Org design around goals, not ideals

At many decentralized organizations, we’re not only building new technology but also trying to use more decentralized organizational designs. These more autonomous models are in line with our vision of future organizations and networks, but they are not easy — especially when the tech that will support them isn’t yet mature. These experiments involve a lot of learning and iteration. We spend a lot of time talking about self-organization, governance, decentralization, decision making models, leadership, and other organizational concepts.

Sometimes we don’t frame these conversations deeply enough in the context of what we are actually trying to achieve: successful organizations.

Great organizations are made by high-performing people who collaborate effectively together. Our org design should not start with a glossary of terms like ‘flat’ or a menu of options for how to make decisions. We should start with an understanding of the trade-offs inherent in org design, and make clear choices between those based on the goals and context of our organization.

There are two factors that impact how effective a group of people is in pursuit of a mission:

How effective each individual is in working towards the mission
How effective the individuals are in working together

Those are the two levers we have to improve the performance of a team. That’s it. This doesn’t make it easier, but it can make things a bit simpler before we dive into the jargon-filled world of organizational design. (Arriving at the ‘right’ mission is a related but whole other ballgame that’s out of scope for this).

Lever 1: Effective individuals

People are effective in their roles when they are (1) motivated to work hard, (2) effective doing their work, and (3) doing work aligned with that mission. Each of these have their own considerations.

Motivation

Daniel Pink, in his book Drive and now-famous TED-talk, says that people are motivated when they have Autonomy, Mastery and Purpose. Autonomy doesn’t require complete freedom, but it does mean no micro-managing. Purpose comes from pursuit of a goal that intrinsically matters. Mastery means building skills and desired expertise through work.

Effectiveness

The organizational model we choose has a direct impact on the effectiveness of individuals in a few ways.

Attracting talent, whether that is entrepreneurs who crave independence or specialists who want clear roles and mentorship
Employee growth, so employees have room to stretch and effective coaching/feedback to improve. Attract entrepreneurs with promises of independence, or train talent with experienced managers and clear promotion pathways?
Role matching that effectively align people effectively with their expertise, skill level and interests (achieving both personal flow and efficient resourcing)

Direction

How are goals and projects determined to make sure that effective work is actually pushing towards the overall mission? How best to promote:

Information flow: are there clear but rigid patterns (a hierarchy makes it clear with only 1 pathway between any 2 nodes) or a fluid but hard-to-track network of connections?
Collaboration: do clusters make it easier to find and work with complementary people and skill-sets, or to share and learn with others in the same domain?
Decision making: One measure of decision making effectiveness is Quality x Speed x Yield — Effort. What trade-offs make sense?

Perhaps the central question of traditional management theory has been how much to centralize or decentralize decision making authority. Should edges with great expertise over their domain be freed to move fast, or should info be filtered to a center where a broader considerations and alignment can be condensed?

Lever 2: Effective working together

The structures we design for our organization impact how and how well we collectively achieve:

The best strategies and major decisions
Alignment on goals to realize those strategies
Effective resourcing (division of roles & work) against those goals
Quality output
Accountability to results and each other to achieve those outputs
Shared learning & adaptation over time

Organizations are not single entities but rather a collections of individuals. How those individuals behave and what they produce will depend on how they organize. One of my favorite examples of the inevitable impact of org design is in the phones we use every day.

Messaging at Apple vs. Google

At one point Google had 7 different messaging apps on Android! They were frequently criticized for this — how could Google let it happen?

This would never happen at Apple. Apple prioritizes integrated customer experiences. That is their strategy and competitive advantage, and so the organization is crafted for that. Everything is orchestrated around end-to-end products. Decisions about what apps are built and make it onto the iPhone or other platforms ultimately roll up to the leaders of those product experiences. This lets them ensure a great and simple user experience, with no room for diversion. It is unfathomable Apple would ever end up with multiple overlapping messaging apps.

