Add a new member
from my phone, on the fly.
A small pocket onboarder takes a name, an email, and a tier from a phone form or a voice note, picks the right product and community, writes the member in, sends the welcome in my voice, and logs the row. No going home, no logging in, no cool-down.
The yes finally turns into a member, on the spot.
Most creators meet someone interested, smile, say “great, send me your email and I’ll add you tonight,” then go home, open the laptop, log in to the back end, fumble through the right tier, and add the person 12 hours later. By the time the welcome lands, the buyer has cooled. There is a different way. Let a small bot listen on your phone, take a name and email from a form or a voice note, pick the right product and community, write the member in through the platform API, send the welcome in your voice, and log the row. Here is who it is for, what goes wrong without it, how it works, and what you get back.
Who it’s for
Course creators, community owners, coaches, consultants, and anyone selling memberships or seats who closes deals away from the desk. Especially useful if you sell in person, on calls, at events, or in DMs and your buyers say yes before you are at a computer. If you have ever told a hot buyer “I’ll add you tonight,” this is for you.
What goes wrong
Without the bot, the gap between yes and onboarded is 12 to 48 hours. The buyer cools, the welcome lands flat, the first 24 hours of new-member energy is gone, and some buyers ask for a refund before they have even opened the community. A delayed onboarding is a closed sale that opens back up.
How the machine works
The trigger fires from your phone, a short form with first name, last name, email, and tier, or a voice note you dictate on the fly. The bot parses the input into clean structured fields, runs a safety check (dedup + email format), routes to the right product and community based on the tier, calls the platform API to add the member, dispatches the welcome message in your voice on the right channel, and logs the row. You say yes. The bot writes them in. The welcome lands before you have walked away.
What you get back
About 25 to 30 minutes saved per add, the back-home-and-log-in roundtrip gone. Roughly 50 hours back per year on a normal 100-add cadence. Around $50 to $150 of extra revenue per hot-on-the-spot conversion that would otherwise have cooled. Roughly $1,500 to $4,000 a year of recovered revenue from converting people while they are still hot, stacking past $15,000 to $35,000+ over five years. The trip home you used to dread is now a tap on the phone.
The trip home was eating every hot sale.
For a long time the worst part of running a community was the gap between the yes and the onboarded. I would meet someone interested, run a quick demo, watch them light up, and hear the magic word. Then I would do the thing every creator does, I would tell them I’d add them as soon as I got home. By the time I was back at the desk, opened the laptop, logged into the right back end, found the right tier, and clicked add, the buyer was on a different feeling. The welcome message that lands at 11pm hits a totally different brain than the welcome message that lands while the buyer is still glowing from the yes. About 25 to 30 minutes per add, every add, every week, and that was only the time. The revenue cost of letting buyers cool overnight was bigger.
The cost was the cool-down. A hot buyer who gets added on the spot opens the community the same day, says hello, sees the welcome, and starts the first lesson before they go to bed. A cold buyer who gets added 24 hours later is already thinking about the next thing on their plate. Some of them stop replying. A few ask for a refund. The same yes turns into wildly different revenue depending on the gap between the yes and the welcome.
So I built a small pocket onboarder. The moment someone says yes, I tap a short form on my phone (first name, last name, email, tier) or I dictate a voice note like “add Sarah Hill, sarah at example dot com, gold tier.” The AI parses the input into clean structured fields. The router picks the right product, the right community, and the right tier. The bot calls the platform API to write the member in. The welcome message goes out in my voice on the right channel, email, DM, or Telegram, depending on what the buyer prefers. The row gets logged with a timestamp so I can see the full path from yes to onboarded. The whole loop runs in under a minute, anywhere I am, including the airport, the car, or the back of a coffee shop.
The total cost of running it is a few cents per add in API calls and a tiny prompt budget. The total time I spend after the yes is around 20 seconds, the time it takes to fill the form or dictate the voice note. Every new member I add now goes through the same warm, on-the-spot welcome, every time, even when I am out and far from the laptop. New-member activation in the first 24 hours is way up, refund rate on hot-on-the-spot adds is way down, and I am no longer chained to the desk to close a community sale.
