How to not sabotage your transformation

Good morning and welcome to re:Invent 2023. It's an honor and a pleasure to kick off this year's re:Invent, even if you do so, probably with 50 other sessions going on at the same time.

I'm Matthias Patzak, an Enterprise Strategist with AWS based out of Munich Germany and I apologize for my horrible German accent and any pronunciation, grammar issues during the talk.

So I said I'm an Enterprise Strategist. No one really knows what an Enterprise Strategist is. My mother, probably thinks I installed printers at Amazon. But in fact, I'm part of a team of 12 former senior leaders from commercial public organizations with transformations as AWS customers.

So I was CTO of Auto1, the leading European marketplace for buying and selling cars online. And in this capacity, we share our own experience on how to transform organizations from a people, process and technology perspective with current customers. We also share best practices from the community and we share how Amazon and AWS innovates at scale.

And so we share with customers and customers share with us, they share with us how they plan to run a transformation and how to set up a new operational model. And sometimes we see a bit of weird things like I brought an example here. Point number four says, "Bring up limit issues as frequently as possible or when possibly refer all matters to committees." It's weird, isn't it?

And actually this is not an operational manual. This is a guide on how to sabotage German, Italian and Japanese organizations during World War Two. But we wouldn't sabotage our own organizations, wouldn't we? Sometimes when I talk to customers, I have the impressions that someone in this organization is sabotaging the transformation.

In this talk, I will talk about patterns that we see and how to overcome them. And it starts with that many organizations are applying outdated principles. These are Henry Fayol and Lyndall. And Henry Fayol was a French mining engineer and was a management consultant. Yes, we had management consultants 100 years ago and they were the thinkers of the day.

So they wrote books and they wrote articles on how to set up organizations and how to run businesses and the books and the articles were based on the assumption and perception that workers are lazy, not really trustworthy and uneducated, and probably they were true with uneducated.

And so this is some of the advice that they provided. The first one is on specialization - divide work to encourage specialization. Or the second one authority - the right to give orders and the power to require obedience. Actually, some of their advice was good and it's still valid. For example, the first one on direction - the entire organization should be aligned and be moving toward a common goal. Or the second one on initiative - encouraging staff to show initiative is a source of strength in an organization. At Amazon, we would call this principle bias for action, one of our leadership principles.

Unfortunately, management paradigms from high volume, low error manufacturing has come to dominate business organizational practices. And this is a pity because for example, in Germany manufacturing sector is just 27% of the GDP. In France, it's 17% and then in the US it is 18%. And even manufacturing organizations do not apply these principles which fit very well for some of their processes in all of their areas.

But many other industries and many other organizations also apply these outdated principles. And as a consequence, this is out of an essential survey - 94% of CEOs said the operating model, so processes and culture, it's not about the people, it's processes and culture, put the growth and performance at risk.

So the majority of senior leaders believe their own organization is a limiting factor to growth and success and it should be the other way around - your people, your process, your culture should be an enabling factor for success.

I attended business school, a German business school and I was taught these outdated management principles. I made my career in startups and SMBs. So I got used to a different way of working, but I was taught these outdated management principles and as leaders in organizations, if you want to change this, we have to start with ourselves and we need to unlearn the management practices that we learned in university and that grew our career and we need to learn new ways of thinking and new ways to set up a culture.

And then we need to walk the talk and help our organizations to learn and unlearn. But we need to be very, very clear on what to keep as well - which part of our culture, which parts of our products and which parts of our processes.

The second thing how organizations sabotage their own transformation is siloed organizations. Many organizations are structured by function - a design department, a marketing department, a manufacturing department - unfortunately, with a hierarchy on top of it. And when these departments want to collaborate, coordinate and communicate very often, they have to ask for approval and these approval processes, they take time. And what does the worker closest to the customers do while they wait? Do they wait? Actually not, they start more work.

