Thank you. Great insights, as usual, and one or two really strong nuggets I'll repost.
One minor note on Conway's Law - the system ends up mirroring the *communication* structure in an organization, not necessarily the organizational structure. And I've seen this in practice - take an existing org structure and add additional structured communications between two siloed teams, and all of a sudden you'll start to see more and better communication between those parts of the system as well, without a reorg.
So, an org structure that perfectly maps the desired system architecture will still produce something else (like system patches to route around the org friction) if the right kinds of communications don't happen between the teams that represent the communication points in the system. And an incumbent willing to invest the resources and disruption to improve communications between siloed teams may be able to change their architecture some without a reorg.
This means, for example, that if your org has communication policies or habits that add friction to communications (default to use email over DMs, hard limits on meeting length or quantity, default private Slack groups, etc.), even a full reorg may change very little in the underlying architecture without upending the communication practices as well.
All that said, though, in many cases there's not much practical difference for incumbents. Because often, the sclerosis of the communication structures is one of the main things that prevents the incumbent from re-organizing in the first place. So, in a real world sense, the importance of communications over org structure may make it even harder for incumbents to compete.
And communication friction increases with org size (Price's Law may contribute here). So, the larger the org, the harder it will be for them to compete, even if they pay the tax to reorg for the new architecture required.
I work a lot in startups and more flat teams where the org structure and communication structure tend tpo overlap quite extensively, but this is an important nuance anyway when it comes to larger groups.
Do you think it’s possible for a company to have the ideal org structure without the communication structure though?
My most common experience has been that senior leadership rarely thinks about what the best org structure is from a system standpoint. They tend think of it from a command and control perspective instead, because that's their area of expertise.
So the system architecture ends up mirroring the current political dynamics rather than optimizing for the best work and information flows. Exactly as Conway's law predicts.
I have, rarely, seen organizations do reorgs focused on the information flow. But even then, they rarely change any of the communication piping in the process. So they end up just moving the bottlenecks around (and amplifying them due to the org chart disruption) instead of fixing the bottlenecks that ultimately cause the problems.
I suspect the reason system structure mirrors communication structure is because of the friction points in the communication. If communication was frictionless across org boundaries, those boundaries shouldn't affect the system structure. We'll never see that ideal state - we humans are natural friction generators. But I've seen that when a focus is made to reduce the friction, the system structure can change pretty quickly even without a reorg.
(Well... at least until the person who drove that friction reduction leaves the org; when that happens, things often revert, and the system breaks down along political fault lines again).
Idle speculation: I almost think it would make sense to have a C-level role whose only (or at least main) job is to reduce communication friction. Because the people usually tasked with it have so many other responsibilities that friction reduction usually takes a back seat until something seriously and obviously breaks.
In any case, I agree with everything you're saying. I just think that a reorg isn't enough. It seems like it usually takes an no-holds barred interim CEO - someone willing break the whole communications infrastructure down and rebuild it - to reduce enough friction for a reorg to make a real difference.
And yes, that means incumbents are almost always ripe for disruption by a competitor who doesn't have to navigate territorial conflict to organize around a new paradigm.
Thank you. Great insights, as usual, and one or two really strong nuggets I'll repost.
One minor note on Conway's Law - the system ends up mirroring the *communication* structure in an organization, not necessarily the organizational structure. And I've seen this in practice - take an existing org structure and add additional structured communications between two siloed teams, and all of a sudden you'll start to see more and better communication between those parts of the system as well, without a reorg.
So, an org structure that perfectly maps the desired system architecture will still produce something else (like system patches to route around the org friction) if the right kinds of communications don't happen between the teams that represent the communication points in the system. And an incumbent willing to invest the resources and disruption to improve communications between siloed teams may be able to change their architecture some without a reorg.
This means, for example, that if your org has communication policies or habits that add friction to communications (default to use email over DMs, hard limits on meeting length or quantity, default private Slack groups, etc.), even a full reorg may change very little in the underlying architecture without upending the communication practices as well.
All that said, though, in many cases there's not much practical difference for incumbents. Because often, the sclerosis of the communication structures is one of the main things that prevents the incumbent from re-organizing in the first place. So, in a real world sense, the importance of communications over org structure may make it even harder for incumbents to compete.
And communication friction increases with org size (Price's Law may contribute here). So, the larger the org, the harder it will be for them to compete, even if they pay the tax to reorg for the new architecture required.
This is a great call out.
I work a lot in startups and more flat teams where the org structure and communication structure tend tpo overlap quite extensively, but this is an important nuance anyway when it comes to larger groups.
Do you think it’s possible for a company to have the ideal org structure without the communication structure though?
My most common experience has been that senior leadership rarely thinks about what the best org structure is from a system standpoint. They tend think of it from a command and control perspective instead, because that's their area of expertise.
So the system architecture ends up mirroring the current political dynamics rather than optimizing for the best work and information flows. Exactly as Conway's law predicts.
I have, rarely, seen organizations do reorgs focused on the information flow. But even then, they rarely change any of the communication piping in the process. So they end up just moving the bottlenecks around (and amplifying them due to the org chart disruption) instead of fixing the bottlenecks that ultimately cause the problems.
I suspect the reason system structure mirrors communication structure is because of the friction points in the communication. If communication was frictionless across org boundaries, those boundaries shouldn't affect the system structure. We'll never see that ideal state - we humans are natural friction generators. But I've seen that when a focus is made to reduce the friction, the system structure can change pretty quickly even without a reorg.
(Well... at least until the person who drove that friction reduction leaves the org; when that happens, things often revert, and the system breaks down along political fault lines again).
Idle speculation: I almost think it would make sense to have a C-level role whose only (or at least main) job is to reduce communication friction. Because the people usually tasked with it have so many other responsibilities that friction reduction usually takes a back seat until something seriously and obviously breaks.
In any case, I agree with everything you're saying. I just think that a reorg isn't enough. It seems like it usually takes an no-holds barred interim CEO - someone willing break the whole communications infrastructure down and rebuild it - to reduce enough friction for a reorg to make a real difference.
And yes, that means incumbents are almost always ripe for disruption by a competitor who doesn't have to navigate territorial conflict to organize around a new paradigm.