 Mary someone that I think Mary and Tom don't need introductions They have been true legends in the field of agile software and Mary was the one who started the whole lean software development movement deeply deeply influenced At least my thinking and a lot of other folks in the community So I want to thank you Mary for everything that you've done over the years and it's so nice to have you part of India back. I think we had you a few years ago, but it's been a while So it's great to see you and the virtual conference made it much more easy and possible to All right, so without much delay, I want to let Mary start with her very important topic Which is more relevant now than ever before with the pandemic. So Maybe if you're gonna, please go ahead and share your slides So good evening or morning or whatever it is where you are everybody. I want to talk about a week several months ago here in Minnesota where I am it was March 17th 2020 and where you are, I don't know what exact week, but it was around about the same time and Everything changed like very fast. It was a call to action It was time for us to well first of all Everybody needed to go home whether they were in school or working and We were called upon to figure out how to make that possible so they needed equipment tools licenses they needed the kids needed internet access and hot spots and maybe even lunch and The professors had to figure out how to use new tools. There was training support How do we put these meetings together? There was security to worry about how do teams talk to each other all kinds of stuff? And it was everybody and it was three days or maybe it was a week or maybe it was tomorrow Depended upon but you know, it was very fast impossibly fast And it wasn't like you could say no, that's not possible because had we known Four weeks earlier that we're gonna do all of that stuff in the short amount of time We would have said not it's not possible, but that wasn't the option in this case So around the globe the impossible had to happen and it did For example my granddaughter got stuck in Morocco and We had to figure out with travel suppliers how to get the two of them Back home through Canada where they were living where their car was parked and back to Rochester, New York And we were there were a whole ton of people trying to move people around the world The professors and the kindergarten teachers and all of those folks had to learn how to teach online The stores that did remain open Had to figure out how to get people to buy things in this far as I was concerned I was not going into a store and Every store around me took maybe one week or two weeks or sometimes a month to figure out how to make it possible for me to Order online drive to the store Call inside they came and put my phone in or whatever I was buying in the car and off I drove and I still shop that way and Maybe a few home deliveries, but mostly drive to local stores and pick stuff up that I ordered online and zoom became a verb and Oddhug teams just swarmed one problem after another I had two stores here one that cons that started out with the capability of having all their inventory and money for about 2% of their Clients they could do this curbside what we call curbside pickup Within a week they had managed to scale and within two weeks you couldn't tell that it was the same service It was so much improved a second store around here two weeks later Nothing had changed because they were waiting for some contract vendor to have some time for them Well at the first store I'm sure what happened is a whole bunch of people got on a team And it wasn't just so far people was all the different people needed to know how to get the inventory right Figure out how to do the massive amount of picking and how to where to store the stuff How to deliver and how to get my Information into the store whole bunches of problems with every one of these things and a team got together one problem after another and figure it out how to solve it and It wasn't so for teams that did this it was software engineers joining teachers or Star workers or procurement experts or call center staff or whoever and Everybody getting together and picking one piece of the problem and figuring it out And then the next piece of the problem and making it workable and then making it better and then making it better Everybody knew what success looked like we were home in three days like it or not Do we have access to our teachers? Everybody in that team didn't have to be told when they were successful They knew when they were making progress and speed was essential and you know what it wasn't like you had a choice It wasn't like these impossible things you could say a winner. No, that just can't happen It was a okay. What's the what can we do towards that goal? so many companies went through this digital transformation that they were thinking about in less than a week for example Twitter had been Working on exploring and doing experiments with working from home And at that point they had two or three people or two or three percent of their workforce working from home And they've been working for two years and it took like a week and everybody was home They hadn't thought it was possible. So this massive experiment which turned out actually quite well for them I'll tell you more about that in a minute But lots of companies took this digital transformation that they were thinking about and they had to make it happen They had to make it possible for everybody to work digital and in this case You see that it was not about software It was about providing the technology infrastructure to accomplish something that was really urgent and much bigger than software And when we are in teams, that's what we tend to be thinking about Something is bigger than the software because actually software is doesn't ever really matter it's always about something bigger and As you will see on this next first part of the presentation I think that it's important for us to be on the team that deals with the bigger problem Not on a sub team that somebody else tells what to do We learned a lot about each other when we were home because we were talking with our coworkers And we could see that guess what they're real people. They have children. They have dogs in The background we saw some guitars hanging and some bikes hanging and back here is all of an assortment of some of the pictures that Tom has taken over the years and We learned empathy we learned to think about our co-workers or somebody not just that corporate person That comes into work every day, but somebody that is sort of a real person And we learned a lot about management Because the supervisors couldn't go into work and see people and check out what they were doing and stuff like that They so repos responsibility just basically had to replace this Micro-management or close supervision that many supervisors had