So many organizations have installed a rigid structure for control and actually what we want to have, we want to have value and flow across the value chain and strong collaboration across it. And this is why Amazon introduced the concept of two pizza teams - small cross functional teams which include all necessary functions and roles that are needed to deliver value from idea to impact.

So one example on communication, so let's assume a person of the marketing department needs support and help from a person out of the ecommerce organization. And in any organization, you would believe you would just ask, "Can you help me?" and the answer would be "Sure." Unfortunately, many organizations and you, you can observe this with the number of people that are on CC or are on your emails in many organizations. What then starts? It's not communication between two persons. Many people are involved, other integrated contributors or more managers, team leaders.

And in this example, it's eight people involved and eight people have to maintain 28 connections, 28 communication channels and 28 relationships. If you would involve 20 people, it would be 190 communication channels. If it would be 16 people, 122. And as a consequence, when many people are involved and decisions are made by a committee, it's unclear who is in charge and who is the actual owner of the decision. And very often the actual decision is the least common denominator with the purpose to make everyone happy. But actually, it makes everyone unhappy.

At Amazon, we act differently. So Jeff Bezos said "Communication is a sign of dysfunction. It means people aren't working together in a close, organic way." And we at Amazon, we should be trying to figure out ways for teams to communicate less with each other. Not more.

So does Jeff Bezos provide and suggest silos actually? Yes, but a different way of silos. I started this part of the talk with siloed by function. At Amazon, we're trying to set up loosely coupled teams - cross functional silos - as I said, two pizza teams with every role that is needed to deliver business value from idea to impact. And this is why we can be fast at scale.

Sometimes organizations also sabotage the transformation because of wrong resource allocation and they allocate resources in the wrong areas of the organization because it's unclear what is the actual purpose, mission, objective and how to achieve certain objectives.

And this reminds me of Alice in Wonderland who asked the cat, "What route do I take?" And what does the cat answer? "Well, where are you going?" And Alice says "Yeah, I don't know." And then if you don't know where you're going, every way is good or every way is bad.

And actually, for organizations, there are ways on how to visualize this with a map. And this would have helped Alice to find the right path, having a map to visualize their value chain and their operational processes.

And this is an adoption of a world map on the horizontal dimension. It has predictability - low predictability and high predictability in your organization. And it shows autonomy that you need to provide, to accomplish this task.

And so I came up with an example by one of my colleagues, the former CEO of McDonald's Philip Brown. And he said, when it comes up to cooking fries, fries in Mumbai, in Munich and in Montreal, they need to look the same. So you need high predictability. So you provide less autonomy to the people producing these fries.

When building a house or implementing an ERP system, it looks a bit different. So the results and the ways to achieve these results are less predictable. So you provide more autonomy to the people. And it's even truer when creating a loyalty program or reinventing retail - reinventing retail is a huge challenge, very unpredictable. So you need to provide a lot of autonomy and freedom and psychological safety to the people trying to reinvent retail.

And you can cluster and structure this map in four areas. And the first one is Genesis - this is the area where innovation happens, where research and development happens with some curious minds in your organizations try to find new solutions for the problems of your customers.

The next area is the Custom Built area - the small and agile teams build custom built solutions to put these innovations into production. And then after some time when some organizations build custom built solutions for the same problem, someone builds a product for it. And over time, some of the products become a Utility.

I'll give an example - so computing in the 1950s, computers were handcrafted, but then it got mass product with IBM. And then Amazon made it a utility with Amazon EC2. And this is not just true for hardware, it's true as well for software. So I'm a computer scientist and applied SQL databases as products a lot. But Amazon made it a utility with Amazon Aurora.

And using more products and utilities enables you to put more of your resources and more of your efforts in your areas of true competitive advantage - in the areas of innovation and custom built products. And this is why many of you are probably here - to learn about the services, the utility services and the products that Amazon provides.

But using utility services is not as easy as it looks. And this is not from a technical perspective, but this is from a psychological perspective and many organizations have inertia, some resistance in the organization, especially when it comes to moving to the cloud.