Proof point: the community this bot writes against is the one tied to my own YouTube channel, and the on-the-fly mechanics show up in 28 Income Streams Revealed and the lifestyle behind it shows up in A Day In My Life, so the time-saved and on-the-spot revenue numbers on this page are checkable against a real business and not a demo.
Three moves that turn any “yes” into an onboarded member, anywhere you are
What made this work was treating onboarding as something you do in the same minute as the yes, not the same evening. The bot is not just a form, it is a stand-in for the back-end admin you would have hired if you had wanted to spend the money. It hangs on three moves: parse the input cleanly (form or voice), route to the right product and community, and let the welcome message land while the buyer is still hot. Done right, it gives back ~50 hours a year and adds ~$1,500 to ~$4,000 a year of recovered revenue from prospects who would have cooled on the trip home.
Parse the input cleanly, form or voice, no fumbling
The bot starts at the phone. Two entry points: a short form (first name, last name, email, tier, dropdown) and a voice note you dictate on the fly. The voice path runs a transcript pass, then a structured-data pass that pulls first name, last name, email, and tier out of the sentence you spoke. If anything is ambiguous, the bot pings you back with a one-line confirmation before it writes anything. Form or voice, you decide on the spot, the output is always the same clean structured record. That is what stops the bot from breaking on a real-world add.
Route to the right product, community, and tier, every time
The router is the second piece of the framework and the part most onboarding bots skip. Every business sells more than one thing, and every new member belongs in a specific product, a specific community, and a specific tier. The router holds the decision logic in one place: if the tier is “gold”, write the member into the gold product, the gold community, and the gold tier. Swap the offer, swap the rule, the whole bot updates from the next add onward. The bot does not invent the routing, it follows the rule you locked.
Land the welcome while the buyer is still hot
The last move is the one that decides if the member sticks. The welcome message goes out the moment the platform confirms the write, in your voice, on the channel the buyer prefers (email, Telegram, BooSend DM). The template is locked once: a warm hook, the two things to do in the first 24 hours, the community link, your sign-off. The buyer reads it while they are still glowing from the yes, opens the community the same day, and lands in the activation window instead of the cool-down. The welcome stops being a chore and becomes the first impression that earns the second sale.
Once those three moves are in place, the gap between yes and onboarded shrinks from 24 hours to under a minute, and the revenue you used to lose to the cool-down quietly comes back.
Before the bot
- ~25 to 30 minutes per add, the trip home, the login, the click-through
- Welcome message landed 12 to 48 hours late, after the buyer cooled
- Some hot buyers asked for a refund before they had opened the community
- I was chained to the desk to close any community sale, even small ones
- Some adds got skipped entirely on busy travel days, the buyer ghosted
After the bot
- ~20 seconds per add, the form or the voice note, no laptop, no login
- Welcome message lands while the buyer is still glowing from the yes
- New-member activation in the first 24 hours is up, refunds are down
- I can close a community sale anywhere, the airport, the car, a cafe
- ~$1,500 to ~$4,000 a year of recovered revenue from on-the-spot adds
Prompt 1: turn a voice note into clean structured fields
The voice note is the fastest way to add a member when you are on the move, but it is messy by default. Use this prompt to turn a dictated sentence into a clean structured record (first name, last name, email, tier) the rest of the bot can act on without breaking.
Voice-to-data parser for member adds
Act as a structured-data parser for a member onboarding bot. I will give you the transcript of a short voice note I just dictated on my phone, and you will return a clean JSON object the bot can use to write the new member in. Voice note transcript: [paste] The list of tiers I currently sell, exactly as they exist in the platform: [paste] The default tier to use if the voice note does not mention one explicitly: [paste] Email-format rule: must contain one "@" and a valid top-level domain; lowercase the whole address before returning it. Return a single JSON object with exactly these keys and nothing else: first_name, last_name, email, tier, confidence. Pull first_name and last_name from the voice note, capitalised. Pull the email and normalise it. Match the tier to one of the tiers I listed, case-insensitive; if no tier is mentioned, use the default. The confidence field is "high" if all four other fields are unambiguous and the email passes the format rule, "ask" if any field is ambiguous or missing. Do not invent any field. If confidence is "ask", set the ambiguous field to the literal string "ASK" so the bot knows to ping me back before writing.