been used or it didn't Because there was a bunch of what's called workforce optimization software You know where you could check and see whether or not somebody was sitting there and what app They were in and all of that sort of stuff and monitor exactly what was going on But you know Sorry to interrupt. I think some people are finding that your slides are completely going dark They're not able to see if you don't mind. Could you just stop sharing and start try once more? Thank you, and if you could just try again. Okay, and Share is over here and Here we go Okay, and just turn your camera back on. Oh Yeah, okay. We can do that camera Okay, all right, let's try that. Thanks. Sorry for the I was turning the camera off if that helps too Let's try it if it doesn't I'll let you know. Okay. I see Hopefully people are able to see I See a lot of hands going up. Thank you guys Barry O'Reilly wrote an interesting blog called why your company isn't profiting from remote productivity And he said in a distributed world where people are all over we have to relearn how you think about management and one of the most important things to think about is You need to create a strong connection between work effort and customer satisfaction You need to have an immediate connection between somebody doing work and Customer receiving the benefit of that and that's motivating So if we think about that And we say has this here been successful this working from home thing If you define success as productivity Then you know what at first there was more output for the same amount of input and later on productivity fell and That's interesting, but that's probably not the right way to define success as Jeff Paton just spent a whole bunch of time telling you Success is about output. It's about the delighted customers It's about the kids at home able to see and the professors able to teach and When teams have responsibility for achieving the customer satisfaction making it possible for me to drive up and pick up my groceries Yes, absolutely working from home was as successful as before teams with proxies Maybe not so proxies by proxies. I mean two things one thing I mean is proxy measurements like things like cost schedule scope member of tasks done stuff like that and I also mean an Intermediary between the team and what it is there and their customers Something that is an intermediary connection and that's that's proxy to like for example The business is a proxy for allowing software teams to be connected with the customers So if you think about teams with proxies, maybe it's been not so successful What Barry O'Reilly says is those managers who controlled who went for outcomes rather than outputs They didn't have a whole bunch of troubles switching to remote work They might not have liked it, but they didn't have much trouble because their natural impulse was Here are the top things figure out how to solve them and here's how we're gonna know that they're they're working Or here's how you can see that they're working and then support the outcomes that helps solve the problem as well Those managers didn't have to really change the way they manage because that worked remotely just as well as it worked while people were in the in in the office But for managers who are used to measuring tasks and having intermediaries Proxies figure out whether or not it was the right thing. That's something different Twitter I said I was gonna mention again They switched very fast to working from home and discovered You know because they had to do it that it was actually really successful and a whole lot of people loved it now That doesn't mean everybody likes it. It doesn't mean everybody wanted home But they estimate they said in May that they're going to allow anybody who wants to stay working from home can and they Estimated about 50% of their people are going to work from home. It's not about letting people Work from home. It's about letting people work from where they feel the most comfortable working from they don't have to ask for Mission they don't have to feel guilty. They work where they would prefer to work And I want to show something more than just that I think the problem still exists. So maybe if I can just request you to turn off your camera, let the slide Thank you so much and does that help Oh Folks, are you able to see the screen now better do a thumbs up? Doesn't seem to does it go there? Okay So maybe it's the video on top. Well, you don't need to see me. That's for sure Okay so I want to talk about job content and This is from a very recent book. It's a book on economics by Roger Martin And it's kind of an interesting book About that was finished just in January before there was any concept of a pandemic and in that particular book He is talking about when more is not better Basically the stuff that we've learned in the last few months now The most interesting thing I found in that book was this chart and this is a chart of the workforce in the United States divided into four groups the first group is the first group is People doing remote work and it is sold locally or excuse me routine work that is sold Locally and in the US about this is the percentage your workforces goes from 50 zero to 50 and about 40 say 4% of the workforce in the US does routine work that's sold in their local area and the next piece is Other routine work routine work that's sold Someplace else for example if you happen to do Outsourcing and people give you detailed tasks and do you're doing routine work? I'll talk about that in a minute and it's sold someplace other than where you live and in the US that's about 20% of the workforce and in Other the next category is creative work that's sold locally Like if you're an architect or something like that and then the final category is other creative work That's sold outside of your local area now the important thing here is how he talks about creative work in Creative work workers exercise meaningful independent judgment and decision-making in order to do their job and that's important meaningful independent judgment and decision-making and It's not here's a task do it It's here's a goal to think about and have your ideas about how to make it work So if we look here And the first cut this here is in the year 2000 over here on the right We have 12 years later in the year 2012 and if you look at here There's two boxes and one is hiding behind the other and this is routine work So locally and it has just about the same what you see is a salary range here bottom to top and the salary range used to be from about say Oh $10,000 a year to about say 30 or $5,000 a year and hasn't changed and here is the next one and It is routine work sold outside the local region and there's fewer people doing it It went from 20% to 18% and the upper salary is just a little bit higher in 12 years Here is creative work sold locally as I said people who are artists and that