So we said we have servers and our servers are now a utility. And the same is true for databases. And there is inertia, inertia that sabotages and slows down your organization. There is a resistance by the people running and maintaining these products. And this is very understandable because what we ask of these people - and it feels for them like we would press a reset button on their knowledge - we ask them not longer to use products that they know very well and instead use a utility.

So we let part of their knowledge go. At the same time, what we also ask them by using the utility services is to create and build up skills and reskill themselves to being able to act in the area of custom built solutions. Because what we want is infrastructure as code.

We ask when moving to the cloud people to use utilities but applied as custom built solutions.

And now these administrators, these operations people, they compete from their perspective with a 22 year old graduate and as leaders, we need to provide the psychological safety to our staff and explain that we might reset some of the knowledge but not all of the knowledge and none of the experiences. So you can overcome this resistance that slows down some transformations in some organizations.

Next topic is forgotten, differentiators. What makes you you and again, let's have a look at worldly maps and this is the original worldly map. So we had on the horizontal axis, the predictability and we have the value chain, the visibility on the vertical axis. So sales on top of this map are very visible for customers. Things at the bottom are not so visible. And this is an example of paying bills in the 19 nineties. So personal banking customers who wants to pay a bill, went to a branch or paid by phone and so the bank created an infrastructure with some custom built products like call center and with products like a mainframe and some utilities like low speed.

One, 20 years later, the map looks different. The customer got more ways to pay a bill by app, by branch, by web or by phone. But we also see that the custom built call center now became a product, a great move. And what we also see that this bank is using a lot of products and just very few utilities. And this is what we observe in, in the banking industry. But with all other industries as well, the traditional organizations work differently and apply this map in a different way that that cloud native and drill native challengers do.

So. Additional native bank, they leverage a lot of utility service, aws services, they use a few products, but they put their efforts in the area of true competitive advantage with a back payment logic or a front end application logic. And their intent is to maximize the work you don't too. So they leverage the cloud, they acquire some software. But they are very, very aware that when acquiring software, there's danger in it because over customized software as a service or commercial software can become an issue and is in many organizations. So what they do, they would rather adapt the business processes to the software than adopt the software because from their perspective and in their culture for them, it's easier and faster to change a business process than a software. And they build innovation, custom built solutions.

And from an investment perspective, traditional organizations put 55 up to 80% of their investments in the area of products and utility. Actually, it should be the other way around the largest part of your investment or your resources and your budget should be in the area of innovative custom built software. The addiction to over efficiency sabotages many transformations as well. The main question here is, are you optimizing for speed or for efficiency? Actually, you need to do both. It depends on where you are.

Remember, alice in wonderland, in the areas of the utilities, the products and the utilities, your main kpi is cost and there you can apply six sigma thinking there. You want to reduce variation and you want to reduce the cost and the suspend because then you can reinvest this money and then resources in the area where you want to compete and there your main kp is speed, speed to deliver an innovation and put it in front of your customers there. You want to reduce the cost of change and learning must be easy and frictionless, simple, efficient and stress free to change and build a solution.

And actually you can apply four ways of working in your organization in the area of genesis and innovation. Think about the pioneers in the wild west. So the cowboys on their horses riding out in this unexplored territory and trying to find places where other persons could make a living. If they found such a place the settlers would follow and they would set up the first houses and make a living there when they proved, and this is in our industry, proved a product and market fit. When they prove that you can make a living there, then would start building a town, building a fire station, building a police station, building a school. And later on like what happened with las vegas and the city planners come and you can apply from a sourcing perspective, not just different ways of working but also different sourcing models. In the area of product and utilities. You can leverage more outsourced and more procured ways of working with external vendors there. It's just you want to reduce variation and there it's also ok working in a non agile way, it's ok to have upfront specification and hand it over to them. But in the in the area of innovation, custom built solution where you want to reduce the cost of learning experimentation you should use in house two pizza teams probably with embedded external experts who have done this type of technology this time of innovation before some owning organization managers are just managers and they fail to lead.

