Org Design for Product Orgs

#CultureOps – Getting conscious about org design

This post came from some thoughts I was mulling over recently. There's a good book on Org Design for Design Orgs, but I have yet to see someone tackle the ways in which a Product Org might look in relation to all the other moving parts of a modern tech company.

An org chart is a bet, much like other bets in product. What it says is "This layout of people is our best bet of how we efficiently and repeatably bring products to market and sustain them over time". If we're going to get better at building product organisations that are resilient, scalable, and repeatable, we're going to have to take some strides to clarify what it takes to design an effective organisation as Product people.

The goal of this post and others in the series will be to open a conversation on getting serious about culture and Org Design as a factor for good product leaders. People are hard. It takes work to manage people and be an effective leader in product management. I've worked through about a dozen reorgs, maybe more over the last 10 years I've been working in tech.

Often reorgs are complicated by unclear logic, poor communication and a lack of open discussion about why and how the reorg helps people be more effective. While every organisation is unique, the things that make bad or ineffective company cultures are surprisingly similar, while the things that make them successful are often unique. When we think of an Org as a kind of product, i.e. a thing that needs intentional design to help someone achieve a purpose (Thanks Randy Silver), we have to be clear about that purpose.

Creating clarity and structure to move people in the right direction is complicated. Janna Bastow wrote a talk in 2017 that highlights this. While the talk Janna gave focuses on stakeholders, it alludes to the ways in which aligning people, which is a core product skill, takes work and effort at multiple layers.

Org Charts, ReOrgs and Outcomes–Oh My

In many places I've worked for and with, org charts and org design get wrapped up in outdated processes and unclear goal structures. Org structures are one of the big frontiers of what i refer to as the "Great Unexamined," the parts of work life that are more taken for granted. The forces that drive reorganisation are often cost savings or mergers and acquisitions, but those forces arent neutral.

ReOrg efforts end up as exercises in political posturing, or short term bets instead of long-term outcomes. These dynamics lead to reorgs that don't fundamentally improve communication, alignment, effectiveness or outcomes, which leads to net neutral customer value at best. Many of the best consulting firms offer a diagnosis to help achieve the perfect structure. Which isn't possible, because organisations are organic, more like a garden than a building. Organisations are organic and like all organic things have minds of their own! (Thanks Ines Liberato)

Often these re-org projects have every intention of improving things but end up caught or slowed by bulky approvals processes and concerns about misalignment. Leaders end up not sharing the right context and thinking and so people end up confused, cynical or anxious about their own position.

There's a real psychological impact on all employees, not just those impacted by organisational changes. Reorganisation can lead to poor wellbeing for employees, in many cases regardless of job cuts. Studies have found that workplace reorganisations can lead to poor mental and physical health for employees, both those who stay and those let go. There can be positive effects for reorgs, but they're mostly correlated with increased individual autonomy in the workplace.

If we take Conway's Law seriously, we should optimise our communication and structures as well as our product. As long as we're making those seams invisible to customers why not ease them for ourselves internally?

What a good ReOrg looks like

If a people org is a kind of product, surely a different way of accounting for people changes can generate better results for the science of org design, whatever science there might be. An emerging important role for the transformation product leader is to provide shape for outcomes and ensure shared understanding of what a good reorg looks like.

A good reorg should begin with a goal statement. It helps set context and frame decision making. Because Org Design is design, it should be guided by an intention we're working towards together. The goal statement of a good reorg might look something like this:
We are better able to deliver value internally and externally better than before and in more sustainable ways. We achieve this with clearer communication, a better shared clarity of purpose at all levels, and a more effective way of sharing context and results cross-functionally.

Fundamentally, the best place to start is with a desired outcome like the one above and work backwards from that. A good reorg starts with people and balances that with clear organisational needs. As always, there's little value in absolute thinking, and is all about clarifying tradeoffs.

A good reorg starts with the friction to customer value and business outcomes and realigns around those two poles. Thinking about people structure as a product helps us begin to start measuring the system. Questions that can help a product leader make tradeoffs in a reorg include:

What are we optmising for by moving folks around? To what end?
What are the outcomes and goals of the reorg?
Who and what is involved in getting value into the hand of customers? How can we align around that better?
What clear things are we not touching or doing right now?

A good reorg is informed by service design. How do we better get value into the hands of our customers through the ways we organise and communicate our work? This isn't just an HR concern, it's a design decision for product leaders.

Looking to the future of product org design

While a product manager needs to know how to bring together other business stakeholders, engineers, and design, the Product Leader has considerations beyond that. The Product Leader isn't just responsible for the viability of good software, they're responsible for the organisation that makes it happen. The capabilites that make that happen have grown and refined over the last decade. Modern product organisations need to think of scale, impact, and distributed teams as much as anything else when it comes to org design.

Modern product management was built on the "Trio" of product, design and engineering. These capabilities are essential for making viable, feasible, and usable digital products. However, the modern product org at scale involves a lot of other capabilities.

There are new and previously unappreciated capabilities in org design for the modern product org. User research, product operations, research ops, data & analytics, product marketing, people & talent, all play a role in org strategy. The Product Leader needs to have an opinion on how to organise all these players into a team that can effectively, repeatably, and reliably deploy software and value into the hands of customers. This means a lot of learning for product leaders from a business or engineering background, as most of the new growth in software support roles is from outside those areas.

These capabilities haven't held as much gravity in discussions of how to shape a modern product org. Partially this is because product leadership is still finding its feet in the C-Suite. Partially this is because former engineers have been over-represented in product leadership and these folks tend to be weaker on design and systems thinking for people. What I hope to do with this series is lay out some thinking about how to do modern product org design at scale.

I would love to hear your thoughts, here with a comment or on Mastodon!