In the early days, when an ox could not pull increasing loads, more oxen were added. Collaboration happened because the hierarchy was clear and the authority was with the driver otherwise, the oxen would have chosen their own way.
“No one can whistle a symphony; it takes a whole orchestra to play it” — Halford E. Luccock
Watching an orchestra is interesting, the conductor does not play a single instrument but unifies the orchestra. She/he translates the vision into glorious music. The group collaborates as they have a conductor who can direct. If the conductor is not bringing collective focus or if anyone in the group is assuming authority, noise emerges. Noise may also emerge when communication fails.
The whole is greater than the sum of its parts. — Aristotle
Collaboration is the act of working together in a group to produce something. As shown in the figure below, there is a task and there are a set of people working together to complete it.
What happens as the task scales? People scale, either people are given more tasks or more people are added.
The primary reason for adding more people is to get the task done faster. Does that mean that the sum of individual efficiencies can linearly increase the group productivity? If a bridge can be built with 10 people in 100 days, can a similar bridge be built with 1000 people in 1 day?
Collaboration at scale needs management and that is the reason why Peter Drucker calls management as the most important innovation of the 20th century. One thing that defines the success of collaboration is velocity and agility to get the task done.
“Collaboration is the act of working together in a group to produce something. The success of collaboration is better productivity, which is a consequence of effective decision making, agility, effective communication, and engagement.”
In 2012, Google started an initiative called Project Aristotle to find the answer for, “Why whole is greater than the sum of its parts for some teams?”. They looked at 180 teams all over the company and collected a lot of data. They studied how people in a team behaved in group settings. The findings concluded that as a team, a group of smart people optimized for peak individual efficiency, may not be collectively more intelligent than a group of relatively less smart people. They found that the first most important setting was psychological safety as written here — ‘a sense of confidence that the team will not embarrass, reject or punish someone for speaking up’. The other four were dependability, structure and clarity, the meaning of work, and the impact of work.
A successful collaboration can result in ‘the whole to be greater than the sum of its parts’.
Hierarchy is a representation of the management structure. There are multiple implementations like push hierarchies, pull hierarchies, wirearchies, etc. The structure is the representation of roles which define the responsibilities of the job.
Hierarchy is mostly implemented by bringing in managers, assigning teams under managers, and managers managing the staff with different styles. Direction and strategy are decided from the top and gets executed by the teams. This is a typical model and is more successfully adopted, as it is familiar. Many companies have built successful teams with this model. This model is currently being questioned and not preferred by millennials. Some of the main reasons being the cost of managers, people spending less time in focusing on collaboration and more on managing up, more time in waiting and convincing bosses for permissions for simple decisions, managers playing favorites, etc.
Like many others, engineers at Google believed ‘management is more destructive than beneficial’ and experimented with a completely flat hierarchy and called it ‘Project Oxygen’. The outcome of the experiment from the source, “And as the company grew, the founders soon realized that managers contributed in many other, important ways — for instance, by communicating strategy, helping employees prioritize projects, facilitating collaboration, supporting career development, and ensuring that processes and systems aligned with company goals”. The Project Oxygen team concluded that managers are indeed required, but had to figure out what the best managers did and created a list of qualities for a good manager.
Basecamp believes in a flat hierarchy with 50 employees serving 2.5 million customers. They believe in hiring people with ‘horizontal ambition’ than ‘vertical ambition’. It means they want people who want to master the art of what they do, rather than to master managing what they do. They believe in self-managing teams, with each team having a rotating team leader. They also have three-day weekends from May to October, with a reasoning ‘when there’s less time to work, you waste less time’.
Zappos implemented self-management with a system called holacracy at a scale of 1500 employees. ‘Holacracy’ is a self-governing organizational operating system. It is a structure inspired by the way city operates, where people and business are self-organized and governed by a constitution. The individuals come together when the work needs to be done and dissolve when it is done. Tactical meetings are used to resolve tensions. Holacracy is hierarchical, the hierarchy is circles run democratically. The higher circle is more authoritative than the lower circle. Zappos is using this system from December 2011, with claimed success and argued questions. Medium dropped holacracy calling it was difficult to coordinate efforts at scale.
Morning Star, where 400 employees produce over $700 million annual revenue, has a radical approach to management called ‘Self-Management’. They have proven for more than 40 years that you don’t need managers to run a successful company. Each year, every employee negotiates Colleague Letter of Understanding (CLOU) with associated stakeholders. Colleagues compete here for value delivered and responsibility, rather than promotion. Hiring is triggered by a colleague when he/she feels overloaded, compensation is decided by peers based on their CLOU, no titles or promotions, anyone can buy any required tool, any decision is based on individual responsibility or consensus.