sort of thing that sell Or if you are working at a local store and doing the software for it or something like that Then the salary range is about 22,000 to about 60,000 used to be and it's gone up now to maybe 75,000 and there's a few more people doing that and in 2012 This was the green spot was really where you wanted to be Creative work sold outside your local region because at that point in time the salary had gone from say 82,000 to 100,000 a year and That was the place where there were the fewest workers only maybe 12% but that is definitely where you would like to live and That's where a whole bunch of software engineers would love to live and should be living and So what I'm saying is look at your job if you're completing tasks That's not a kind of job you want you want to look for a job that gives you the Responsibility to figure out how to delight customers not have somebody else tell you what that means So I'm going to talk about this word respect. I Could call it responsibility, but in this case. I think it also shows respect when workers solve critical problems without proxies between them and the results of their work and I'll tell you I'm big on this. This is not the first time I've talked about it, but I think is really important to think about Respecting the people who do software engineering. I Want to say how we do this in our world because this is not something that's hard to do There is a model and it's I'm going to call it the responsible engineer model It's it's it's if you look at SpaceX. Here's a Falcon rocket and it has various stages This is what systems engineering is all about and you can see the different stages And there will be a chief engineer or a responsible engineer for every single one of those components Landing legs feel fins you name it and that there will be a team and within that team the responsible engineer will have some people responsible for subsets of that components and there will be a responsible engineer for the entire rocket and the concept doesn't have any SpaceX doesn't have a software department the the the payload has a soft has a Hardware piece and software engineers and all the people necessary to do everything in the payload and Every single one of those components Has a team and their job is to make sure that their piece works and does its job as part of the whole thing now This is February 6th 2018 when three of those rockets were tied together and put the Falcon heavy up into the sky and they Landed the booster lock rockets because they've been trying for five years to figure out how to get those things to come down safely and Be able to be reused instead of dumped into the ocean or burned up in the sky Because if you can drop the price of putting stuff in space dramatically you can put a lot more stuff in space and do interesting things Between 1970 and 2000 for all those many years the cost to put a kilogram into space was a you know $18,000 per kilo but now it costs more just under 3,000 us dollars to put a kilo into space that's a factor of seven reduction and That reduction came through I would propose really intensive and good engineering and They operate on what they call the philosophy of responsibility. What does that mean? That means engineers are responsible for the design and development of a component and for making sure they understand how their component should operate and how What its job is as part of the overall system and they're not just responsible for making it work they're responsible for making sure that it does its part in the overall system and the launch director at SpaceX John Mertroni says there's no engineering process in Existence that can replace the philosophy of responsibility for getting things done right and getting things done efficiently and That's how they figured out how to make this happen. They didn't go for perfection on every test launch. They went for Everybody did their part. They had to instrument it because every time that one of those boosters crashed while they were developing it and they crashed a lot the Elon Musk was going to call within 24 hours and He expected to talk to the engineers that were responsible for the problem And they were going to tell him how it was never going to happen again So this concept of the responsible engineer Can put space rocks into science or into the into the space if it can do that It can probably do the stuff that you guys have to accomplish Efficient supply chains This is sort of part two of my talk and I know I've got to get going here because I have like only maybe 20 25 minutes left But what happened is that efficient supply chains? Just plain collapsed Was soon as the everybody got locked down Because there were dramatic demand shifts that couldn't be accommodated for example personal protective equipment There just wasn't enough Where it needed to be and there needed to be more made there were cleaning supplies I can't sanitize or insult that we couldn't get for the longest time food food shortages mostly in Well, what we saw interestingly enough was a persistent shortage of flour and the reason was because the flour was packaged in Commercial packages and sent to bakeries and when the bakeries closed they had all the flour They had plenty of mills, but they didn't have the Packaging supplies to put it into we put it in five or ten pound bags and sell it in grocery stores They didn't have either the equipment or the packaging to do that and there was a two three month strong shortage of flour in the US Other countries had different shortages, but it was a supply chain problem usually now infrastructure actually Didn't tend to if you had water and electricity you tended to still have water and electricity if you had internet It didn't tend to go away In fact our internet almost every vendor started supplying higher speeds and more of it fast So the question is why do supply chains break and why doesn't infrastructure break and there are really only two reasons Why a supply chain will break? first of all the demand changes faster than the time in the supply chain, so if you have Personal protective equipment and it's made in China And then it has to go on a container into a container and it makes its way to a port and then the sun a ship and Then another port and then it has to make its way to a distribution center and eventually get to a hospital That's a long time if demand change is faster than demand change like within a week Then the long supply chain that supply chain breaks and secondly If demand could be exceeds capacity at a link in the supply chain like for example We had plenty of flour, but we couldn't package it at that link the demand exceeded the capacity So those are the two reasons why a supply chain will break If we go back to the first one and look at it closely Durable good supply chains anything like refrigerators and appliances and even bicycles they're they're very slow