The output of this prompt is the structured record the router and the platform call both depend on. Lock it once for the business, the bot reuses it on every voice add.
Prompt 2: route the new member to the right product, community, and tier
The record is clean. Now use this prompt to decide which product the new member gets, which community they join, and at which tier, so the platform call writes them into the right place on the first try with no manual fix-up later.
Product and community router
Act as the routing layer of a member onboarding bot for my business. I will give you the structured record from prompt 1 and the routing rules I have locked, and you will return the exact destination for the platform call. Structured record from prompt 1: [paste] The routing rules, as a table of tier -> product_id + community_id + role: [paste] The list of platforms I use and their API endpoints (Circle, Skool, Discord, BooSend, etc.): [paste] Any tier-specific add-ons (a free bonus course, a private channel, a one-on-one call slot): [paste] Return a single JSON object with exactly these keys: platform, endpoint, product_id, community_id, tier, role, add_ons. Match the tier from the record to the rules table and pull the corresponding product_id, community_id, and role. Pick the platform that hosts that community. Build the endpoint string the bot will call. List the add_ons for the tier as a JSON array, empty if none. Never invent a tier, a product, or a community that is not in the rules table. If the tier from the record does not match any row in the table, return a single field error with the string "tier_not_found" so the bot pings me back before it writes.
This is the prompt that decides where the new member lands. Lock the rules table once, every future add follows the same routing.
The 3-minute overview of how this works
Before the build steps, watch this short overview. It’s the exact video from the Automations Made Easy page, and it walks through the mechanics behind machines like this one. 1,000+ students have used these mechanics to save two hours a day, with zero coding.
Want to learn the mechanics behind pocket onboarders like this one?
I teach the same mechanics that make this pocket onboarder run inside Automations Made Easy. 1,000+ students have used these mechanics to save two hours a day and close community sales from the airport, the car, or the back of a coffee shop. No coding required.
Week one: seven adds, all from the phone, all before I noticed.
Most people quit a system like this in the first week because the first add feels too quiet. There is no big spike, no traffic surge, no inbox flood. The first week of a pocket onboarder is meant to feel small. The point is to confirm the form and the voice note both fire, the parser returns clean fields, the router picks the right product, the platform API writes the member in, the welcome message lands in the right channel, and the row logs cleanly.
Week one looks like this. I have seven hot yes moments across the week, two at an event, three on calls, two in DMs. Each time, I tap the form or dictate a voice note from wherever I am. The bot parses, routes, writes the member in, sends the welcome, and logs the row before I have walked away. Seven new members onboarded the same minute they said yes, instead of seven members onboarded 24 hours later from the desk. That is roughly 180 minutes of trip-home time I did not spend, around three hours of the week back. Each hot-on-the-spot add earns ~$50 to ~$150 more than the same add 24 hours later, because the buyer is still glowing when the welcome lands. Quietly, in the background, while I keep enjoying the day.
The point of the first week is not the volume. The point is to prove the loop closes: yes happens, phone fires, fields parse, router decides, platform writes, welcome lands, row logs, all without you opening the laptop.
From there the maths is simple. ~25 to 30 minutes saved per add across ~100 adds a year is ~50 hours, around 1.3 working weeks bought back every year. ~$50 to ~$150 of recovered revenue per hot-on-the-spot add across that cadence is ~$1,500 to ~$4,000 a year. The business earns forever, so over five years the compounding stacks past $15,000 to $35,000+ from the same exact conversations you were already going to have.
All of this runs while you travel, meet people, sleep, or take a week off. The bot does not care. The phone listens, the parser cleans, the router decides, the platform writes, the welcome lands. The community grows on a schedule that matches the conversations, and the back end finally catches up to the speed of the yes.
