The Context-Switching Tax a Solo Operator Pays
Context switching is the real cost of running many companies alone. Here is what the switching tax actually is and how a solo operator drives it down to near zero.
The hardest cost of running twenty companies alone is not the work. It is the switch. Every time I move from one venture to another, there is a reload: the context, the code, the state of the last decision, the tone of the last conversation. That reload is the context-switching tax, and if you do not manage it, it eats more of your day than the actual work does.
Most people running a portfolio think their problem is capacity. It is usually switching. You are not out of hours. You are spending a third of them paying the reload tax over and over because you organized your day around companies instead of around the work.
What the switching tax actually costs
The tax is not the two minutes of opening a different folder. It is the several minutes after, where your head is still in the last company and the new one is not loaded yet. During that window you are slow, error-prone, and likely to make a decision you would not make with full context. Do that thirty times a day and you have lost hours to a cost that never shows up on any list.
It compounds in a nastier way too. A bad switch produces a shallow decision, the shallow decision creates a thread you have to revisit, and revisiting means another switch. The tax pays for itself in more tax. This is exactly the kind of leak the never drop a thread system is built to catch, because a half-loaded switch is where threads get dropped in the first place.
Batch by kind of work, not by company
The single biggest lever is batching. I do not schedule my day as "company A, then company B, then company C." I schedule it as "all the content, then all the deploys, then all the reviews." Same shape of work, back to back, across whatever ventures need it.
The reason this works is that the reload cost lives in the kind of work, not the name of the company. Writing keeps the same headspace whether it is for one venture or five. Shipping keeps the same tools loaded. When I batch by layer, I pay the setup once and amortize it across the whole portfolio. This is only possible because every venture sits on a shared foundation, so a deploy is a deploy no matter which company it is.
Make the reload cheap with written state
You cannot eliminate every switch, so the second move is making the unavoidable ones cheap. The reload is expensive only when the state lives in your head. When it lives in writing, picking a company back up is reading, not remembering.
So every venture keeps a short current-state note: what I was doing, what is decided, what the next action is. When I switch in, I read three lines and I am loaded. When I switch out, I write three lines and I am clear. The note turns a five-minute mental reboot into a thirty-second read. It sounds trivial. Across twenty companies it is the difference between the portfolio being holdable and not.
Stop switching for things that can wait
The cheapest switch is the one you refuse. A lot of switching is self-inflicted, triggered by a notification that felt urgent and was not. If I let every incoming ping pull me into a different company, I would never finish a single block, and the tax would be the whole day.
So interruptions get captured, not chased. The thing that just arrived goes on a list and waits for its scheduled block. Almost nothing genuinely cannot wait two hours. Protecting a work block from unplanned switches is worth more than answering fast, because fast answers built on half-loaded context are usually wrong answers I have to fix later.
Why this is a solo problem specifically
A team hides switching cost by parallelizing: different people hold different companies, so no one person reloads twenty times a day. Alone, there is no one to hand a context to. Every switch is mine. That makes the tax the defining constraint of solo operation, not a footnote.
It also makes it the thing most worth engineering. Drive the switching tax toward zero and one person can hold a portfolio that looks impossible from the outside. Leave it unmanaged and you cap out at three or four companies, exhausted, wondering why. The tax is the ceiling, and lowering it is most of the job. I lean on automation for the mechanical parts of the reload, which is what tools like ServoAgent and Girard AI exist to carry.