Bonus Onboarding Script In The End.
0) Goal of this onboarding call (say this first)
“Today I want to do 3 things:
- Understand your current process + where it breaks
- Confirm what a ‘successful automation system’ looks like
- Lock scope, timeline, access, and next steps so we can start building immediately”
1) Set the frame (premium positioning)
“I don’t build one-off workflows.
I build reliable automation systems that keep working even when:
→ data is messy
→ APIs fail
→ volume increases
→ edge cases happen
That’s the difference between a $100 workflow and a $5000 build.”
2) Quick business context (5 minutes)
Ask:
- “What team will use this daily?”
- “What’s the main outcome you want from this automation?”
- “What happens if this automation fails for 1 day?”
- “What’s the biggest pain right now: time, errors, missed leads, delays, reporting?”
Confirm back (repeat in 1 line):
“Perfect. So the goal is: [Outcome], for [Team], and failure costs you [Impact].”
This makes them feel understood + justifies premium.
3) Process mapping (the real onboarding)
Say:
“Now I’ll map the process exactly like a system.
We’ll go step-by-step from input → processing → output.”
Ask them to explain their current workflow:
“Walk me through it like I’m a new employee.
From the moment data enters… to the moment it’s done.”
While they talk, capture:
A) Inputs
- Where does the data come from?
- Form, website, email, CRM, WhatsApp, spreadsheet, webhook, Stripe, etc.
- What fields do you receive?
B) Processing
- What decisions happen?
- What conditions matter?
- What transformations happen?
C) Outputs
- Where should the final result go?
- CRM, Slack, Notion, Google Sheets, Airtable, email, dashboard, etc.
D) Ownership
- Who needs to be notified?
- Who approves what?
4) The “high-ticket” questions (this is what upgrades the price)
These are the questions that turn you from workflow builder → automation architect.
A) Data quality + validation
“Before we automate, I need to know what ‘bad data’ looks like.”
Ask:
- “Do you ever get missing emails / wrong phone numbers?”
- “Do names come in different formats?”
- “Do you want the system to block bad entries or fix them automatically?”
Decision:
- “If email is invalid → do we skip, correct, or send to review?”
B) Duplicate handling
“Duplicates are the silent killer of automations.”
Ask:
- “What counts as duplicate?”
- same email? same phone? same company?
- “If duplicate exists, should we update the old record or ignore the new one?”
C) Failure handling (most important)
“APIs fail. It’s normal. The question is: what should our system do?”
Ask:
- “If an API fails temporarily, should we retry 3 times or 10?”
- “Do you want fallback logic?”
- “Should failures notify you instantly or only after multiple retries?”
D) Rate limiting + scale
“Even if you have 100 records today, we build for 10,000 tomorrow.”
Ask:
- “How many records per day/week?”
- “Any seasonal spikes?”
- “Any tool limitations you’ve hit before?”
E) Human approval points (if needed)
“Should this run fully automatically, or do you want approvals at specific steps?”
Examples:
- Send proposal email automatically vs approval first
- Push to CRM automatically vs review first
F) Compliance + security
“Since this touches business data, we’ll do it properly.”
Ask:
- “Any sensitive data involved? (payments, health, personal IDs)”
- “Who should have access to credentials?”
- “Do you want separate dev/prod environments?”
5) Define the scope clearly (avoid scope creep)
Say this:
“To keep this clean, I define scope in 3 layers:
Must-have, Nice-to-have, Later.”
Must-have (Phase 1)
“Let’s lock what must work 100%.”
Examples:
- Capture leads
- Validate data
- Enrich info
- Push to CRM
- Notify team
Nice-to-have (Phase 2)
Examples:
- AI scoring
- Smart summaries
- Auto-proposals
- Advanced routing
- Analytics dashboard
Later (Phase 3)
Examples:
- Multi-language
- Voice agents
- Full agentic workflows
- Additional channels (WhatsApp/SMS)
Confirm:
“For this build, we will deliver Phase 1 fully. Phase 2 is optional add-on.”
6) Explain your build structure (premium delivery)
Say:
“I’ll build this like a production system:
→ Main workflow hub
→ Sub-workflows for validation + retries
→ Clean transformations (Code node when needed)
→ Logging + monitoring
→ Separate dev and prod versions
→ Documentation + Loom walkthrough”
This makes the client feel: “this is not a freelancer.”
7) Access checklist (what you need from them)
Say:
“To start building, I need access to the tools involved.
I’ll send a checklist right after this call.”
Ask for:
- n8n access (or confirm if you’re hosting)
- API keys / credentials for:
- CRM
- Google Sheets
- Email (SMTP / Gmail)
- OpenAI / OpenRouter (if AI is involved)
- Any scraping tools (Apify, etc.)
- Sample data:
- 10–20 real examples (good + bad cases)
- Expected output format:
- screenshot or example of “ideal final result”
Important line:
“I don’t need admin access to everything.
I only need the minimum permissions required.”
8) Success criteria (make it measurable)
Ask:
“At the end of the project, how will we know this was a win?”
Give options:
- “Time saved per week”
- “Lead response time reduced”
- “Error rate reduced”
- “No missed follow-ups”
- “Accurate CRM data”
- “System runs without manual babysitting”
Then lock it:
“Perfect. Success = [Metric].”
9) Deliverables + handover (this is where high-ticket closes)
Say:
“Once it’s live, you’ll receive:
→ the complete workflow system
→ a Loom walkthrough (how it works + how to edit safely)
→ a README doc (inputs, outputs, logic, failure handling)
→ monitoring setup (alerts if something breaks)
→ a short training session for your team”
This reduces fear and increases trust.
10) Timeline + communication rules (super important)
Say:
“Here’s how we’ll run this smoothly:
→ I build inside dev first
→ You review output
→ We test edge cases
→ Then we push to production”
Confirm communication:
- “Where should I update you?”
- “How fast do you want updates?”
- daily / every 2 days / weekly
Set expectation:
“If something blocks us (missing access or unclear requirement), I’ll pause and message you instantly.”
11) Final confirmation (repeat the whole plan in 20 seconds)
“Just to confirm:
We’re building an automation system that takes [Input], processes it with [Rules + validation], and outputs [Result].
It will include retries, batching, monitoring, and clean documentation.
Next step: I’ll send the access checklist today, and we start build immediately.”
12) Close with next steps (clear + professional)
“Right after this call, you’ll get:
- Access checklist
- A 1-page scope summary
- Start date + delivery milestones
- First test run plan
Once I have access + sample data, I start building.”