Prompt 3: write a warm welcome message in my voice, on the right channel
The welcome message is the first impression the new member gets of the community, and it has to land while they are still hot. Use this prompt to write the welcome in your voice, on the right channel (email, Telegram, BooSend DM), so the activation window opens instead of closing.
Welcome message generator (voice-locked)
Act as a copywriter standing in for me on the welcome message to a brand new member of my community. I have a locked voice profile and a welcome template you must fill, not replace, so every new member gets the same warm first impression no matter which tier they bought. Structured record from prompt 1: [paste] Routing decision from prompt 2 (product, community, tier, add_ons): [paste] My voice profile (the words I use, the words I never use, the tone I keep): [paste] The channel the buyer prefers (email, Telegram DM, BooSend, in-platform notification): [paste] My welcome template (hook line, the two things to do in the first 24 hours, the community link, my sign-off): [paste] Write the welcome message as plain text, ready to send through the channel I named. Open with the hook line addressed to first_name. Name the product and the tier so the buyer knows exactly what they just joined. Spell out the two things to do in the first 24 hours, in plain language, with one short sentence each. Drop the community link on its own line. Sign off with my sign-off, no extra closing line. Keep paragraphs short, three lines maximum. Use my voice from the profile all the way through. Do not use the words "welcome aboard" or "we are excited to have you".
The bot runs this pass the moment the platform confirms the write. Every welcome lands in the same voice, on the right channel, while the buyer is still hot. Lock the template once, every new member gets the same warm first impression.
Prompt 4: run the safety check before any platform write
A clean parse is not the same as a safe write. Use this prompt as the last gate before the platform call, so the bot dedupes against existing members, validates the email format, and refuses test entries that would clutter the back end.
Pre-write safety check
Act as the safety gate of a member onboarding bot. I will give you the structured record from prompt 1 and a snapshot of the current member list. Your only job is to decide if it is safe to write this record to the platform, or if the bot should stop and ping me back. Structured record from prompt 1: [paste] Current member list (email, tier, joined_at for every existing member, last 24 months): [paste] Email-format rule: must contain one "@" and a valid top-level domain, lowercase only. Banned patterns: any email that contains "test", "example", "asdf", or my own personal email domains: [paste] Return a single JSON object with exactly these keys: decision, reason. decision is "write" if the record is safe, "ping" if anything is off. Reason is a short string from this fixed list: "duplicate_email", "invalid_format", "banned_pattern", "tier_mismatch_with_existing_record", or "clean". If the email already exists in the list at a different tier, return "tier_mismatch_with_existing_record" so the bot pings me before either creating a duplicate or silently downgrading the existing member. Never write a record on your own, you are a gate, not a writer.
The bot runs this gate before every platform call. No duplicates, no broken emails, no test entries hitting the back end. The cost of being wrong is wasted seats and confused members, so the gate is worth the extra few cents per add.
The exact build, step by step
Wire the phone trigger: a short form and a voice note, both pointing at the bot
The trigger is the source of truth for the whole bot. Point a small workflow (n8n, Make, or Zapier all work) at two entry points on your phone: a short web form (first name, last name, email, tier) saved as a home-screen shortcut, and a voice note shortcut that records audio and pushes the file to the bot. Both entry points fire the same workflow with the same downstream stages. The form is the calm path for when you have ten seconds. The voice note is the fast path for when you are mid-conversation. Either way, the trigger fires once per new member and carries the input forward.
Parse the input into clean structured fields (form skips, voice runs)
When the trigger fires, the bot reads the input source. If the input is the form, the fields are already structured, the parser skips ahead. If the input is a voice note, the bot pushes the audio through a transcript pass, then runs the structured-data prompt against the transcript to pull first name, last name, email, and tier. If anything is ambiguous, the bot stops and pings you back with a one-line confirmation message on your phone before it writes. Both paths converge on the same clean record.