Today, leaders run the engine from a budget perspective. If we already saw most of the budget is spent on just running the engine in the areas of utilities and of products. And just a slow part is spent on innovation. And the leaders, they work heavily in the system. They are heavily involved in day to day operation and they spent just a small part in the area of talent and strategy.

Tomorrow, modern managers, they shape the future. They spent some of the budget in running the engine but spent most of the budget in the area of innovation. They work a bit in the system, but mostly they work on the system, optimizing the system. They spend a lot of their time in attracting hiring, developing and retaining talent. And they spent time in co creating a strategy and driving its execution. And this is the most often issue that i see with strategy coming up with a strategy presentation is very easy but driving the execution, this is the hard part.

And the question is, is your leadership team a team of leaders or as i would say, a group of leaders or is it a leadership team? So we had this team that went out on the off side with the goal, we want to be more customer centric and then they come back and they translate this objective into the organizations. And the one leader says we need to cut cost. The next says we need to build a product. The next says we need to reduce the number of products we sell. And the last one says we need to open most law. They are not aligned on the objective and they are not aligned on their communication. And leadership teams work differently. They apply what we learned for the last 20 years with agile teams in software development. Even as a leadership team, they are co located, they share common objectives and they have a closely aligned communication strategy.

So what does leaders do first, they define a mission and the broader vision with me and i'm going to read it out loud and it might fit for your organization as well. We will become a data enabled organization leveraging our cloud synergies to achieve world class customer centricity, unmatched by our omni channel. Competitors might be a fit for amazon. Actually, the amazon amazon mission is this to be earth's most customer centric company. It's a totally different way to set up mission for an organization.

What leads us also to they codify culture with leadership principles in organizations. You should make pretty clear what is the culture you expect? What is the behavior you expect? And they brought and we believe the principles are the organizational glue. What really makes a difference? And i brought some examples with me. The first one is we should have integrity when we work with customers and prospects openly honestly and sincerely, this is from enron who remembers en these organizations who went out of business 2001 because of fraud. This was their leadership principles. Or we should value our employees. We expect to be customer preferred choice. And this is what i really love. We operate within the letter of the law in these leadership principles. They remind me on 10 things you should have learned in kindergarten. Like share everything. Don't hit people say sorry when you hurt somebody, put things back where you found them, play fair, clean up your own mess, wash your hands before you eat. Don't take things that aren't you. If you have to codify things, you should have learned in kindergarten. In your leadership principles, you have a serious issues and you should immediately think about your hiring process, your promotion process and the process where you beside whom you let go.

I brought other examples with me. This is out of my own transformation. When we at out of scout decided to move and transform our organization from a monolith on oracle, operated on in an on premise data center to a microservice architecture and organization cloud native in abs with scala and the monolith, we used microservice technology. So this was a huge change and we try to provide some guidance for our staff. And these principles were facilitated and created by my chief i tech christian die. And we had, for example, this one heavily inspired by amazon. We made this transition in 2014, you build it, you run it. The team is responsible for shaping, building, running and maintaining its products fast feedback from life and customers help us to continuously improve.

We also ask our people to be bold. Please be bold. Get into production, early value monitoring over tests fail fast. Please dear engineers please fail fast but recover and learn. Optimize for meantime to recover, not meantime between failure. But we balance this and we had more leadership principles. We balance this. For example, with this one security compliance and data privacy, be with least privilege and data privacy in mind, know your threat model. So be aware of your risk, fail fast but be aware of your risk. And when you fail, limit your blast radius, we also had the aws first principle. So for us, it was pretty clear we want to be cloud native and aws was the preferred supplier.

