooting Agile. Hi there, Jeffrey.
Jeffrey: Hi Squirrel.
Jeffrey: So this week you were telling me that you were in an experience where you saw people doing a lot of thrashing. That sounded really interesting. Let’s talk about people thrashing.
Squirrel: Absolutely. And it’s funny to talk about people thrashing because thrashing is something that I heard about first, I think in my operating systems class at university. It’s something that system admins would think about a lot. So it’s something that a computer normally does. But I’ve got a situation where I’ve got a client and I’m going to, as normal, protect their confidentiality by not saying too much about who they are. But I’ve got a client who has their team thrashing and let’s define first what thrashing is.
Squirrel: So in technical terms, if you’ve got a computer thrashing, then what you do is you try to run way too many processes on it, all of which are using computer resources like memory and CPU and so on. And the computer spends so much time trying to allocate the resources to the different processes, all the different things it’s trying to do that it never actually does anything. So you could imagine opening 500 tabs in your browser or opening 500 copies of a game or something like that, or trying to play 500 games at once on your computer. The computer would try to handle that. It would it would it would do its very best to try to get the memory in use by one to be freed up, to be used by another and so on. But actually the time that it takes to do that switching takes all the time that the computer has. And so you can wind up with a machine that is at something like 95 percent utilisation. It’s very, very busy and nothing is happening. And that’s precisely the situation of the humans at my client.
Squirrel: So to briefly describe their situation, they have a very ambitious set of goals, as lots of my startup clients have. And they are very complex types of activities, so you should think, very, very detailed scientific type activities that they’re doing that require a lot of preparation, a lot of physical processes that they’re working with. So these are not things that you can just whip up in a morning to try out a new lead generation method, a new landing page for your website. This is much more substantial effort and therefore the switching is much bigger for them. And they also try to document their activities in tremendous detail. So a lot of their activity is, in fact, writing 20 or 40 page strategy documents about what they’re doing. And the result of all this is that they’ve created a spreadsheet and I couldn’t quite tell whether it was percentages are just decimals, but it was it was too many decimals for me. They had people allocated to like 0.03% to this project and 0.07% to that project. And you look at that and you say, how on earth can these people possibly keep track? And in fact, they’re not getting where they need to go because they’re spending so much time on the resource allocation. Something I often say to folks, and I’ve said it to them, is, you want to be looking at throughput: that is, what are you actually getting out to real customers where people are getting benefits from what you’re doing rather than utilisation. And it seems to me they’re really optimising for utilisation here, which is a dangerous thing to do. And Jeffrey, I think you’ve seen the same thing often.
Jeffrey: Oh, absolutely. I’ve seen that many times. And in fact, this idea of people who are working and being busy as opposed to getting output is something that’s very common and that various sort of processes that you come across in Agile try to come back. And one of them is to have, you know, WIP limits, work in progress limits for exactly this reason to make sure you’re not spend all your time thrashing back and forth between different projects. The interesting part about the story when you told me this all resonated with me, if I’d seen this before, but really made me think this is a good podcast material was what happened when you had some intervention there. And suddenly they they did something where they stopped thrashing and they suddenly had the opposite. And can you tell us what happened next then? At least for a short time this period where you the thrashings stopped.
Squirrel: Absolutely. Well, and I shouldn’t take any credit for it. They did it on their own, which I was very proud of. So I gave them just the tiniest push toward being more focussed and darned if one of them didn’t say, ‘hey, let’s be really focussed’, even used in Agile term. He said, ‘let’s have a sprint’. And so they went off in a cross-functional way across the whole very small company. Everybody worked on this particular activity and actually made tremendous progress toward a result, which is the first time in a while that they had this type of result in this kind of predictable way. So that all sounded wonderful to me. But the problem was that others in the business said, ‘well, hang on, we have some plans. And their situation is one where, like I say, they can’t just whip up something in a morning, there’s quite a lot of preparation and physical items to get to get ready, not just not just software. And the difficulty is that when they actually went and tried to do this sprinting behaviour, so they were they were focussed on a small set of things. And in the sprint, only on one, and getting people across the business to collaborate on it.
