Author of Blitzscaling
Chris Yeh is the co-author, along with Reid Hoffman, of Blitzscaling, the book that explains how to build world-changing companies like Amazon, Alibaba, and Airbnb in record time. As writer, investor, and entrepreneur, Chris has had a ringside seat in the world of startups and scaleups since 1995. His books help founders, venture capitalists, corporate leaders, policymakers, and everyday people better understand how the internet has changed the way we work together to build amazing organizations.
Read Chris' inspiration while doing research for Blitzscaling
Why is good organizational design so important while growing? One of the things that you rapidly discover when you’re scaling up a company is that the nature of the company is going to change. A 100 person company is not simply a bigger version of a ten-person company. If you go ahead and solely rely on momentum inertia, you're going to get it wrong. You need to be intentional about the organizational design. Growing is going to require you to change the design.
What is the most common mistake in designing organizations? Simply not realizing that the process of designing the organization is not static. Designing an organization is something that you do on an iterative basis; you design and redesign. In the book Blitzscaling, we talk about the different stages of the organization based on the orders of magnitude of the number of people that you have. As your company grows, you go through the stages that we named: family, tribe, village, city, and nation stages, and the rules change. You don’t need to wait for things to fall apart; you can anticipate change and make changes proactively. The other element is monitoring the key metrics, such as revenues per person or employee satisfaction, by observing these on an ongoing basis. From the results, you can tell that something has changed, and you can work on the intelligent redesign.
Designing an organization is something that you do on an iterative basis; you design and redesign.
Be an infinite learner.
What characteristics should a founder have to be able to scale a company? Be an infinite learner. You’re playing a different game at every stage of growth. In each stage, you need the ability to learn how the game has changed, learn the new set of rules and strategies for the new game, and then implement the necessary changes. People who are infinite learners look forward to the change, to learning new rules, and to the challenge of adapting and modifying what they do.
When should a scaleup founder start thinking about the organizational design? They should think about it from the start. Many things about your organizational design are not easy to fix later on when you don’t get it right upfront. As an example, one of the issues that people talk a lot about these days is Diversity & Inclusion (D&I); it’s much easier to build in D&I when you have five people than when you have 500 people who look and talk the same. The key stage for rethinking everything is the transition between the tribe (stage 2) and village (stage 3). A company with up to 99 people is a tribe, and a village is 100 people and up. It’s not a binary swift; the transition depends on your company dynamics. It’s the crucial transition from the informal and unstructured to the formal and structured stage. The early employees especially can have a hard time making this shift.
It’s much easier to build in D&I when you have five people than when you have 500
Who will your people turn to if they need help making sense of what’s going on?
What are the steps in discovering an organizational model that suits a company? The default model that’s going to work best for most startups is a classic functional organization, where you have people in charge of different departments or functions, such as a VP of engineering. This is best for the early stage because when you’re a startup, you have to be tightly focused on one or two things. Everything in the company is then organized functionally, which means that everyone works as a unit and the functional organization facilitates rapid change and adaptability. As the company grows, you need change, depending on the nature of the business. Think of a business unit or a divisional structure, the scale becomes difficult to manage, and there are too many things on which to focus. Therefore you need a more decentralized organization that allows the different components to focus on the things that matter to them. While discovering your model, make sure that you involve the right people, not just your managers but, more importantly, also the informal leaders of the organization. Who will your people turn to if they need help making sense of what’s going on?
What should I take into account when choosing to centralize or decentralize? There are certain things that should be centralized. There is a great book called ‘Scaling up Excellence’; in it, the writers describe two different types of scaling. They jokingly refer to them as Catholic scaling and Buddhist scaling. Catholic scaling is centralized scaling. There’s a single pope; somebody who’s the voice of truth, and things are designed to be the same everywhere. This is appropriate for cases where there’s a relatively stable situation, and precision and replication are important e.g., Intel corporation has a principle named ‘copy exactly’. Buddhist scaling is where globalization comes into play; we will adapt our approach based on the local conditions. In most cases, markets are different worldwide, not just regarding language and culture, but also legislation and consumers. This is why ‘copying exactly’ is not going to work. Decide if Catholic or Buddhist scaling is appropriate for your company, and then decide if you should centralize or decentralize.
You need to hire ‘miss right now’, not ‘miss right’.
Keep the number of direct reports of a manager at less than ten people
What is your preferred organizational design model?
I think you want to have a flat organization, but not to the extent that it creates significant problems. I’m a bit skeptical about Holacracy because I believe it goes against some of the elements of human nature. People want clarity over how decisions are made, and holacracy can be confusing as to how you resolve different conflicts. Traditional hierarchy is not something we need to eliminate completely; still, we would like to keep a flat organizational structure. You can’t flatten it too much, people management takes time, and one can only manage a certain number of people. An example of why it’s important not to flatten the hierarchy too much is Google. In their early days they didn’t like all the layers of mid-management; they wanted their organization simplified. They had all 400 engineers report to the same person. Of course, this experiment didn’t last very long. It’s simply impossible for one individual to manage 400 people directly. It totally destroys the coordination and collaboration of the organization. Because of this, the best advice is to keep your organization flat but not too flat; keep the number of direct reports of a manager at less than ten people. As a result, you’ll have a greater hierarchy when the organization grows and changes from a strict functional organization to a more divisional and business unit organization. It’s simply the only way to keep the decision-makers close enough to the front line to be able to make good decisions.
Find people who really understand how to get you to the next stage.
What do you think about using C-titles at an early stage? One of the things I hate seeing is when a team shows up as a five-person company having five C-titles. It’s a warning sign for troubles later on and it tells me a variety of things, such as too many egos at work. The focus should be on building the business instead of fancy titles. The chance that those five people will still be leading the company when they’re at 1000 employees is very low. The person who’s best at getting the company from stage one to stage two is not necessarily the best for getting the company from stage two to three.
What are your tips and tricks for scaleup founders when it comes to designing their organization? One of my biggest lessons learned is a counter-intuitive rule of Blitzscaling; you need to hire ‘miss right now’, not ‘miss right’. There’s a strong tendency to hire the best person from the world’s leading company because they presumably know how to do the best job. It’s true to some extent, provided that your company is at the same scale as his. If you bring in someone from Google, was he one of the first 100 employees at Google, or employee number 75.000? Will they be able to manage and make things happen in your organization without an army of people and an unlimited budget? The fact remains people really tend to be best in one or two stages, but not all of them. Find people who really understand how to get you to the next stage.
The focus should be on building the business instead of fancy titles.