and very unresponsive because make it the most efficient cheapest place in the world and Gradually sort of move it over to where it's needed and when you have to rapidly change a supply chain. It's too slow Perishable good supply chains actually are much faster and much more responsive and I'll talk about that in a minute and infrastructure Infrastructure does not tend to be designed for a single thing it tends to be designed for the long term not current demand So if you're gonna plant a tree in a park That tree is gonna be small for 10 15 20 years before it even starts to be big But that's okay. It's infrastructure and so you do it because you're thinking long ways out in the future when you plant the tree When you look at the second one Okay, demanding exceeding a capacity at a link in the supply chain resilient supply chains Have some sort of slack at every single step and if you don't you are not going to be able to respond to rapid change in demand and Another thing that you want to look at in a supply chain links is you do not want links to be tightly coupled So that when one thing changes you get a cascading failure across the whole network We see that oftentimes in power is interesting Lena Loose coupling limits what we have a habit of calling the blast radius of failures Those are the two things to think about when you look at the links of the supply chain with infrastructure That the it's designed for aggregated maximum demand So it doesn't actually care if the demand is coming from homes or offices the flower does but the infrastructure doesn't and so Therefore infrastructure tends to be much more resilient So let's going back to Roger Martin's book on the relentless on When more is not better and he says the relentless pursuit of efficiency has created very fragile supply chains And boy, could we see that so What we've been trying to do is optimize supply chain for efficiency at every single step Whether it's low wages at the creation step big batches in the transport step Long distances for distribution. It's very slow. There's no redundancy and strong dependencies and if we want to live in a world that's going to change fast and Unexpectedly then we've got to stop doing that. We have to start looking at Something more like the bizarre you might remember in the late 1990s Eric Freeman wrote the book the Cathedral and the Bazaar and the bizarre or the marketplace was and in his metaphor Cathedral was a strongly architected Development processes and the bizarre was the way open source was developed. Everybody had their own little shop and did their own little thing And he said he couldn't figure out how the bizarre could possibly work but it turns out it works a lot better for software and In fact, it works a lot better for a lot of supply chain For example, if you look at perishable goods The last thing that you wanted to do with milk or meat or fruit perishable fruits Most vegetables is optimized for resource efficiency or for high utilization of resources doesn't make any sense So if this woman with her bike and all her bundles of vegetables and fruits We're going to be really efficient She would take the stuff to a truck that was at a distribution center and then that truck would get driven to the market But that doesn't make any sense those things need to go and small batches straight to the market and she can get it there fast You never accumulate large batches when you're talking about perishable goods because small batches move faster And they don't sit in great big piles of inventory Well, you also Have to with perishable goods aggressively discover and eliminate any kind of bottleneck so we get a lot of our fresh fruit from South America and If the airplanes don't fly we don't get it and so you have to look is this a bottleneck? Is it redundant and then if it has to wait at the airport because there aren't enough trucks to distribute it then you got to handle that and With a perishable good supply chain you replenish what people buy and You use that to pull your replenishment and lastly you leverage customer feedback to make product supply decisions. What is she going to plant? She's going to plant with people buy and Maybe that's going to have a little bit of a lag time, but not that much so the thing that we always talk about for any kind of thing that's optimized for speed rather than for resource efficiency is Don't batch and queue Okay, so what is the batch and what is the queue? Well? Projects are batches not a good idea backlogs are cues Not a good idea either prioritization is just ways to Reshuffle the queue. That's kind of a waste of time. You don't need the queue in the first place Estimates there another way of figuring out which piece of the queue am I going to work on next? Every one of these things belongs to a push System they are about forecasting demand But in times of uncertainty push systems are going to be wrong because things change so fast So all of the forecasting in the world that people did in January certainly didn't help in March The thing you want to do in a uncertain environment even in ours is to pull from demand So what does this mean for you? Well, what it tends to the way to think about it is to establish a regular cadence of getting stuff done so You should know the rate at which you get stuff done. Let me give you an example. There is a Author who was interviewing Tom and me for an article. She was writing for a website. I think it was an employment website and she was going to interview us about Estimates and she wanted to know what estimates were all about and I said well, let me ask you something How many articles do you write in a month and she said eight? I Said oh you write eight articles in months. Would you ever accept 12? She said of course not I write eight a month and She doesn't plan out more than you know a few weeks or something like that So she wouldn't even imagine that it would make any sense to accept 12 a month You know your delivery rate of stuff that's done. You've got the data somewhere Don't accept the man at a rate higher than you deliver It's pretty simple Don't accept stuff and put it on some sort of backlog at a rate higher than you deliver and while you're at it You only need a short buffer of work to absorb variation between the demand and the delivery and if you have a short buffer and you only accept work at the At the rate that you complete it and and not at any higher rate You basically have a very rapid way to respond to pull from demand So that's the second section of what I want to talk about now. I'm going to go into the last one Because the last thing we saw was a recession or a depression or something like that, which we are still in the middle of Now we know about disruption in our world This is a picture that Tom took and it is on the cover of our first book and he took it in May of 2002 and that's important because this is Tom's first digital