Hire a “Send and Delete” Person
The one hire that protects founder bandwidth — fire the request, delete the task from your brain, trust it gets done
← Productivity · Personal · AI Chief of Staff · 5 Roles
Source & Credit
Shared by Graham Stephan (@GrahamStephan) — real estate investor / YouTuber — after sitting with a “liquid billionaire.”
Original post: x.com/GrahamStephan/status/2078301190832664827
The Idea (in one line)
Hire a “send and delete” person. You fire off the request → immediately delete the task from your brain. They handle it. No follow-ups. No mental overhead. Once you hit send, you consider it done.
If you ever have to remind them, they’re not an asset — they’re slowing you down and stealing bandwidth. Fire and move on as quickly as possible.
Per Graham: this one hire changed how that operator scales everything.
Why It Matters
- Bandwidth is the scarce resource — not ideas. Every “did that get done?” tax is a hidden tax on growth.
- Trust is the product — the role only works if completion is default, not optional.
- Follow-ups are a failure mode — chasing people is management debt; the right person removes it.
- Scale pattern — founders often stay bottlenecked on ops, vendor, admin, and “just handle this” work that should leave their head forever.
What “Send and Delete” Looks Like in Practice
You do
One clear request (outcome + deadline + constraints). Hit send. Stop tracking it mentally.
They do
Own end-to-end: clarify if needed once, execute, close the loop only when done (or truly blocked).
You don’t do
Nudge, re-explain, re-open the ticket, or keep a mental backlog of “stuff I delegated but still own.”
Pass/fail test
If you have to remind them, the relationship fails the send-and-delete standard. Replace quickly.
Who This Role Is (and Isn’t)
| Often is | Usually isn’t |
|---|---|
| Chief of staff / EA / ops lead who closes loops without babysitting | A task collector who waits for you to re-prioritize every hour |
| Someone who treats “sent” as a commitment they own | Someone who needs constant status meetings to move work |
| Clear judgment + follow-through under ambiguity | High output that still requires you as the project manager |
Related team design on this site: 5 roles in AI-first companies (who owns what), AI chief of staff (orchestration layer), management pods if you use that structure.
How to Hire and Run It
- Define the surface area — what can leave your brain forever? (travel, vendors, calendar, payments chase, packing ops, follow-ups, research briefs…)
- Write requests as outcomes — “Book X by Friday under $Y” beats “look into X.”
- One channel — email, Slack, or Asana — not five inboxes.
- Default trust for 30–90 days — then audit: how many times did you re-open a “sent” item?
- If reminders appear, don’t “coach forever” — Graham’s standard is blunt: fire and move on. Protect bandwidth.
AI Angle (Optional Extension)
Agents can be partial “send and delete” for well-scoped work with verification loops (see stop babysitting agents, agentic setup, 4 steps of AI adoption).
But Graham’s point is human: a person you fully trust to close the loop without reappearing in your head. Use AI to reduce load; use a send-and-delete hire to own the load that still needs judgment and follow-through.
Checklist: Are You Still Carrying It?
Related Guides
AI Chief of Staff
Orchestration layer that multiplies a small team — human CoS is the classic send-and-delete seat.
Read →Clear Mental Fog
Dump every open loop in 10 minutes — Delegate lines go to a send-and-delete person.
Run the protocol →Idea shared by @GrahamStephan — View the original post on X →
Comments
Approved comments appear below. Log in once with GFAVIP — it applies across the whole site. GFAVIP login
View comments archive