What does leaders do as well? They provide gura and mechanism and this leads to the next thing how organizations sabotage their own organization, bureaucracy. We believe that and it's easier said than done that organizations should not be complex. But when organizations grow, having more products, more areas, regions and countries where you sell these products, more roles, more teams, you need more processes, more bureaucracy, modern indi native organizations don't need more bureaucracy. They act in a different way what they do. They provide clear objectives and a clear purpose. They hire great talent, provide very, very crisp and clear culture and then they just get out of the way. And for example, you can read it in the netflix document of culture. You find the latest version on the netflix job page. You will also find the netflix document of culture on slideshare the first version of 2009

And they explain very well the mechanics they apply while having increased complexity in the organization. They try to stay flat on the bureaucracy and think about your last week and how much of your time you spent on activities, not adding value to your customers, or the times you were waiting for colleagues or peers or managers to give you an answer and support your work.

Lack of ownership is also an issue. Many organizations, there is not a clear accountability. At Amazon, we have this concept of a single threaded owner and the illustrated owner owns all resources for initiatives and they are also responsible in the long term for this product and this service. And this is differently how many organizations work with the concept of a product manager, a product manager. He facilitates moderates and coordinates resources, but this person does not own the resources and they don't have long term ownership for the product.

A project has a defined a start date and a defined end state and at the end state, they hand over the result of the project, whatever it is, hopefully on time on budget and on quality to another organization. Both roles, remove barriers and dependencies and both roles motivate against a common outcome.

What we also see that organizations hiring just for skills and not for attitude as you actually should do because we believe that leaders need to raise the bar in the organization with every new hire. So we say hire for attitude hire people that are a great fit for your culture and have shown the behavior that you expect them to show in the future, that they have shown it in the past and skills you can just learn. So higher learners and why higher learners, 65% of children in preschool today will work in jobs that don't exist yet. And just 10% of classroom learning is retained after three months, but 100% of classroom learning that is immediately applied learning on the job is retained after three months.

And when it comes to recruitment, there are different desires between organization and the new joiner. So usually an organization wants to pay, everyone says we pay on market, but usually they try to pay below market. They provide modest bonus and some equity. They require specific skills. The new joiner should be a fit for the manager, should take direction and should be flexible based on task.

But what does talent require? What are the desires of talents? They want to have a salary on top of the market? They would love to get some equity but they, they want to have ability to learn. They expect a great tax take and tooling and actually this was the reason why my organization auto scout went in the cloud. We got a new ceo and he, yeah, he pushed me through a three minute it strategy process. In the very first meeting, he asked me two questions. The very first question was matthias. He was an australian matthias. Do you attract talent? And my answer was yes. 2013, we were releasing 12 times a day on our on premise data center on an oracle database. I thought we were great.

And then he asked me the second question, matthias, how does a 21st century text de look like oracle microsoft technology on premise data center. And then i realized we hired good people, but we were not really able to hire great people from our industry. When we looked up the top 100 european websites, just two of them had the same technology stack as as we did. And so we changed for this very strategic question. Do you attract talent, our whole technology stack and the way of working?

But what employer also require are they want to feel ownership and they want to be aspired by a purpose? And by outcome, this is edwin cockcroft, a former cto of netflix and former vp at amazon. And in one of his talks, he described how netflix operates and innovates its scale. And then in a q and a session, he got the feedback from, from ac t for the fortune 100. And this ct said we can't copy netflix because it has all the superstars engineers and we don't have the people. What is adrian answer? Pay them more?

Actually, what he answered was we hire them from you and then we get out of the way. It's not, as i said, with 94% of co believe the operational model is a limiting factor for growth and success. It's the operating model, processes and culture. It's not most of the time, it's not the talent, it's not the people. And this is how netflix grows and this leads to motivation or de motivation as a factor on how to sabotage the transformation.

We believe that employees are retained by being engaged. And when is peak motivation of an employee in a smaller forum, i got the answer on the weekend on payday or when they leave the company, i believe. And this is one of the valid series about a motivation. Employees are on the peak of the motivation when they sign the contract or on the very first day. And from the very first day on the motivation starts different workplace. The team is not as ready as, as promised. The project is not there what was promised. And this again, if you want to motivate people, try not to demotivate them, just get out of the way, hire them. Set a great purpose, put in the right culture in your organization and then get out of the way when it comes to retention, they again are different desires between organizations and employees organization.