Wirearchy, as defined here, is ‘a dynamic two-way flow of power and authority, based on knowledge, trust, credibility, and a focus on results, enabled by interconnected people and technology’. It operates on champion-and-channel — championing ideas and innovation, over command-and-control. This model believes in effective knowledge sharing and stresses on people working together through connection and collaboration.
The company that developed the games like Half-Life, Counter-Strike has movable workstation which allows you to move your workstation to any floor and move to any team and work on whatever the employee feels interesting. There is also a choice to allocate time for ideas and teams. At Valve, employees working on the same project, rank each other on productivity, skills and other contributions that are used for the overall leaderboard, which determines salary.
Mumbai Dabbawala is an impressive way of sorting, transporting, delivering and returning nearly 200,000 lunch boxes by nearly 5000 dabbawalas at a very low cost. It has been running for the past 125 years. Each day they do nearly 260,000 transactions in 6 hours and the error rate of 1 per 16 million deliveries. The workforce is semi-literate, self-managed and collaborates without the use of IT systems or mobile phones. They have a flat organizational structure with nearly 200 units of 25 members each and these members operate with autonomy. They follow a set of guidelines but manage themselves on hiring, conflict resolution, customer acquisition, and retention. Dabbawala is a proof of successful collaboration at scale, without people with extraordinary talents.
“Scaling business is about scaling people to effectively solve more problems.”
In global companies, authority and hierarchy become confusing, especially when globally distributed teams need to collaborate. The reason is the difference in expectation and the actual style of management. This article gives a good explanation on how this works across cultures.
American managers are egalitarian but typically practice top-down decision making. They are also quicker and flexible with the decisions. The Japanese managers are hierarchical but have a strong tradition of building consensus. Japanese managers follow two methods of consensus. The first one is nemawashi — the practice of speaking individually with stakeholders to agree on advance. The second ringi — the practice of passing proposal across layers starting from the bottom. They take time to decide, but once it is done, it is not very flexible.
Both Japan and America being successful economies, it is hard to say whether hierarchy is right or wrong. However, in global companies, authority and hierarchy can become confusing, especially when globally distributed teams need to collaborate. But having people across cultures, it is better to educate and be explicit about how collaboration happens.
Choosing or changing an existing model is not easy, it depends on the size of the organization and the duration of the current model. People with senior roles may not be happy with the change, relatively to people with junior roles. For a person who is EVP or VP, who would have earned the title after years of hard work, the title may have been the inspiration to join the company. It may be a representation of career growth for him/her and it would be hard to give it up overnight for a new model. There can also be resistance from people who believe in vertical growth and who have moved from their mainstream work to management.
It is relatively easier to start the model earlier in the company when it is small, as Tony Hsieh believes if he could have done anything better, he would have done it earlier. Morning Star started it quite earlier.
Choosing a model at the executive level and then pushing that may not also work. It is better to experiment in smaller groups and then sell the results and findings to other teams. The results may also show that the present model is more productive than the new one and the only way is to try more models based on the learnings and feedback from the experiments. It is better to hire people who believe in your style of collaboration.
For collaboration to happen, since many people are involved, it becomes very important to understand how decision making works. Everyone should understand and agree upon how authority works in the group and how it is communicated and later executed as a group. Understanding this is important because it becomes one of the important dimensions of culture.
A good organization will have a good intersection of representation, implementation, and execution of hierarchy. Though many of these companies have interesting models, it may not be easy for everyone to work on these models. A successful model for one organization may not be the same for another. As Valve handbook starts with the sub-title
“A fearless adventure in knowing what to do when no one’s there telling you what to do?”
Collaboration becomes difficult with organizational growth. So, hierarchy may not be easily eliminated as the organization grows but can be implemented well — by respecting diversity and better ways of adopting and implementing authority and decision making. Choose a model that is not coming in the way of collaboration and works in the context of the organization.
“First, we shape our structures. Then, our structures shape us.” — Winston Churchill
In the present economy where people move across companies very fast, it is better to be explicit about how hierarchy, authority, decision making, and communication work in the company. It is better to hire people who believe in your style of collaboration. It is good to keep in mind what they aspire for. For a few people, title and promotion are more important than a salary hike.
“If you are planning to change the collaboration model, it is better to make the framework simple and familiar, so it can be adopted easily.”
The next decade may see dilution of command-and-control model and it will be interesting to see more innovations coming in hierarchy implementations. It will be interesting to see models that can handle organizational scale, keeping the basics right and can make the whole be greater than the sum of its parts.