Party School · Foundations · Your first hand-off

Hand off one task.

Somewhere in your week is one small thing you do the same way every single time, barely thinking about it. That sameness is exactly what makes it the perfect first thing to hand to AI. This lesson gives you one simple rule for spotting it, the once-a-week test, and then walks you through the whole hand-off: do the task with AI once, save that work as a helper you can reuse, and let it be done for you next week. No wiring, no code, no new subscription. New to all of this is a completely fine place to start.

Updated August 1, 2026 Refreshed monthly A free account is plenty
Core concept 01

The once-a-week test: spot a repeat, then hand it off.

You do not need a big strategy to start. You need one task. The way to find it is a simple rule called the once-a-week test: spot something you do the same way at least once a week, do it with AI one time, then set things up so it comes back mostly done. That is the whole move, and it is the entire scope of this lesson.

Notice the word repeat. The task worth handing off first is not the hardest thing you do or the most important. It is the most repeated. Zapier's team, who have watched people automate work for years, points to the same signal: the tasks worth handing off are the ones that are frequent and follow the same steps every time. If you write roughly the same kind of email, caption, or reply over and over, you have already found your first hand-off. You just have not called it that yet.

Source: Zapier, "When you should automate a task"

Scroll the diagram sideways →

Doing it yourself vs. having it done EVERY WEEK, FROM SCRATCH Doing it yourself 1 · Open a blank chat 2 · Type the whole ask again 3 · Wait, read, copy it out 4 · Next week: start over Same effort, every single time. SET UP ONCE, THEN REUSE Having it done 1 · Open your one saved helper 2 · Paste this week's specifics 3 · Read the draft it hands back 4 · Next week: same helper The hard part is already done.
Same task, two ways · the right side does the thinking once and keeps it
Core concept 02

Your first hand-off needs no wiring at all.

When people hear "automate," they picture a tangle of apps plugged into each other. Set that picture down. Your first hand-off is much smaller and much friendlier: you build one reusable helper and open it yourself when you need it. No apps talking to each other, nothing running while you sleep, nothing that can break at 2am.

A reusable helper is just your instructions and your examples, saved in one place so you never have to type them again. Three easy shapes, all on a free account. The simplest is a saved prompt: write a really good ask once, keep it in a note, and paste it in each week. A tidier version is a custom GPT in ChatGPT or a Project in Claude, where your instructions and a few good examples live permanently, so you just open it and go. Anthropic, the team behind Claude, describes this middle rung plainly for builders: a workflow is a set of steps you define once and reuse, not something that reinvents itself each time. That is precisely what you are making, minus the code.

Go deeper: why "set up once" beats "ask again" every time

The magic is not that AI is smarter inside a Project than inside a blank chat. It is the same AI. The difference is that a blank chat starts from nothing every time, so you supply the context by hand, over and over. A saved helper holds that context for you: your voice, your examples, your rules, sitting there permanently.

So the week you feel rushed is the exact week a blank chat lets you down, because you skip the context and get a generic draft. A helper cannot skip it. The good version is baked in.

Source: Anthropic, "Building Effective Agents"

The once-a-week loop STEP 1 Spot the repeat the weekly same-way task STEP 2 · DO IT ONCE Do it with AI, then save it as a reusable helper STEP 3 Reuse it open it, read, done SAME HELPER, EVERY WEEK
Spot it once, set it up once, then reuse the same helper for good
Core concept 03

You always take the last look.

Here is the rule that makes a hand-off safe: the AI does the drafting, and you do the deciding. Nothing it writes goes out to a client, a customer, or the public until your own eyes have passed over it. That last look is not the boring leftover step you cut later to save time. It is the part that keeps your name on the work.

OpenAI, the team behind ChatGPT, makes the same point in its guide for businesses: a person reviewing before anything is sent is what lets you actually trust the system and keep using it. So build the review in on purpose. Ten minutes with your coffee, or one read before you hit send, is plenty. The draft is the machine's job. The send is always yours. That single habit means a wrong first draft costs you one quick edit instead of an awkward apology.

Source: OpenAI, "A Practical Guide to Building Agents"

Is this task ready to hand off? · three quick questions
Ask yourselfWhat a yes gives youIf it's a no
Do I do it the same way every time? SamenessThe steps barely change from week to week. Great, that is exactly what a saved helper is good at. Pick this one.
Are the stakes low if it's a little off? Safe to draftA wrong first draft costs one edit, not a client. If a mistake is expensive or personal, keep doing it yourself for now.
Is the finish line clear? KnowableYou can tell a good result from a bad one at a glance. If even you are not sure what "done" looks like, the helper can't be either.
The worksheet you leave with

Add up the time one hand-off gives back

Before you build anything, do this quick sum for your candidate task. It turns a vague "this eats my week" into a real number, and that number is usually bigger than you expect.

