Rigel Studio
Private working document

Enter the password to continue

This scorecard is shared privately with clients and readers of the studio. If you were sent this link, the password came with it.

That password isn't right. Try again.
Don't have the password? Message Kateryna.
Scorecard

What to Ask an AI Customer Service Vendor Before You Buy

We built this scorecard out of our own negotiations with customer service AI vendors. A demo is the product at its best. These 8 questions are about what happens after, in production. Use the scorecard interactively, or load it into ChatGPT or Claude and run it as a prompt.

{{ q.num }}
 

{{ q.title }}

{{ q.context }}

Good answer

{{ q.good }}

Red flag

{{ q.bad }}

Score your vendor

Reading the verdict

Your answers update this in real time. Below, the total tells you whether you're looking at a partner, a risk, or a walkaway.

Strong answers
{{ strongCount }}
out of 8 questions
Acceptable
{{ midCount }}
out of 8 questions
Red flags
{{ redCount }}
out of 8 questions
Answered
{{ answeredCount }}
of 8 · {{ remaining }} to go
Verdict
{{ verdictTitle }}

{{ verdictDesc }}

Red-flag count
{{ redCount }}
0 — 2 red flags
Pass

The vendor has operational transparency. Proceed with a contract review focused on SLA language and data ownership clauses.

3 — 5 red flags
Proceed with caution

Clarify flagged questions in writing before signing. Red flags in pricing, data ownership, and maintenance scope are contract-level risks.

6 — 8 red flags
Do not proceed

This vendor cannot answer basic operational questions. That inability is itself a preview of how your team will be treated post-contract.

Or run it with an AI assistant

Turn this into a live evaluation session

Copy the whole framework as a prompt. Your assistant will ask for your context and the list of vendors you're comparing, then walk you through each of the 8 questions per vendor, keep a running scoring table across all of them, and return individual evaluations plus a side-by-side recommendation at the end.

C
Run with Claude

Opens Claude in a new tab. Paste (⌘V) into a fresh chat.

G
Run with ChatGPT

Opens ChatGPT in a new tab. Paste (⌘V) into a fresh chat.

A dialog opens with the full prompt. Copy it, open the assistant, and paste into a fresh chat. Nothing is sent from this page.
Evaluation framework prompt
Copy this into {{ modalTarget }}
{{ modalStatus }}

Running an evaluation and want a second pair of eyes on the answers?

Message Kateryna