A brief history of a solved problem
In the early days of manufacturing, factories worked like most knowledge workers manage email today: each item was handled as it arrived, moving through the shop floor one piece at a time. A machinist would set up their tools for one part, run it, change the setup for the next part, run it, repeat.
The setup time — the transition cost of switching from one task to the next — was enormous relative to the actual work. And because transitions happened constantly, setup time dominated the production day.
The lean manufacturing revolution solved this with a simple insight: group similar work into batches and run them together. Instead of setting up for each part individually, run all the parts that require the same setup at once. The setup cost is paid once per batch, not once per unit. Total throughput increases dramatically even if you are processing the same number of units.
This is not a metaphor. It is a mathematical property of any system where transition costs exist.
The setup cost hiding in your inbox
When you check email, you are paying a setup cost whether you realise it or not.
Your brain shifts context from whatever it was doing — thinking, writing, designing, planning — to email mode. You load the current state of your inbox: who has written, what is unread, what was previously read but not acted on. You make triage decisions. You read. You switch back.
That mode-switching has a cost, just like the machinist's tool change. And like the factory floor, if you pay it for every unit (every email check), setup cost dominates your day.
Research from cognitive science corroborates the manufacturing intuition. Context switching between distinct types of tasks — the kind of switch that happens when you move from deep analytical work to reactive email triage and back — costs 20–40% of productive time according to estimates from the American Psychological Association. That is not the email reading itself. That is purely the transition overhead.
What batch processing looks like in practice
Applied to email, batch processing means: choose when you process email and only process email during those windows.
A two-batch day might look like:
- Morning batch (9–9:30am): Review overnight email, respond to anything actionable, set a flag on items that need more time.
- Afternoon batch (4–4:30pm): Clear the day's remaining items, process anything flagged in the morning.
Everything outside those windows is not ignored — it is deferred to the next batch. The inbox is not checked between batches.
The cognitive benefit is asymmetric: two 30-minute focused email sessions produce better work than fifteen 5-minute reactive checks, even though the total time spent on email is the same or less. The protected deep work windows in between are genuinely protected — not punctuated by interruptions that each carry a recovery tax.
The complication: raw email is not batch-ready
In manufacturing, batching works because the incoming units are known and standardized. You know what came in; you group it; you process it.
Raw email is neither known nor standardized. Your inbox receives newsletters, project updates, automated notifications, client requests, and one-line replies all mixed together. Processing this as a batch is better than constant checking — but you still spend much of the batch time on triage and parsing rather than on the actual work of responding and deciding.
This is the gap a digest fills. A pre-processed digest — where AI has already summarized, categorized, and extracted action items from incoming mail — is genuinely batch-ready. You open the digest with a clear mental model of what is in it, the triage has already been done, and you spend the batch time on decisions rather than parsing.
The factory equivalent would be not just batching the parts, but having them pre-sorted and labelled before they arrive at the machinist's station. Setup time drops to near zero.
The objection worth addressing
"My job is email. I cannot batch it."
Some jobs genuinely require real-time responsiveness — customer support, trading, crisis communications. For everyone else, the feeling that email requires immediate response is more cultural than operational. Expectations are malleable. Most clients, colleagues, and managers adapt quickly when you tell them your response window.
The friction is in setting the expectation, not in the batching itself. Once the expectation is set, the batching works. The days get longer. The work gets better. The email still gets answered.
The factory figured this out before your grandfather was born. The principle did not expire when knowledge work arrived.