  1. 01The taskname the one repeated thing, e.g. "reply to booking inquiries"
  2. 02Times per weekhow many times you do it in a normal week, count honestly
  3. 03Minutes eachhow long one goes, start to finish, including the dread
  4. 04Minutes per weekline 02 × line 03
  5. 05Hours reclaimed per month = (line 04 × 4) ÷ 60example: 5 times × 12 min = 60 min/week × 4 = 240 min ÷ 60 = 4 hours a month back
Fresh from the lab

What changed this month

ChatGPT can now hand you back a finished document, not just a chat reply. On July 9, OpenAI launched ChatGPT Work, a bigger helper inside ChatGPT that can work on its own for hours and hand back an actual document, spreadsheet, or slide deck instead of a paragraph you still have to format. It runs on Scheduled Tasks, the same "set a day and time" feature this lesson already teaches, so if you build a helper here, this bigger engine can plug into the same button later.

Even the best AI still gets it wrong more often than right on real business tasks. On July 6, an independent test called AutomationBench-AA gave AI agents 657 realistic jobs, like updating a CRM or replying in Slack, and checked whether they followed the rules along the way. The best AI in the test got it fully right only 48.6% of the time. That number is your best argument for taking the last look every single time, on every hand-off, no matter how good the tool gets.

OpenAI is folding its standalone browser app back into ChatGPT. Also on July 9, OpenAI said ChatGPT Atlas, a separate app for browsing and clicking through tasks, stops working on August 9, 2026. The same abilities move into ChatGPT itself and a new ChatGPT desktop app, so if you'd started using Atlas for anything, the exact same ability just moved home, one app instead of two.

Sources: OpenAI, "ChatGPT is now a partner for your most ambitious work", July 9, 2026 · Artificial Analysis, "Announcing AutomationBench-AA", July 6, 2026 · OpenAI Help Center, "Evolving Atlas into ChatGPT", July 9, 2026

Vocabulary

Five words, in plain terms

You do not need the jargon to do the work, but these five words come up constantly. Here they are with no mystery attached.

Workflow

A fancy word for the same few steps done in the same order every time. Your first hand-off is a tiny workflow: spot it, do it once, reuse it. That's the whole thing.

Trigger

The thing that tells you it's time to run the task. For your first hand-off the trigger is simply you noticing it's that time again and opening your helper. No app required.

Template / Project

Your instructions and examples saved in one place so you never retype them. In ChatGPT it's a custom GPT; in Claude it's a Project; anywhere, it can just be a saved prompt.

Reuse

The point of the whole exercise. You do the hard setup once, then open the same helper next week and the week after. The effort you spent stays spent, working for you.

Human in the loop

The safety habit: a person, meaning you, reads and approves the draft before anything is sent. It's the one step that's never optional, no matter how good the drafts get.

Deeper how-tos, straight from the builders: ChatGPT automations docs · Claude Help Center · Zapier

Make it yours

Hand off your #1 repeat

Same simple method, your one real repeat. Pick your line of work and build one reusable helper: no apps wired together, no code, a free Claude or ChatGPT account is plenty. Budget an hour to set it up, then let it save you time every week after.

For female founders

I'm CEO, marketer, bookkeeper, and intern in one body, and the actual building only happens after 9pm.

The weekly posts helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: for most founders it's writing the week's few social posts, done from scratch every time.
  2. Do it with AI once: paste in three posts you're proud of and say, "This is my voice. Draft this week's three posts on these topics to match it."
  3. Save it as a helper: put those instructions and sample posts into a Project or custom GPT named "My weekly posts," so your voice is baked in.
  4. Reuse it next week: open the same helper, drop in this week's topics, and read what comes back.
  5. Take the last look: edit anything that isn't you, then post. The draft is done, the judgment stays yours.

The payoff: the posts stop eating the 9pm building hours.

For artists

My statements and applications eat the exact studio time the work itself needed.

The application-answers helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: the application answers and statement rewrite you redo from scratch for every open call.
  2. Do it with AI once: paste your artist statement and CV and say, "Draft answers for this open call in my voice, matching their word limits."
  3. Save it as a helper: keep your statement and CV inside a Project named "Applications," so they're always there.
  4. Reuse it next call: open the helper, paste the new call's questions, and let it draft.
  5. Take the last look: read it once out loud, fix the voice, submit.

The payoff: one deadline stops costing you a full studio week.

For actors

Between survival jobs, self-tapes, and submissions, the admin of the career crowds out the acting.