They want employees who stay in their role. They should take direction. Please don't ask for a raise. They should be comfortable with a performance monologue, the manager and i've been a manager for 20 years telling you how he perceives you and please reskill in your own time on the weekends. Employees desire is differently. They want to have clarity on their career path. They want to have autonomy to work on great things. They expect to be proactively re leveled on the market. They want to have fairness of evaluation and they want to have continuous learning environment and to motivate people. They need autonomy, mastery and purpose. And it's on us, on us leaders and everyone is a leader to help to improve, to move the organization in a more loosely coupled set up to work on mastery, provide mastery and also require mastery. This is the leadership principle, amazon that we call, insist on the high standard and create a great purpose.

And sometimes in my organization, i i felt it was a bit hard to make the purpose tangible for my engineers. And what we then did, we started to get out of the buildings, we went out to car dealers and we went out to people, private persons that needed a new car because they had an accident. So there was some urgency or they just needed a larger car because the women were pregnant and they were expecting a baby in six months. And this made the purpose of our organization very tangible for all of us.

At amazon, we provide a clear career path for managers and leaders, but as individual contributors as well. And at amazon, you can as a technically individual contributor without having direct reports become a senior vp and distinguished engineer who has the same rights duties, salaries, bonus equities like you have as an svp, what we see when organizations are more advanced in the transformation. They're working in a more agile ways of working that they still apply traditional ict suppliers and very often this is not the best fit for the operation model.

So enterprises desires when they act in the area of innovation and custom built solutions, they want to achieve outcomes, not just deliver features. They expect the people from the system integrator to be co located agile and experimental. They want to reduce costs through automation. So they want to set up infrastructure as a code in the aws cloud. They expect from the system integrator, highly skilled mixed tax experts that they can inject in their agile. Two pizza teams, people who have done whatever they are doing right now, done before and they would love to pay for results and share some risks.

Traditional system integrators, they have a different desire, they want to generate revenue. They expect predictable contractual and documented requirements. They want to preserve the margins. They provide you mix skills because they have talent, but they have to provide the talent for all of their customers and they want to be paid for time and they really don't want to share your risks.

We're coming to an end. So the last slide on your journey don't photocopy, someone's else organization who knows the spotify model. The spotify model actually was published in 2014 by hendrik ng. I read it by myself, got very, very inspired by it. And it's a great, a great document. Actually, it's a presentation. This comic style and visualizations how spotify operated or parts of spotify is operated in 2014. Spotify is not longer waking this work working this way, but still many consultants sell exactly the solutions to their customers. And the spotify model is great for getting inspiration but you should not copy someone's else operation model.

What what we recommend is that you experiment as a leadership priority and don't over plan or over estimate the pace of change. But be stubborn on the vision. You need to have an objective like alice in wonderland should have and then be flexible on the execution. Don't just rebrand roles and processes. I can't tell you how many deos transformations i've seen where nothing changed. Despite of the title, this is not how you, how you make progress on your journey and please don't ignore resistance but double down on change management and what it means, provide psychological safety.

Most of the changes are hard for the people, as i said, for them, it feels you're resetting their knowledge, their experience and they compete in an unfair game. So i'm 48. I have two kids. My time is very limited that i could invest on learning whatever i should learn on the weekends by my son. He's 18. It feels like for them, the day has 28 hours, he just sleeps, two of them, whatever he wants to learn, he would be much faster than i am. So if you transform your organization, provide psychological safety in your organization and don't delegate change, you can get agile coaches, change managers from external partners in your organization and they can support you. But this is your company, this is your culture, this is your change, engage in this change and delegate other boring bureaucracy processes and stuff to externals.

So, thank you very much. I have two asks. First one is amazon is a data driven organization. So please provide feedback on the events app and I would love to continue to engage with you after this talk, right or left at this stage or continue conversation on linkedin. Thank you very much. Yeah.