camera and by that time Time had been a photography for a photographer for a lot of years We have a whole bunch of scanned pictures of all the colored prints and slides that he's made but he took this picture on this on this trip because it was a little tiny thing that he could fit into the Hem into the into the bin on the bike and you know what it's three megapixels and He never went back to film If you look at Kodak versus Fujifilm This is a graph of the demand and on the left it starts in 1994 and the peak where the arrow is at the top is March 2003 that was the peak demand for color film and by 2012 it was down at the bottom sort of near the top of the bike here right here, okay and It just fell off a cliff just like that lots of things have done that but the thing to see is that Kodak didn't respond they kept trying to make what they were doing work better and within 10 years they had lost basically 50% of their revenue on the other hand Fujifilm In the same time period their revenue went up 60% almost How is that happened? Well as soon as they start seeing the demand fall out fell off They accepted it and said you know we got to do something different They did a little bit. They quit doing research on film. They supported a few of their most Important businesses for a while, but pretty much they tried to switch to something else and that's something else turned out to be pharmaceuticals because And beauty products because that's where their their assets and their Intelligent people could use the same skills that they've been using before towards a different product in 2012 Kodak declared bankruptcy and By that time Fuji film was doing just fine. In fact, they quit calling themselves Fuji film and called themselves Fuji And we have the same story here with Netflix, which you can see you probably have Heard of Netflix, but have you heard of blockbuster because blockbuster was a Storefront video rental place in the United States We had about 3,000 of them in to the bike 2004. They had six billion dollars of revenue coming in They were very successful you go in in the rented videos And they had late fees which really annoyed people because those late fees were kind pain and so then Netflix started up and they had a different scheme DVD by mail and it was flat fee three a month You return one you got no one you would never have late fees. Whoa So people loved it. They switched and then Netflix said, okay, no late fees We're gonna do DVD by mail to return it to star and you can get your new video faster and everybody thought that would be great But then blockbuster said, oh, how about streaming you can have a you can add two or three hours of streaming a month to your DVD subscription how about that it became popular and That the same part point net blockbuster was not able to get enough revenue so they returned the late fees and They then decided with down with streaming what they were going to do is put a He asked in each store where people could download movies bring it to home I bring it home on a chip put the chip in some sort of a player that they sold and as you can imagine that didn't go over as well as streaming and So they kept falling in revenue. They kept adding additional late fees and Annoying customers even more at the same time Netflix switch to a streaming only subscription And let's just say the rest is history So we know how to survive disruption and the first thing we have to do is stop doing what doesn't work We know we have to stop doing what doesn't work We have to stop annoying our customers and you'd be surprised at the kinds of things we do the airlines in the US charged all kinds of fees for Luggage and for changing flights and stuff They had to drop all those fees just to get customers to come back Well, we have one airline that never had any of those fees and it's doing quite well. Thank you You need to preserve cash in any kind of a downturn And you need to stop spinning it on the old stuff that you used to be doing and figure out how you're going to spend it Next you need to find some sort of unmet need And leverage your existing expertise and assets towards that need with a lot of short customer-focused feedback loops This is our motto at 3M that we always had try a lot of stuff and keep what's work keep what works So there's a story and there's these stories all over the world But I like this one of the pizza factory in Chicago that started making face masks instead They make about five thousand a day Can you imagine that of these face masks for medical workers? Because they found that the pieces of plexiglass could be treated just as if it were they were pizzas He didn't in the oven bent to shape And then they added a few accessories and they were able to package them up and so on and The world is full of people who've figured out how to repurpose assets and people Towards a different concept What I'm talking about is something that I'm going to call maneuverability Manoeuvrability these are impala. They are the fastest animals. I've ever seen, you know, they will hold this pose I'm amazed that Tom got the picture because that didn't last for more than a second And then there has we're up and they were looking for the lions and the leopards that wanted them for lunch so in their case maneuverability is survival and maneuverability I think in a times of great uncertainty is survival this is An f86 plane in the early 50s in 52 John Boyd was a saber pilot and he was supposed to get into fights with Soviet MiG-15s, which were faster They had heavier firepower. They had a narrower training radius, but you know what they did not actually Have the advantage you would think because by the end of the air-to-air combat the sabers these little slow planes Had a 10-to-1 ratio of success So boy thought this was really weird and he wanted to know what wasn't about that plane that made it so successful and He did a lot of studies and a lot of simulations And he decided that he discovered that the success is a byproduct of it's an ability to transition from one Maneuver to the next maneuver faster than the opponent and I would propose that that's what we need to be able to figure out how to do So here is his OODA loop observe orient decide act I'm going to call it a maneuverability loop and you have to observe what's going on Orientation I'm going to talk about in a minute and you have to decide based on the orientation What are you going to do next you call it a hypothesis and then you act and your action is a test Did it work or not and then you go through the loop again? So if we take a look at this central thing right here What we're doing in orientation is we have a mental model of how the world works and What you're doing with maneuverability is you're saying What's wrong with my mental model because it isn't working? How am I going to change my mental model of how the world works to match the way it works today? And