Run the safety check: dedup, format, banned patterns
This is the gate before any platform write. The bot pulls the current member list, checks the email against every existing record, validates the format (one “@”, valid TLD, lowercase), and screens for banned patterns (test, example, asdf, your own domains). If the record passes, the bot moves on. If anything is off, the bot stops and pings you back with a one-line message naming the exact reason (duplicate, format, banned, tier mismatch). The cost of getting this wrong is wasted seats and confused members, so the gate is non-negotiable.
Call the platform API and write the new member into the right product and community
This is where the routing decision turns into a real member. The bot reads the routing output (platform, endpoint, product_id, community_id, tier, role, add_ons) and calls the platform API with an OAuth token from your account. The platform writes the member, returns the new member ID, and confirms the role and tier. The bot also writes any add-ons the tier carries (a free bonus course, a private channel, a one-on-one call slot) in the same pass so the new member lands in a complete state, not half-onboarded.
Send the welcome message in your voice, on the right channel
The platform has confirmed the write. Now the bot fires the welcome message generator from prompt 3, fills the locked template with the new member’s first name and the tier they joined, and dispatches the message on the channel the buyer prefers (email, Telegram, BooSend DM, in-platform notification). The welcome lands while the buyer is still hot from the yes, opens the activation window, and the new member starts the first 24 hours on the right foot. Lock the template once, every new member gets the same warm first impression.
Log the row to the buyers database and ping me on Telegram
The last step is the receipt. The bot writes one row to your buyers/members database (Airtable, Notion, Supabase, whichever you keep) with the parsed fields, the routing decision, the platform confirmation, the channel the welcome went out on, and a UTC timestamp. Then it sends me a tiny Telegram message (“gold member added: Sarah Hill, welcome sent on email + telegram, 14:32 UTC”) so I can see the loop closed without opening anything. No login, no Studio tab, no back end check, just a confirmation in my pocket.
Phone fires
The hot yes happens in person, on a call, or in DMs. You fill the form or dictate the voice note, the workflow fires once with the input.
Inputs cleaned
Voice runs through transcript + structured-data parse, form skips ahead. Safety gate runs dedup, format, and banned-pattern checks. The record is ready.
Member written
Router picks the product, community, tier, and add-ons. Platform API call writes the new member in, returns the member ID, confirms the role and tier.
Welcome and receipt
Welcome message lands in the buyer’s preferred channel in your voice. Buyers database logs the row. Telegram pings your phone with a tiny receipt. Loop closed.
Build this pocket onboarder inside the same playbook 1,000+ students use
Automations Made Easy teaches the mechanics behind bots like this one. Step by step, no code, plain English. Close community sales from the airport, the car, or the back of a coffee shop, and let the welcome land while the buyer is still hot.
The six months after I switched it on
Here is the shape of the first six months after I turned the pocket onboarder on for the business. The line is intentionally modest in the early months and steady from there, because that is how on-the-spot adds actually behave: every hot yes that used to cool overnight now lands as a same-day member, and the layers stack.
Hot-on-the-spot adds converted same-minute, per month
Caption: Monthly count of new members added on the spot through the pocket onboarder instead of from the desk hours later. Three things matter on this chart. The line stabilises around 16 same-minute adds a month and stays there. Every one of those adds is a member who would have cooled if the welcome had landed the next day. And every single one of those months happens without you opening the laptop.
What other students built
I teach the simple skills behind machines like this in Automations Made Easy. Students who built their own version sent back what changed in their first month.
“I used to tell hot buyers I would add them tonight, then forget. Now I dictate a voice note from the parking lot and they are in the community before I start the car. My refund rate on community seats dropped to zero.”
“I sell two tiers and I always added people to the wrong one when I was tired. The router picks the tier from the form, the welcome lands on the right channel, my back end stopped getting messy.”
“Built the voice path first because I close most sales on calls. I dictate the name and email between two calls, the member is onboarded by the time I dial the next one. Saved me about half an hour per add.”
“The Telegram receipt is the part I did not expect to love. I see every new member land on my phone, no back-end check, no anxiety about whether the add went through. The loop closes in my pocket.”
