← All skills

Six Frames · Quick-mode worked example

Example: framing an AI vendor's pitch

A full Quick-mode run. The source is a made-up vendor email, written for this example and typical of what lands in a small business owner's inbox. "FlowPilot" is not a real product.

The input

/six-frames

Purpose: I run a 12-person accounting firm. Should I take the demo call?

Subject: Cut your admin by 80% with AI agents

Hi there, FlowPilot's autonomous AI agents handle your emails, scheduling, invoicing and client follow-ups, so your team gets back 80% of their admin time. Over 10,000 businesses already trust FlowPilot. Setup takes 5 minutes, no technical skills needed. Our agents learn your business in days and never make mistakes. Book a 15-minute demo this week and get your first month free.

The output

Six Frames: FlowPilot cold email

▲ Purpose

Decide whether a 12-person accounting firm should spend 15 minutes on a demo. "Enough" means knowing whether the claims hold up and whether it fits how an accounting firm actually works.

● Accuracy

"Cuts admin by 80%": doubtful. No baseline, no source, and 80% across email, scheduling and invoicing is far beyond typical published results.
"10,000+ businesses trust FlowPilot": unverified. Easy to check against reviews and the company's age.
"Setup takes 5 minutes": possible for a login, doubtful for agents that touch email and invoicing.
"Never make mistakes": doubtful. No AI system can honestly claim this. It's a red flag on its own.
"Learns your business in days": possible, but "learns" isn't defined.

■ Point of view

A vendor selling a demo slot. Every claim is written to get the call booked, not to inform.
Missing: anyone who has used it, anything about client data handling (a big deal for an accounting firm), and what happens when an agent gets something wrong.

♥ Interest

"Never make mistakes" tells you more about the vendor's honesty than about the product.
Notice what isn't listed: no integrations with accounting software.

◆ Value

Useful: it's a prompt to work out which admin tasks actually eat your team's week, with or without this tool.
Noise: the 80%, the 10,000, the free month. None of them help the decision.

▬ Outcome

Conclusion: Don't book the demo off this email. The claims are inflated and it says nothing about client data.
Confidence: high on "inflated", medium on "the product is bad" (the email is bad, the product is untested).
Next action: Reply asking two questions: how client data is stored and which accounting systems it integrates with. Book only if both answers are specific.

Follow-on offered: "Want me to look up FlowPilot's reviews and company age to check the 10,000 claim?"

What to notice

The Triangle did the heavy lifting. "Should I take the call?" is a much smaller question than "is this product good?", and every other frame stayed focused because of it.

The Circle and Square stayed separate. The vendor's bias (Square) doesn't automatically make every claim false (Circle). "Learns your business in days" was rated possible, not doubtful.

The Slab ended with a cheap action that resolves the uncertainty, not a vague "be careful".