Google is organized differently because it’s strategy has been different. They won with engineering leaps and experimentation, and a wide base of products across the web and other platforms. They are organized around this. As messaging became an important battleground, of course many different teams at Google would see an opportunity and build a product aimed at that market.

Google is not one entity, its a collection of people and teams. Hangouts, Gmail, Messages etc. all came out off different teams. Some started on desktop, some mobile, some other. They expanded and converged. They did not roll up to a single owner or product group to manage away the overlap. Google is not built that way. This hurts coherence, but allows experiments, like the wildly successful Gmail, to emerge. Different strategy, different trade-offs, different org, different outcomes.

Choosing the right org structure

Various models for organizational design with different trade-offs. Concepts like ‘self-managed’ or ‘flat’ are not good themselves, they are designs and each has their own merits and priorities.

Hierarchy

Hierarchy is an organizing tool that gives relatively clear blueprints: do the work in your scope, share and elevate whatever seems important, trust your seniors and subordinates to do the same. It’s not necessarily efficient, but there are clear routes for information and decision flows which dramatically simplifies things. There is only 1 route between any two boxes in the org.

Hierarchy does not need to be command and control. Many good (servant) leaders try to push as much as possible to the edges. But there are defined pathways to elevate major or unexpected situations.

Matrix

Intel CEO and management guru Andy Grove has said that every organization of a certain scale inevitably adopts a matrix, with people belonging to both vertical (e.g., Line of Business) and horizontal (e.g., function a la Finance) organizations at once. This adds more lines of communication, flexibility to shift resources and expertise, and dynamic sharing. It also adds complexity. Org design often becomes about how much power (decisions, budget) rests in the vertical vs. functional organizations and their leaders.

Nested circle

In some ways, forms of nested circles (like holocracy) have a lot in common with the matrix organization, but with more than 2 dimensions. They are better in fast-changing, dynamic environments as they flex to many forms of expertise, organizational groupings, and overlaps. But the same trade-offs also scale, with more ambiguity about where responsibilities ultimately lie.

Fully decentralized network

Removing the organizational units altogether in favor of a web of individuals or roles maximizes flexibility and autonomy. But achieving shared information, agreement, and clarity will naturally be far harder. Because the structure is so dynamic, it is impossible (past a small scale) for individuals to keep a mental model of who works on, knows about, has responsibility over what. So information flow, advice gathering, and decision making all must lost clarity because they can flow in so many ways rather than shared and anticipated ones. This may be worth it for the other factors, but (without much better technology) is an innate trade-off to these org designs.

Design for goals not ideals

Rather than starting with a menu of org design options and a desire to ‘be flat/decentralized/agile,’ it makes more sense to start by weighing trade-offs. Which levers matter most for your organization: motivation, individual effectiveness, aligned direction, good strategy, efficient resourcing, accountability, shared learning? With that rubric in place, org design decisions can optimize for the right things.

The right model depends on what the company is optimizing for. Is the path to winning integrated experiences or innovative leaps? How important is speed in decisions versus broad alignment? How critical is it to attract proactive entrepreneurs, or how much of your workforce will need to progress with steady training? ‘Flat’ might mean the motivational autonomy and speed with no management overhead. But it also might mean less room for coaching, nobody responsible for giving honest feedback and supporting employees’ growth.

On our first retreat for 3Box we went through an exercise as a team, weighing various trade-offs and talking through the toughest ones so we could figure out what we valued, how we should prioritize and compromise, and so how we should design our team.

Over time, we’ll return to this and see if our situation and values have changed, and adjust accordingly. Starting with a clear focus on the goals of our org design, the two levers that impact productivity — individual effectiveness and teamwork — and what we value will make that adaptation far clearer to all and easier to achieve as the needs of our organization evolve.

3Box is building a more human web3, and we’re working hard to create the best team possible. If that sounds exciting, reach out or get involved.

Join our community discord
Check out our open jobs

Originally published on Medium/3Box.