What’s inside Automations Made Easy
AME isn’t a library of pre-built automations. Every business is slightly different. What’s reusable across all of them is the underlying mechanics: how to set up little machines that listen, write, and follow up while you sleep, and how to wire the pieces together without writing code.
The program walks you through six modules: The Right Tools (the cost-effective, no-code stack I actually use), Task Selection Mastery (which automations are worth building first), Design Secrets (mapping an automation before you build it), Zero to Hero (complete beginner to confident automator), Real-World Application (we build a full automation together, end to end), and Monetization Mastery (turn the skill into a side-business).
It also includes done-for-you templates you import in two clicks, over-the-shoulder training videos, and the same playbook 1,000+ students have used to save two hours a day. No coding required. If you can copy and paste, you can build this.
The pocket onboarder: common questions
Pulled from what readers and Automations Made Easy students ask most.
What if the voice note gets the name or email slightly wrong?
The parser is the front door of the bot and it is built to ask once instead of guess. If the voice note is ambiguous on the name spelling or the email is missing the domain, the bot pings me back with a one-line confirmation message on my phone (Telegram or iMessage, whichever I wired in) before it writes anything. I read the confirmation, hit yes, and the bot keeps going. If the parse is clean, no ping. In practice the parser gets the email right on the first pass over 90% of the time because I follow a short habit of spelling out names letter by letter when I dictate. The safety check (dedup + format) is the second line of defence, so even when I miss it, no broken row hits the platform.
Which community platforms does the router work with?
Whatever you already use. The router does not care if your community lives on Circle, Skool, Discord, a private Slack, a Telegram group, a Memberstack site, or a course platform like Teachable or Thinkific. Every one of those platforms has a public API for adding a member, and the router holds one small connector per platform with the credentials, the group ID, and the tier IDs you sell. You add a new platform by writing one more connector, you do not rebuild the bot. The point is that the routing decision (which product, which community, which tier) is made once in the prompt and the connector is just plumbing.
Can I still review the add before it goes through?
Yes, this is the default mode. The bot has two settings: a review mode where it parses the input, runs the safety check, picks the platform and tier, then sends me a one-line confirmation message on my phone with the parsed data and the routing decision, and waits for my yes before it writes the member in. And a direct mode where it writes without asking, which I only use for tiers I sell at one price with no upgrade path. I run review mode 95% of the time. The extra two seconds of tapping yes is worth not having to fix a wrong-tier add in the back end.
Will the welcome message sound generic if the bot writes it?
Not if you give the bot a short voice profile and a locked template. The welcome message generator does the same job as a copywriter standing in for you on the welcome email or DM, which is the first impression the new member gets of the community. The template is a placeholder document with your hook line, the two things you want them to do in the first 24 hours, the link to the community, and your sign-off. The bot fills the placeholders using the member’s first name and the tier they bought, in your voice, on the right channel (email, Telegram, BooSend DM). Lock the template once, every new member gets the same warm first impression, every time.
How much extra revenue does adding people on the fly actually save?
Conservatively, every prospect who says yes in person and is added on the spot is worth around $50 to $150 more than the same prospect added 24 hours later from the desk. The reason is simple: the buyer is hot when they say yes, the buyer is cold the next morning. At roughly 100 on-the-fly adds a year, that is around $1,500 to $4,000 a year of recovered revenue from prospects who would have cooled. Over five years, the compounding stacks past $15,000 to $35,000 from the same exact conversations you were already going to have. The pocket onboarder pays for itself the first time you add someone outside the house.
More Money Makers like this one
Built from the same handful of mechanics. Each one captures, converts, or earns money that would have walked out the door.
Run this pocket onboarder yourself, or learn the mechanics inside Automations Made Easy.
If you want to learn the mechanics behind member onboarding automations like this and build your own at home, Automations Made Easy is the playbook. Step by step, no code, plain English. If you want to talk through which trigger, parser, and router would work for your specific community first, I take a small number of consulting clients each month.
€497 one-time · Lifetime access · 1,000+ students