The cover-note helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: the cover note you rewrite from zero for every single submission.
  2. Do it with AI once: paste your resume and say, "Here's a breakdown. Draft a tailored cover note and pick which two credits to lead with."
  3. Save it as a helper: put your resume and a note you're happy with into a custom GPT named "Submissions."
  4. Reuse it next breakdown: open it, paste the new role, and read the draft.
  5. Take the last look: fix anything that doesn't sound like you, then send.

The payoff: submissions go out the same day, not the same week.

For musicians

I'm a musician, not a manager, but I'm the one booking gigs, chasing deposits, and answering "do you do weddings?" at midnight.

The booking-reply helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: the venue and wedding inquiry replies you type again and again.
  2. Do it with AI once: paste a reply you like and say, "Match this friendly-but-firm tone. Reply to this inquiry with my rate, process, and next step."
  3. Save it as a helper: keep your rates, process, and tone in a Project named "Bookings."
  4. Reuse it each inquiry: open it, paste the new message, and let it draft.
  5. Take the last look: read once, tweak, send before it goes cold.

The payoff: inquiries get an answer before they go cold.

For fitness pros

I'm booked teaching all day, and then it's 9pm and I still haven't posted, answered DMs, or chased the clients who ghosted their re-sign.

The weekly programming helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: writing each client's weekly programming, redone from scratch every Sunday.
  2. Do it with AI once: paste your usual format and say, "Here are this week's notes per client. Draft their adjusted programming in this format."
  3. Save it as a helper: keep your programming style and rules in a Project named "Programming."
  4. Reuse it each week: open it, drop in the week's notes, read what comes back.
  5. Take the last look: approve each program before it goes to the client.

The payoff: Sunday admin drops to one sitting, not a whole evening.

For coaches

I sell transformation, but I spend my week on scheduling, session notes, and the follow-up emails I keep meaning to send.

The session-recap helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: the recap and follow-up email that should go out after every session and usually doesn't.
  2. Do it with AI once: say, "From these three bullet points about today's session, draft the recap email and one clear next step in a warm, encouraging tone."
  3. Save it as a helper: keep your tone and a good example in a Project named "Recaps."
  4. Reuse it after each session: open it, paste your three bullets, read the draft.
  5. Take the last look: approve it right then, before you open the next tab.

The payoff: clients hear from you the same day, every time.

For therapists

Between sessions I'm doing intake calls, insurance, and admin, and I went into this to help people, not run an office.

The intake-reply helper, with blanks

Tool: a saved Project or custom GPT · draft only, no client names
  1. Spot the repeat: the intake and insurance FAQ replies you retype for every new client.
  2. Do it with AI once, with blanks: "Draft a warm intake reply using this template, with [NAME] and [INSURANCE PROVIDER] left blank for me to fill in."
  3. Save it as a helper: keep the blank templates in a Project named "Intake." No client information ever goes into a general AI tool; notes stay in your HIPAA-compliant system with a signed BAA, a separate tool entirely.
  4. Reuse it each time: open it, generate the blank reply, fill the blanks by hand.
  5. Take the last look: read once before sending, always.

The payoff: the admin hour between clients goes back to notes, lunch, or nothing at all.

For authors

I want to be writing the next book, and instead I'm writing about the last one, forever.

The newsletter helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: the monthly newsletter you draft from a blank page every single time.
  2. Do it with AI once: paste two past issues and say, "Match this voice. Using these notes on what I've been reading and thinking, draft this month's newsletter."
  3. Save it as a helper: keep your voice samples and structure in a Project named "Newsletter."
  4. Reuse it each month: open it, paste this month's notes, read the draft.
  5. Take the last look: add the one line only you would write, then send.

The payoff: the newsletter stops eating the mornings that belong to the book.

For chefs

I'm cooking or shopping all day, so menus, quotes, and event logistics live in my head and my texts, until something slips.

The catering-quote helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: the catering quote email you assemble from scratch for every single inquiry.
  2. Do it with AI once: "Using my per-head costs and standard terms, draft a quote email for this headcount and date."
  3. Save it as a helper: keep your pricing and terms in a Project named "Quotes."
  4. Reuse it each inquiry: open it, paste the client's details, read the draft.
  5. Take the last look: adjust anything unusual, then send.

The payoff: quotes go out same day, and the kitchen never waits on the paperwork.

For consultants

I'm delivering client work all day, so my own pipeline goes cold every time I'm busy, which is exactly when it shouldn't.