you have to make you have to update your mental models and you have to do it fast So you have like genetic heritage like so the some polys are really fast You have cultural traditions if you look at air pilots, they have different kinds of traditions than I don't know army or something like that There's previous experiences if you've been through this before it helps a lot There's new information because this is different than before you put that all together in synthesis now If you want to think about this We are looking for a broad view of what the mental model could be changed to to try something new and I would propose that if you have a team a small team has to be small because otherwise it can't act fast and it'll have a broad view of the environment it can form some mental models and It has some diversity of opinions so that you don't keep doing what you used to do Lots of different minds not one expert and so when you have that in the orientation box You are much less like more likely to come up with a mental model That's going to work for the next thing for what's going on now if you have a small diverse responsible team So I think the thing we need to do is to figure out how we're going to solve for maneuverability and in our world, this is how you do that. This is a book by Jeff Jeff and Josh and it's called sense and respond and the thing to the important thing to know about those two guys Is they are designers? They design things and this is the way designers think about the world So they say what you need to do is maintain a continuous Two-way conversation with your market talk Figure out both directions try some see what they say You want small multi-discipline teams and I don't mean software teams I mean everybody involved in the invite the lighting customers You want a very clear goal? You want to know the outcomes? And you want to know what are the constraints if your constraint is the a deadline that can't move like three days That's a constraint if your constraint is there's only so much money. That's a constraint a constraint isn't a list of tasks It's the real thing that that without Hitting that you can't violate The team needs freedom to act asynchronously separately without permission from colleagues without permission from management By themselves And that is usually an architectural problem as much as anything in our world The team has to be responsible for building the right thing I'll just Remind you of my response philosophy of responsibility. That's what their responsibility is figuring out what's right and making it work There should be frequent delivery to customers so they can learn their way forward That's a quote from sense and respond learn your way forward Success has to be customer outcomes not how much output there is not proxies, but customer outcomes so the new normal Today Resilience isn't a luxury. We have discovered that resilience is pretty fundamental when the world changes so fast So what we need for that is three things one is responsibility Workers should solve customer problems without proxies and we need Maneuverability a freedom to act asynchronously and without permission if you look at AWS For example, one of the things that makes them so maneuverable is they're basically composed of Small teams that can act asynchronously without permission in their whole architecture In fact, the whole cloud architecture is designed around making that possible and Lastly, you should pull from demand Manage by rate Know your rate. That's the rate at which you can accept not just deliver Maintain a very short first-in-first-out queue and that's all you need So if you have those three things in place Then you're going to have a lot more resilience in your organization Then if you don't and so with that, I would like to say thank you and I think it's time for Questions, so what should I do nourish stop sharing? Great. Thanks. Maybe if you could just turn on your camera stop sharing. We'll go through a few questions here Okay, is my camera on? There we go Yeah, I can see you thank you. Okay, and Tom is going to join too And I'm going to make this a little bit bigger Great, there we go Perfect, you can see all the wonderful pictures Tom is taking the back right Having traveled with both of you a little bit. I thought Tom's destination to taking pictures. It was amazing Yes Now we've got we have a big TV here showing our pictures from all the many years and it's kind of reminds us of all the great places We've been to and it gets a really nice pictures from from India, too Yeah Very very informative very engaging Talk, I think also very relevant To what people are going. So thank you very much Mary for that There is a whole bunch of questions that people are asking so without much delay. I'll quickly jump And start taking questions So the first question I see here is from Rajiv So Rajiv is asking in your opinion. What could be one good software related trend? That has its birth in 2020 during this COVID time that you think will continue beyond 2020 Carolyn can I think it was one of the points Mary made here, which is that Software is a contributor to a team that solves problems It's not the solution All by itself. It's not that somebody tells People with software skills what to do. It's that they collaborate with other folks with other Capacities to solve emerging and rapidly evolving problems I Let's say one thing. I don't think it's going to be So we heard a lot about artificial intelligence was going to become really big in 2020 But what has happened? Is that the database is used to train artificial intelligence? Have suddenly not become relevant anymore because everything's changed. So a huge amount of the ai training that's that had going on has become suspect if not irrelevant. So um So I don't think that's it But I do think that we need to figure out how to respect The people that are solving these problems The ones that are on the front lines making stuff happen and give them Give them the the respect of Telling them here's what a solution looks like and let's make it happen. I think that We've seen a trend this way and I think it's a revert a reversion to where software used to be back when I first started coding Um And I'd like to see more of it. I think instead of let's do the tasks some other people line up We have to start expecting software engineers to be part of a problem solving team And and when things have to be really fast there just is no time For any of the other processes and any of the other gatherings or anything else You just have to get everybody together and get them in the room and say you've got three days make this happen And I think that that'll continue more more than continue Awesome All right. So moving on. Uh, thank you very much for that. Uh, we there's a question from venea Where she's asking about what are your thoughts on kanban? Seems to check many of the boxes you