Jono Catliff: Give Different Inputs a Common Shape
How can I prepare messages from different places so an AI assistant can work with them without losing the source or inventing missing details?
Choose, practice, reflectA practical learning exercise
The Lesson
A message arrives by email. Another arrives as a voice-note transcript. A third is copied from a form. You want an assistant to organize them, but each uses different labels and leaves out different information. Before choosing a model, decide what each incoming item should look like.
The La Growth Machine academy chapter featuring Jono Catliff recommends standardizing incoming message types before they reach an assistant. Jono’s official site connects that work with his Automatable teaching. The useful starting point here is input preparation: make the material easier to interpret before asking AI to act on it.
A common shape does not require every message to contain the same facts. It means putting available facts in predictable places and making absence visible. A form might contain a requested date; an email might only say “next week.” Moving both into a field called requested date should not turn “next week” into an invented appointment.
For a practice workflow, keep the original wording alongside your organized version. Include where the item came from, what it asks for and what remains uncertain. That lets a reviewer compare the prepared input with its source. A tidy record is useful only if it remains faithful to the material.
In n8n, the current Edit Fields documentation describes creating or changing fields and choosing which input fields remain in the output. That is one technical building block. Our exercise starts on paper so you can decide what should be preserved before configuring a workflow.
Reflection
Think about one repeated task that receives information from several places.
- What facts does the next person or tool actually need?
- Which labels mean the same thing across those sources?
- What could become misleading if you removed the original wording?
- Which missing details require a question rather than a guess?
Keep the first task small. You are defining a handoff, not trying to automate an entire business.
Practice
Original SelfGrowthVideos exercise: Build a common input sheet. This worksheet is our learning exercise, not a named Jono Catliff method or an endorsement.
Use these three fictional messages:
- Email: “Could you send me details about a beginner workshop next week?”
- Voice-note transcript: “I want the Saturday session, I think. Please confirm the time.”
- Form response: “Topic: spreadsheet basics. Preferred day: Saturday. Contact preference: email.”
Create one record for each message using these fields:
| Field | What to record |
|---|---|
| Source type | Email, transcript or form |
| Original text | The exact fictional message |
| Requested task | What the sender is asking for |
| Stated preferences | Only preferences present in the message |
| Unanswered question | A detail that needs clarification |
Do not combine the three people into one record. Do not infer that the workshop exists, that Saturday has space or that a time has been agreed. For the second message, preserve “I think”; it describes uncertainty, not a confirmed booking.
Now write a short instruction for an assistant: use each prepared record to draft a clarification question, preserve uncertainty and make no booking. You can try it with a tool you already use or write the response yourself. Compare each question with the original message.
Review
Check your three records before improving the prompt. Did you keep each source attached to its item? Did a preference become a commitment? Did the assistant ask for information already supplied?
Change one field definition that caused confusion, then repeat the same small exercise. Keep the earlier version so you can see whether the change helped. Before applying this to real work, define the permissions and review needed for those actual messages.
At your next check-in, explain the input sheet to someone else. Can they distinguish a stated fact, a tentative preference and an unanswered question without asking you?
Go Deeper
Explore Jono Catliff’s video library and AI Agents & Automation. Continue with Liam Ottley’s business context brief or Wes Roth’s reviewable agent handoff.
Use this practice in Side Hustles: AI Services & Business Workflows to describe one bounded task you could demonstrate. A worksheet is evidence of practice; it does not establish customer demand or a working service.