Squirrel: The problem is that that wasn’t very comfortable because folks had made plans for things that they were going to do. They had physical processes going on that take a certain period of time to mature and they were not able to switch those off. And that felt very, very uncomfortable and very inefficient. The problem is that before in the much more ‘efficient’ situation where there was tremendous resource planning and very detailed slicing of tasks into small pieces for everyone to do a tiny bit of, they weren’t getting anything done. So my response to them was, well, I recognise this is uncomfortable. And one of their very good points was, look, we didn’t have any warning that this was going to come. And it’s kind of at the last moment and we’re doing it against a deadline. And and this is all very unplanned and inefficient. And I said, yes, you’re absolutely right. We would be better if we planned to have this kind of focus. No argument. But this is what focus feels like when you are focussed. It means you say no to lots of other things and you don’t thrash.
Jeffrey: Yes. Suddenly you had a very clear priority, like one single priority. This is what we’re all going to do. And then suddenly great progress is made. But actually, it was the resistance to that, that people felt is what really resonated with me, because this is something that I’ve also seen and been seeing where people will say, ‘oh, this has come down now. We’re working on this thing’. And people will say, ‘oh, it’s a fire drill, right. Fire drill.’. Clearly a bad thing. It’s like, well, you know, in a sense, maybe it’s not a fire drill, maybe it’s an actual fire. This is the real thing that I think that’s that’s the challenges you’ll have. Something that actually is very important is the most important thing. But it’s not so clear as a fire, you know, when something is a fire and it’s like life life threatening, then people will understand we have this interruption. We have to we have no choice. But it’s difficult for people to voluntarily discipline themselves to actual focus, to actual prioritisation that this is the most important thing. And therefore, we’re going to do this one thing. And I was trying to understand why this is. Why is it that people suddenly feel so uncomfortable with it? Is it just a case that they have routines that are being disrupted and they they just like the routines? Is it goal displacement where they had wanted to have a good outcome? And so they designed a good process to get a good outcome, and then later, when something disrupts that process, even if it has a good outcome, they are attached to that process when they should be attached to the outcome. And there’s many different factors that are possible. And I think that’s what worked for me, I thought was so interesting to talk about, is how comfortable thrashing can feel, that having your time sliced across lots of different projects. So you’re always busy at work.
Squirrel: You’re never bored. There’s always something happening that I’m going to be working on Project A or B or Q or X or Alpha, something.
Yeah, that’s so efficient. Right. And if I can’t make progress, I’d like to really make progress on ‘A’ but I’m blocked on someone else who’s not available. No problem. I’ll go to Project B and if that fails then Project C, you say there’s there’s a cascade of something that’s under my control while I’m waiting for all the things out of my control. People seem to adapt to that pretty well. So they might be working on priority Q but they’re happy because they’re like ‘yep I’m doing my part.’
Busy Doesn’t Mean Well Utilised
Squirrel: If you think about the opposite, if you think about what happens if you’re not fully busy, if you’re not fully utilised, that feels really empty and useless and it can feel like you’re stealing from the company because you’re doing work, you’re turning up to earn your salary, but you’re not actually doing the thing that everyone else is working on because it may actually be most useful for you to sit on your hands for a bit and let us work at the at the bottleneck to clear the bottleneck in theory of constraints terms and get that cleared. But you won’t be fully utilised and that can feel really uncomfortable. I’ve especially encountered this with folks who have a history in agencies where the goal was always to remain billable. So make sure that every minute you’re spending sitting in the chair at your cushy office, writing software or whatever it is you’re doing, it is a minute spent that is billable to the customer. And if you always do that, if you’re always billable at an agency, you’re making money. Whether you’re producing value is a different question. That’s what can happen, if you have a billing agency kind of mindset, is you get divorced from value and you get worried about utilisation. That’s why I’m always talking about we want to focus on what the throughput is. Are we actually getting the results from the process that we’re following? Because we could follow the process all day, really spread ourselves thin across a lot of different projects and get nothing whatsoever done. Make no money for the the investors.
Jeffrey: This tendency of people to want to be utilised, it has a natural counterpart in management who also want people to be utilised. I can’t tell you how many times I’ve heard the phrase ‘well we have to keep the developers busy. We’ve got to make sure the developers have something to do’. Like the worst thing in the world would be to be paying these expensive people and then not have their time used 100 hundred percent. It feels like waste, what are we doing? You know, we’re chumps if we’re paying people and we’re not using every second of the time that we’ve paid for. But that’s so wrong. And if we take a small step back. This is mistaken for lots of different ways. And the question is, why is it so hard to see? One thing you and I were discussing is if you’re there, the person making the widgets and maybe uselessly. But you’re like ‘look, I mean a widget, there’s a thing.’
