The Bridge Between the Analog Organization and the Digital Output
Those of us who have been around for a while grew up with a machine mindset, both in how we built systems and in how we ran organizations. A system was specified, built, finished and then maintained. An organization was drawn, installed and corrected from above: draw the chart, write the process, add a control where it went wrong.
Neither ever worked that way. With systems, most of us have unlearned it. With organizations, mostly we have not. An organization is grown. It responds to conditions, not to instructions. The work is to cultivate those conditions. It is analog in the sense that matters here: people, agreements, habits, trust and history, none of which runs on instructions.
The work that comes out of it, though, is increasingly digital. Code, data, integrations, platforms, and lately AI. That side is precise, fast and literal. A deployment pipeline does not know what the steering group meant, and neither does an agent.
So there are two sides, and something has to connect them. That something is the work harness: what whoever does the work actually has to act with. A clear outcome, the right context, boundaries, the right to decide, tools, and feedback on what happened.
The harness does not depend on how far an organization has come with AI. A team with no AI at all needs exactly the same six things as a team working beside agents. The same rules and the same conditions help everyone. Setting them up is not an AI project, and waiting for an AI strategy is no reason to postpone it.
The harness is the bridge between the analog organization and the digital output. Friction is what the crossing costs.
My paper, Friction — The Silent Killer, makes this argument on both sides of the bridge. It does not point at the bridge in any one place. This post does.
The analog side
The paper starts where the machine mindset breaks. We Stopped Building Machines describes how software people spent fifteen years learning to treat systems as living things, and then went on managing the organizations around them as machinery. The Picture I Keep Coming Back To replaces the machine with a tree: roots as capabilities, soil as culture, flow as circulation. A living thing responds to conditions.
Part II shows what that means in practice. It Grew Like That says nobody designed the shape the organization has now; it arrived one reasonable decision at a time. A Meeting Grew Here shows what grows where nobody owns a thing. The Conditions Don’t Come in the Box explains why a borrowed structure never brings along what made it work.
Part III is cultivation. Tending What Grew: you cannot redesign a structure nobody designed, but you can prune it. Leading Without Deciding and Growing What Cannot Be Installed are about what no programme can install: design done on purpose in rooms that never thought of themselves as designing, and trust built in small moments, repeatedly.
The digital side
The same paper is just as clear about the other bank. Faster, and Still Wrong: technology multiplies a capability and has no opinion on whether the capability was any good. What AI Lands In: AI brings the same friction, amplified, and far less forgiving of the conditions it arrives into. You Have to Know a Guy and The Long Way to a Yes show what happens when the way into a capability is a person and the way to a decision is a queue.
And it names what digital work needs from the organization before it can move: decisions settled once, in Settling the Few Things; meanings agreed, in Agreeing on What Things Mean; architecture carried by the people deciding, in Equipping, Not Reviewing; and the part that stays human, in Owning What AI Can’t.
Speed makes it visible
None of this is new. What is new is the difference in speed between the two sides.
When building something took six months, a three-week wait for a decision disappeared inside the build. Nobody noticed it, because the work itself was the slow part. Continuous delivery, cloud platforms and now AI have made the digital side fast. The wait is still three weeks, and suddenly it is most of the elapsed time.
Friction did not grow. It moved. It used to sit in what took long to build. Now it sits in what takes long to decide, to agree on, to own. The two speeds side by side make the gap impossible to miss. A team that can deploy in an hour and waits three weeks for a decision is standing on the bridge.
The bridge
That is why friction is not a separate topic next to cultivation and technology. It is the price of the crossing. The digital side is ready. The analog side has not settled what it wants, who may decide, or what a word means.
The Meter Was Always Running shows the same crossing paid in two currencies. A person pays in handoffs, meetings, waiting and questions, spread across a calendar where nobody counts it. An agent pays in context, tool calls and tokens, itemized on an invoice. AI is simply the fastest example so far, which is why it shows the cost most clearly. The bridge was always there.
The paper’s four indicators, in Observing Organizations, are readings taken on that bridge: ownership clarity, handoff loss, decision latency and learning cycle time. They work the same whether the crossing is made by a person or an agent.
Why the machine mindset builds the wrong bridge
When an organization with a machine mindset notices that the crossing is expensive, it reinforces the bridge. More rules. More mandatory context. More checkpoints, approvals and sign-offs before anyone may start. Each addition is sensible, and the harness gets thicker every quarter.
A thick harness is a symptom. Every rule and checkpoint in it exists because something on the analog side was never settled. Adding to the bridge does not make the crossing cheaper, because the problem sits on the bank nobody is tending.
Cultivation is what makes the bridge short. When ownership is clear, the harness needs no approval step. When a word means one thing, nobody needs three documents explaining it. When the decision sits with the people who have the information, nobody has to carry the information to the decision. This is the order Reducing Friction insists on: make ownership unmistakable, move the decision to the information, and only then scale with technology.
The harness worth having is the thinnest one that still lets someone act well. It gets thinner in one way only: by cultivating the organization it leads out of.
In short
- Cultivation tends the grown organization.
- The harness carries its intent into digital work, for everyone who does it, whatever the AI maturity.
- Friction is what the crossing costs when the first is neglected and the second is overbuilt.
- Speed is what makes the crossing visible, and what moves the friction to where the organization is slowest.
That is what the paper has been saying all along, from both sides. Digital transformation does not fail on the digital side. It fails on the bridge, and the bridge is only as short as the organization behind it is well tended.