mentioned Could you read the first on kanban? So, um kanban is a term that came from Uh, japan and was used in lean manufacturing back when I did lean manufacturing um, and kanban was a mechanism to Pull from demand and we set it up and used it in our plants And if you think about the the software version of kanban, it is a mechanism to track pulling from demand and making sure that You use some sort of a pull mechanism Kanban actually means card and in our manufacturing plant. We actually use those cards to pull So in my background, which predates software kanban by a bit I've always thought of kanban as a tool for pulling from demand Now software kanban Is a lot of other things it's It's idea of minimizing work in process Isn't the is is a fundamental important piece of it But I think this whole concept of pulling from demand Is something that whether you call it kanban or something else It's it's it's a fundamental aspect of being resilient of being able to respond fast to events the way that they unfold and um Many of the tools in kanban are interesting tools for figuring out how to pull from demand um other pieces of kanban As you sort of expand it and talk about The culture that surrounds it wouldn't necessarily be any different than other agile things. I don't think but i'm not sure Okay, thank you. Uh moving to the next question, uh from shalini Amazing and thank you so much for so much the need of the art Uh, how do you build this culture? sustainably in an organization that starts from the top of the organization the philosophy of what people think are important um So if you have an organization, it's basically successful And they've been doing things in a certain way for a lot of years You can't just go in and say change what's been successful Where I see sustainable change in culture is when there are shockwaves that come through Either from a stream event like we've just had or from very surprising competition like for example kodak and fuji film faced and uh You know digital competition and suddenly what you were doing before it doesn't work and the senior management has to think differently the only Real successes that I've seen in totally changing the whole company culture have come when The the heads of the company get their head together and communicate it down through the ranks We're going to have a just a different way of doing things a different philosophy It doesn't just mean the top management thinks it's good Means it's not management is able to change the way the structure of the company is and what's rewarded And what's thought of as good to make it happen And as martin faller always like to say you have to either change your company or change your company And no, okay Perfect, uh, we'll take one last question here. Thank you. Uh, this is uh, this is from uh all right here You mentioned about cues my question is during the uh, during the Unpredictable sorry, uh unpredictable disruptive times What could be some efficient ways we can manage cues or is there a better way? Than managing cues for better outcomes. You just stop having cues Delete from get rid of them That's the most efficient thing and in fact Any time that you're working on cues backlogs or anything like that you're just wasting time I consider cues a waste um, we were talking to once upon a time a healthcare company in seattle that Um was just so resistant to what we were talking about and we discovered That in the vicinity of 20 to 25 of the people and they were all management people in a room of about 20 Spent full time managing cues And they of course couldn't really accept what we were saying because their job was going to go away But you don't need to set priorities. You do not need to have backups and they're just plain bad ideas It is not necessary to have cues and if you have cues You have to just figure out how to get rid of them There's no good way to to to take something that you should not have and say how do I make it better? It's like get rid of it A cue is essentially a push mechanism You decide ahead of time what you're going to do and then you commit to doing it and if you're learning Indicates that there's something else that you should be doing instead. Well, you've already committed um Push mechanisms are very inefficient They are very expensive They are not very effective So instead of pushing with the cue you pull with Bulls You have some goals You do whatever it takes in order to achieve the next most important thing um Okay Is it fair to say that if you know, you may have uh, you may have a list of things that you've got from different places And you want a place to to just capture them as a as a log or as a list But then you start pulling based on goals from it. Is that that's still considered as a cue? Remember you can only complete work at the rate at which you can complete work And if you're accepting work at a higher rate, you're saying where can I park it? Well, that's an interesting concept, but you're accepting work At a higher rate than you can accomplish it. It's much better to say I've got three things now next to never. Okay This is what i'm doing now This is what i'm going to do next and everything else goes into my never list Now you can have a never list Every swathware organization. I know of almost with 95 percent have more work and they can possibly handle So they will always have stuff coming in that has to go into this never list or on the floor one or the other um, I was working with the very senior executive from a long long company here that was quite successful after that in the u.s and and He got this concept and then we went and visited He was in one of our classes really early on and a few years later We gave a talk in his company which by that time was quite big and he said, you know The best thing you ever told me Mary was have a never list Because if you put stuff on your never list, you know what I will bet you never look at it again Because what's urgent now is what you're going to be paying attention to and if you're refreshing what you think about doing from the most current problems instead of stuff that's five and ten and you know seven months or weeks old you are actually doing a better job of choosing Than stacking this stuff up. It is better to say to the people who are sending demand at you Look, here's my capacity I'm going to tell you when you ask me something if I have capacity to put it on my list and if I don't have capacity I'm going to tell you right away and that gives that Demand generator a way better at lots of options than maybe hoping that someday you're going to get around to their thing By the time that you do it will be obsolete But hey at least they got it on your list That doesn't that doesn't do your demand generator is any good your customer is any good And it doesn't do you any good to have this so if you gotta have a list put it on a never list And I'm going to bet you