The proposal helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: the proposal you rewrite from a blank doc after every good call.
  2. Do it with AI once: "From these call notes, draft a proposal using my standard structure and the two case studies most relevant to this problem."
  3. Save it as a helper: keep your structure and case studies in a Project named "Proposals."
  4. Reuse it after each call: open it, paste your notes, read the draft.
  5. Take the last look: set the price and scope yourself, then send same day.

The payoff: the pipeline stays warm while you're billing, which is the whole trick of consulting.

For health advocates

Every client is a crisis, and the paperwork is endless. I'm doing intake, appeals, and case summaries with no time left to grow the practice.

The appeal-letter helper, with blanks

Tool: a saved Project or custom GPT · draft only, no client data
  1. Spot the repeat: the appeal letter and intake packet you rebuild for every new case.
  2. Do it with AI once, with blanks: "Draft an insurance appeal letter template for [DIAGNOSIS TYPE], with [PATIENT] and [DETAILS] left blank."
  3. Save it as a helper: keep the blank templates in a Project named "Appeals." No client or patient information goes into a general AI tool, ever.
  4. Reuse it each case: open it, generate the blank letter, fill it in your own case file.
  5. Take the last look: read the finished letter before it goes out.

The payoff: the paperwork mountain becomes a fill-in-the-blanks library, and client hours go to clients.

For makers

I make the thing, then I photograph it, list it, ship it, and market it, and there's no time left to actually make.

The product-listing helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: writing a title, description, and caption for every new piece.
  2. Do it with AI once: "Here are two plain sentences about this piece. Draft the full listing: title, description, materials, care, and a caption, in my shop's voice."
  3. Save it as a helper: keep your voice and format in a Project named "Listings."
  4. Reuse it each piece: open it, paste your two sentences, read the listing.
  5. Take the last look: read each one before it goes live, five takes ten minutes.

The payoff: listing night shrinks to listing hour, and the studio gets its maker back.

For nonprofit leaders

I'm chasing grants, thanking donors, running the program, and writing the report all at once. I'm the whole org, and the mission is waiting on my inbox.

The donor thank-you helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: the donor thank-you letter that should go out within a week and sometimes takes a month.
  2. Do it with AI once: "Draft a thank-you letter for this gift, personalized by amount and by whether they're a first-time or repeat donor, in our warm tone."
  3. Save it as a helper: keep your tone and a great example in a Project named "Thank yous."
  4. Reuse it per gift: open it, paste the gift details, read the draft.
  5. Take the last look: add the one human line, then send.

The payoff: every gift gets thanked the same week, and the mission stops waiting on your inbox.

For photographers

I'm booked shooting on weekends and editing all week, so inquiries wait and the blog died years ago.

The inquiry-reply helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: the inquiry reply that used to go out same-day and now waits until editing week ends.
  2. Do it with AI once: "Reply to this inquiry with my packages, pricing, and availability, matching my usual warm tone."
  3. Save it as a helper: keep your packages and tone in a Project named "Inquiries."
  4. Reuse it each inquiry: open it, paste the new message, read the draft.
  5. Take the last look: personalize the first line, then send.

The payoff: inquiries get a same-day reply without hiring a VA.

For realtors

My database is a graveyard. I forget to follow up with last year's buyers, and every new listing is a scramble of descriptions and emails.

The listing-description helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: writing a fresh listing description for every new property.
  2. Do it with AI once: "Here are the details for this listing. Draft the description in my style, plus one email to my buyer list about it."
  3. Save it as a helper: keep your writing style and a description you love in a Project named "Listings."
  4. Reuse it per property: open it, paste the new details, read the draft.
  5. Take the last look: add a personal line for anyone you actually know, then send.

The payoff: every listing gets written up the day it's ready, not whenever you catch up.

For stylists

I'm behind the chair all day, so rebooking, DMs, and content pile up, and the clients I forget to rebook just don't come back.

The rebooking-text helper

Tool: a saved Claude Project or custom GPT · free tier · no wiring
  1. Spot the repeat: the rebooking text you mean to send and don't, until the client's already gone elsewhere.
  2. Do it with AI once: "Here's a client and their usual service. Draft a friendly, specific rebooking text in my voice."
  3. Save it as a helper: keep your tone and a good example in a Project named "Rebooking."
  4. Reuse it weekly: open it, list who's due, read the drafts it hands back.
  5. Take the last look: send them yourself in a slow ten minutes between clients.

The payoff: an empty chair becomes a choice, not an accident.

The house rule

Lauren's line, for your very first hand-off: what belongs to you and keeps your humanity, you keep. What frees you to be more human away from your laptop, you hand off. The drafting is the machine's. The last look is always yours. Spot one repeat, build one helper, and reuse it until it feels normal. That's it. That's the whole first hand-off.

Next lesson: Put something real out there →