Squirrel: It feels good. Yeah. Look, I produced a thousand lines of code today. I’m not quite sure where they’re going to be used or if they’ll be deployed or if they’ll be erased tomorrow. But at least there’s a thousand lines code. I feel like I did something today.
Jeffrey: Yep. I went to these meetings and I took some notes and I sent some emails and I checked the documentation and I wrote some documentation, the various kind of work products I can point to. I made a diagram. I have a PowerPoint and it feels like progress. But there’s it’s so much harder to have a system view to say, well, what’s how is this work in the context of the whole system? And this is when we’re talking about this, I thought, well, good thing IT Revolution has this fantastic book by Dominica DeGrandis called Making Work Visible, which is about exactly this, that it’s so easy to have our work. The tangible work product is visible, but the overall context of the work and the implications of it, when we have things like too much work in progress, like when we have unknown dependencies, unplanned work, conflicting priorities, neglected work, that these things all end up costing us time. But they do it invisibly in the sense painlessly when we try to address it. When I said, right, actually we’re going to focus and some people can be really busy and some people will be doing it again to contribute and someone else might not have anything to do on this most important thing that’s so uncomfortable.
Know the Critical Path
Squirrel: It absolutely is. I’m reminded of something I read once, and I’m not sure if it’s absolutely true, but I want it to be true. So I’ll pretend it is that if you walk up to anybody who works at Space X, one of Elon Musk’s crazy projects and trying to get spaceships to Mars, if you ask them, what are we doing here?
Squirrel: They said, well, well, I’m helping us get to Mars. In theory, you’re able they’re able to then say, well, you know, this strategy document that I’m writing is all about getting approval from NASA to get to the next stage of our tests, which will then allow us to build this, which gets us to Mars. They can they can draw a line from whatever it is they’re doing to getting to Mars. And I think far too often when work is invisible, it feels like, well, I must be doing something. I must have something to do with where we’re going. It probably would get us to Mars. But you know what? I’m not really sure if it’s going to help us at all, get people to be really honest about it. They might say for an awful lot of the work that they’re doing, this is this is utilisation. I’m being utilised for sure. Throughput not so sure.
Jeffrey: I think a lot of people do think they’re going to have throughput because they’re only looking locally. You know, they say, yeah, I have throughput, stuff comes into my inbox, it goes out of my inbox, it goes to someone else. There’s plenty of throughput because again, the lack of the system. I find it really interesting, this challenge between the people’s attachment to a process and the routine and the conflict with that, with learning and improvement and in a sense trying to optimise for the situation you’re in to be context dependent. And I have a trick that I often used, which is to say, look at it in terms of, you know, well, what’s preventing us from being done today. Mm hmm. You’re worried that we have a deadline in September. I’m not worried about September. Can we get it done on Monday? Is there anything that prevents us from being done this afternoon?
Squirrel: Oh my God Jeffrey you’re absolutely nuts! What we’d have to do to get done by Monday is we’d have to do project A and we’d have to cancel Project B, and then we’d have to see whether we could even find anybody to outsource C for us. And that would take us at least two weeks.
Jeffrey: Ok, so you’re saying we could be done in two weeks?
Squirrel: Oh, well, yeah, I guess we could.
Exactly. I used a very provocative question to service what are the real obstacles to being done? And very often those obstacles that surface are things that we hadn’t actually considered doing before, they give us new options. And very often we want to take those options and are things that would not have come up, in my experience, don’t come up in the sort of standard, routine approach where we have tried to plan things forward looking. For me there’s a kind of pull system. I’m trying to start from the end and saying, right, I’m going to pull the output. What prevents me from having it now and have that cascade back up the chain and so often from this, I describe what I’m doing to people- not to people. I describe to people, what I’m doing in the project, in the project management space, saying I’m trying to do a critical path analysis. I’m trying to find those things that are on the critical path to keep us from shipping and anything that’s unknown is on the critical path.
