The work was still getting done

I once resigned from a company I loved because I could no longer reconcile what the organization wanted to accomplish with what it was asking people to absorb in order to accomplish it.

At the time, I was regularly working twelve-hour-plus days. My team was under-resourced even as the work around us continued to expand. I was moving between licensing and auditing, customer negotiations, new-product introductions, technical support, change management, data, and operations. My team supported a growing go-to-market organization while the infrastructure underneath it was not growing at the same rate.

The interesting thing is that the work was still getting done. That was part of the problem.

From the outside, an organization can look extraordinarily capable when what it actually has is extraordinarily capable people compensating for the organization.

My competence was hiding the constraint

For a long time, I thought being able to hold that much complexity was evidence that I was particularly good at my job. In some ways, it was. But I eventually understood that my ability to compensate for the system was also preventing the organization from fully experiencing the cost of the system.

Every time I caught something before it broke, translated between teams, stayed late, built the missing bridge, or figured out how to execute a decision that had not accounted for execution, the organization received an outcome it had not actually built the capacity to produce.

My competence was hiding its constraint.

Activity can hide the absence of capability

Visible activity feels reassuring. Messages move quickly. Calendars are full. People respond immediately. Problems are escalated and resolved. Everyone seems to be doing a great deal.

But activity and output are not the same thing, and output and capability are not the same thing either. A team can produce enormous output while spending much of its energy compensating for the conditions under which the output is produced.

Someone manually reconciles data before every forecast. A manager attends meetings because decisions stall without them. An operator maintains a private spreadsheet because the official system cannot represent what the work requires. The work happens, but the organization has borrowed the capacity to make it happen.

Early-stage companies often depend on extraordinary human flexibility for good reasons. Roles are fluid. Information changes quickly. People cross boundaries because the boundaries themselves are still emerging.

The trouble begins when a temporary operating condition becomes an identity. ‘We move fast’ can mean planning is someone else’s problem. ‘Everyone pitches in’ can mean ownership is permanently ambiguous. ‘High agency’ can mean people are expected to solve around structural constraints without authority to change them.

What began as flexibility becomes dependency. The organization still experiences the benefit, but the cost has moved somewhere less visible.

Borrowed capacity eventually becomes a constraint

Human beings are remarkable shock absorbers. We remember what the system forgets. We translate between teams whose language no longer matches. We notice exceptions. We maintain relationships that keep brittle processes moving.

But borrowed capacity has limits. Eventually the person leaves, burns out, stops volunteering the extra hour, or simply encounters another demand they cannot absorb. The organization experiences this as a sudden failure even though the structural deficit existed all along.

The useful question is not only ‘How much are people doing?’ It is ‘What are people spending their capacity making possible?’

Sometimes the constraint is not that people need to do more. It is that the organization needs to stop borrowing so much from them.

The system can mistake compensation for capacity

I have an old note that says, almost breathlessly: I hear everything. I see everything. And I feel I must take action on it all.

That is personal, but it is also an organizational hazard. Systems are very good at discovering who will absorb ambiguity. The person who notices the missing handoff gets the handoff. The person who can translate gets more translation. The person who can carry several functions at once becomes evidence that several functions do not need to be staffed.

Eventually competence becomes a subsidy.

I have also written that I will map systems, draw diagrams, pull reports, and join keys all day—but if we do not learn to execute in a shared motion, we will overburden our most valuable resources: our people and our teams. That is the line I care about. The system should learn from human adaptability, not consume it.

The work between the work

Organizations are good at naming visible work: launch the product, close the deal, ship the feature, publish the report, resolve the ticket.

We are less precise about the work required to make those things possible across boundaries. Someone translates what Product means into language Sales can use. Someone notices Finance and Customer Success are using different definitions. Someone repairs a relationship after a difficult handoff. Someone carries context from one meeting into another because the system does not.

Coordination is not absence of execution

When coordination is treated as overhead, people learn to hide it inside other work. The meeting appears on the calendar, but the preparation, translation, follow-up, and relationship maintenance disappear.

Then leaders ask why execution takes so much time. Part of the answer is that execution includes creating enough coherence for multiple people to act as though they are part of the same system.

Translation has a cost

Every boundary creates translation work: technical to commercial, strategy to operations, policy to practice, customer language to product language.

Strong operators often become valuable because they can cross those boundaries. The danger is allowing the organization to depend on a few people to translate forever rather than learning where shared language or better interfaces are needed.

A failed handoff does more than delay a task. It changes expectations between teams. If those expectations are not repaired, the next handoff begins with less trust and more defensive behavior.

That makes relational work part of throughput. Trust changes how much checking, escalation, documentation, and protection a system requires.

The work between the work is not always something to eliminate. Sometimes it is the connective tissue. The design question is whether the system recognizes and supports it—or simply hopes someone will keep doing it for free.

Translation is a mechanism, not a personality trait

For someone who sees patterns and shapes, translation has become a mechanism through which I can be understood. I do it in organizations too: technical to commercial, data to story, vision to execution, one team's language into another's.

But translation should not depend indefinitely on the existence of a gifted translator. If the same boundary has to be manually crossed every day, that is information about the interface.

I will map systems and draw diagrams and pull reports and join keys all day. But unless we learn to execute in a shared motion across the organization, the translation itself becomes another form of invisible infrastructure.