you'll never look at But it'll be there just in case you're thinking you should watch it And hopefully if something's really important it will come back again And you you know you can deal with it when it comes back And if it's really important it'll come back But if it comes back two and three times it means you basically don't have the capacity to deal with it Because you're using your capacity to something that you you have decided time and time again is more important so That means you either have to increase your capacity or it has to go somewhere else if it's really important But it is so much better to be honest about that than to pretend that maybe someday you are going to be able to do it And it could be reframed in a way that makes it easier to deal with or faster to deal with or Something that really is more important than the the competing idea from some other source of demand but I don't think putting stuff on a list to hopes that maybe you'll By thinking about it harder get around to doing it I don't think it does anybody any good not the people that have the request Not the team that has to work on it not anybody. It's just a bad idea All right Thank you so much one last question. I know we're running a little late, but I think I love this question. This is from Sakshi She's asking You talked about providing freedom to the team sometimes freedom turns to chaos Especially in in meetings any suggestion how to limit these situations Okay, so remember I talked to also about the philosophy of responsibility all right so The team knows what it needs to achieve And it needs to know when it needs to achieve that It needs to know the constraints around the achievement and it needs to know it's peace in the bigger system all right And in my experience if I look at the companies that are really successful Most of these small teams have responsible leaders In the chief engineer model There's a chief engineer that is the responsible leader that owns Responsibility for delivering on that thing that the team is responsible for And at some point in time that leader has to say okay, we've had enough discussion We're heading this direction And we we have to do this here's our goal. So let's say you are trying to do a Curbside pickup software and you had to have it done in a week There's one person who's leading a team of seven and that person is responsible for having it And there's some arguments about which way is best those are good because it gives a lot of ideas And at some point in time The leader says okay, we've spent three hours on this. That's all the time we have we're going to make a choice And move on and we're going to go with that choice and try it out because it's all the time we have So there has to be a way to Move forward and not argue if you take a look at open source There is always the the what the nevel and dictator That does that occasionally not very often but occasionally to keep things moving on And in a chief engineer model, there is a responsible engineer Who owns responsibility and passes it to the team? To have a team that has nobody that has the capability of saying We're going to head next isn't an engineering model. Okay the team has freedom about how to pursue Solutions to the problem it's trying to address but the freedom Comes with constraints schedule constraints cost constraints Um access to skills and resource constraints, but within those constraints You have freedom to choose how to proceed um the control really comes from the constraints There's a really interesting book called think like amazon by I forget you have to look it up He was uh, he was ahead of amazon when amazon put up its its merchant website And they had an interesting way of doing it. They said you have a small team And the first thing that this guy did when he joined and took this job on was to negotiate his What was he was responsible for and his responsibility? He said goes like this um A merchant will be able in the middle of the night To log on to our platform And by morning will be able to have a sale That to the customer will look just exactly like it was an amazon sale That was the definition of success And that definition was what the team worked for Until they were done. That's how they if they met that goal If they were able to put something up that that met the goal then they got to proceed They had a limited amount of time. They had a customer focused goal that was pretty clear And um, and he led a team and he was responsible for that team that made that happen And when you have a customer focused goal almost every aws service works like that They have a customer focused goal that they can say that defines success They have a leader. They have 10 12 people. They have constraints And within that window, they have to figure out how to make the service work And if you look at SpaceX, it'll be the same thing You have a launch a test launch. It's in three months And the purpose of this test launch is to To test how the landing legs connect with the ship out in the wavy water and just see if they can lock in Okay, and you got three months and You know May 31st the launch will happen and you know what it's not going to change. It's going to be may 31st It's going to go up and it's going to try to land in the ocean and if your legs aren't ready to lock in well too bad Um because it's going to be tested and you You have to be ready And so you have a team that has that goal They know what they have to do. They know where the legs fit in They have to work with the crew on the ship to make sure they know how the locking mechanism works And you can have arguments about the best way to do it But may 30th isn't going to move And you can only have arguments up to a certain amount of time before you have to start trying stuff And you have to have two or three different Different strategies because when may 30th comes one of them has to be in place. I'm working and when you have that You know the constraints, you know the deadline, you know when things are going to be tested And you have a a leader typically with SpaceX So there'll be an engineer who is responsible to make sure that these five people get that landing like figured out Communicate with the boat people and how it's going to lock in and then test and see how it works and that test will happen and it will work or it won't and The engineering team and the chief engineer are responsible for whatever happens And in that world, you know, you can absolutely have to have arguments. It's good to have arguments You also have to move forward and you have to have everybody responsible for making sure that they're going to be ready for that next deadline People are loving that All right, I think we've uh taken a lot of your time Thanks, uh, thanks a lot. Uh, Mary and Tom, uh, it was such an honor to have you back And uh, such an amazing topic. So, uh, thank you very much for that Thank you