Jeffrey: So the first job is to drive out uncertainty, and so we need to know if something’s on the path or not. And if it’s not, well, great. I’m not going to probably pay attention to it right away until we have all the uncertainty done. Until I can say yep, this is what we know, this is the critical path. Everything else that was planned, that was considered. You know what? If we don’t need to do it, it’s not a critical path. It may not get done. It is certainly not going to get done right now. We’ll figure out a way. But first, let’s figure out this this critical path. Now, I was really happy to use the term critical path analysis here, because I’m not sure I’m using it exactly correctly. And I’m hoping that one of our listeners will write in and tell me ‘no Jeff, that’s exactly right.’ Or they’ll give me the correct name for what I’m doing here, which is to try to construct the critical path for the project and particularly with this goal of driving out uncertainty.
Squirrel: Fantastic, well, I really like the thought experiment of why couldn’t we go live Monday, it’s not typically something that you absolutely have to do that you’re not. I’ve even said things like, ‘well, if we if we had guns to our heads, if if aliens were going to blow up the Earth, unless we had it done by Monday, what would we do?’
Squirrel: And that’s not because I secretly have a bunch of aliens who are going to blow up the earth if the team doesn’t get done. That would be a threat that I couldn’t deliver on. But it really does focus the mind on what would actually be needed. And it often turns out to be much less than you thought. For example, I was working with a different client and they’ve been working on a project for a very long time. And I said it did the same question, why couldn’t we do it? And we discovered that actually we could go live next week. So I think that’s what we’re going to do. So that was a case where it turned out there was nothing preventing us from doing the surprisingly possible delivery because there wasn’t a bottleneck, there wasn’t something preventing us. The only thing that was preventing us was this thrashing that so many of my clients get into and that I suspect our listeners have seen as well.
Jeffrey: And to kind of come full circle on this, I think what you and I are describing and we have kind of internalised a process of disrupting process rather than having laid out sort of things we’re actually going to do. Our process involves things that are going to potentially change what we had intended to do. And that kind of comes back where in the beginning people are laying out ahead of time. They’re planning a process forward and then they become attached to their plans, attached to the process, attached to the routine. And suddenly when focus comes in, it feels disruptive and uncomfortable because it’s disrupted their plans and routines and process. I think, though, this if we can look at the system view and think about what you really value, the advantage of having a process that generates disruptions, generates learning, generates new information is a more effective way to go. And that sort of attachment to the meta approach, I think, and for me replaces the attachment to the sort of daily routine. I don’t know how my next week is going to be compared to the last week. I know they’ll be different, but I don’t know how. But I’m comfortable that what I’ll be doing will be focussed on this sort of surfacing of unknowns, driving out uncertainty and converging onto a critical path and building alignment on the team as we go. So I have a meta approach that that I’m attached to and feel comfortable with. But of course, I’m also happy to change it if it turns out there’s a more effective way. So this difference in mindset.
Squirrel: It certainly is. And it’s a mindset that may feel threatening or or challenging. That certainly is not one that I’ve seen people adopt wholeheartedly and say, boy, I can’t wait for everything to change and for uncertainty to reign and for me not to be sure whether I’m going to be doing something useful tomorrow or not, that when they look at it that way, rather than the very positive way you just looked at it, they often wind up with a real challenge in adopting it.
Squirrel: The advantage is that if you if you can get yourself to adopt it, if you can get to that mindset and get your team to that mindset, you’re much more flexible and you can focus on the critical path and deliver things in a surprisingly, much more efficient way than you would have in your supposedly efficient super thrashings, super high utilisation method.
Jeffrey: That’s right.
Squirrel: Wow. OK, well, lots to digest there, I imagine some of our listeners have been thrashing themselves and would be interested in talking more about thrashing, which of course, we’re very happy to hear about.
Squirrel: You can get in touch with us on Conversationaltransformation.com, where you’ll find our book, Agile Conversations. You’ll find links to us and free material and things to learn more from and videos and all kinds of good stuff. So have a look there. Get in touch with us, Twitter, email, lots of other ways, all starting on Conversationaltransformation.com. And of course, you can come back and hear what the listeners had to say next weekend, whatever we come up with to talk about next time. And that will be something that you can get to if you hit the subscribe button on whatever app or programme or whatever it is that you’re using, because we’re going to be here next week. Excellent. I’ll see you then.
Squirrel: Thanks Jeffrey.
Jeffrey: Thanks Squirrel.