Explore

Ticketing

Options on the Table

a) Purpose-Built Ticketing Tools

Examples: Zendesk, Freshdesk, Help Scout.

Pros: Mature, robust workflows (SLAs, escalations, knowledge base).

Cons: Can feel heavy, expensive, and generic. Doesn’t differentiate your agency.

b) Custom Fabric Hub Ticketing (with AI assist)

Use your existing hub (Coda/SharePoint/Notion) as the database of tickets.

Layer AI (ChatGPT) on top to:

Parse inbound emails/Slack messages into structured fields (issue, urgency, client, category).

Auto-prioritize based on rules (VIP client? urgent keyword like “down,” “not working,” “cancel”?).

Suggest draft responses.

Pros: Stays inside your ecosystem, flexible, branded, integrated with your dashboards.

Cons: You’ll need to design the workflows (intake → triage → assignment → resolution → close).

⁠

2. Can ChatGPT Take a Customer Prompt and Turn It Into a Ticket?

Yes—very doable. Workflow looks like this:

Inbound Message → comes from email, web form, or Slack.

Webhook/API call → passes the raw text to ChatGPT.

Parsing Prompt → “Extract account, issue summary, priority, urgency, suggested category from this text.”

Ticket Record → ChatGPT outputs JSON that your hub/database ingests. Example:

{

"Account": "Somma Cruises",

"Summary": "Instagram ad campaign reporting incorrect spend",

"Priority": "High",

"Category": "Paid Media",

"Owner": "Ava Johnson",

"Status": "New"

}

Dashboard View → Appears in your Ticket Table, with AI suggesting first-response draft.

⁠

3. Best Practice Blend

Keep it simple: Use your Fabric Hub as the single source of truth (Ticket Table).

AI’s role: Intake, classification, and response drafting.

Automation’s role: Routing (assign to right AM), SLA nudges (“ticket open 48 hours”), client comms.

Optional: Sync with Teams/Slack so AMs can manage tickets in the same flow as daily digests.

⁠

4. Recommendation for You

Don’t buy Zendesk unless you’re running hundreds of tickets a week.

Instead, build a Fabric Hub Ticket Module with:

Ticket Table (ID, Account, Issue, Priority, Status, Owner, Dates).

AI intake agent (ChatGPT → structured record).

SLA automation (like we did for sales).

Client-facing form/email intake that pipes into the system.

This keeps everything under your “AI + Creativity” narrative—you’re not just answering tickets, you’re running a responsive, branded, AI-augmented client portal.

⁠

FabricHub_Ticket_Intake_Prompt.txt

1 kB

⁠

FabricHub_Tickets_Table

FabricHub_Tickets_Table

Ticket ID

Account

Issue Summary

Category

Priority

Status

Owner

Created Date

Due Date

Source

AI Draft Response

TK-001

Somma Cruises

Instagram ad campaign showing incorrect spend totals

Paid Media

High

New

Ava Johnson

9/5/2025

9/6/2025

Client Email

Thanks for flagging this. We’ve identified a discrepancy in reporting. I’m reviewing the ad sets now and will update you by EOD.

TK-002

Two-Rock Amplifiers

Website contact form submissions not reaching inbox

Web/Tech

Critical

In Progress

John Schimpf

9/4/2025

9/5/2025

Support Form

We’re aware contact form submissions aren’t coming through. I’m working with our dev partner to restore functionality and will confirm within 24 hours.

TK-003

Bayview Law Group

Need updated creative for September LinkedIn ads

Creative

Medium

New

Ava Johnson

9/5/2025

9/10/2025

Slack

Got it—we’ll draft updated creative for LinkedIn ads this week and share for approval by 9/8.

There are no rows in this table

⁠

FabricHub_Ticket_Lifecycle

FabricHub_Ticket_Lifecycle

Stage

Definition

Max Days in Stage

Automation

New

Ticket created, not yet reviewed.

Notify Account Manager on creation; assign default Owner; auto-prioritize using AI if not set.

In Progress

Assigned and actively being worked on.

Daily reminder to Owner; if no activity logged in 48h, escalate to Ops.

Waiting on Client

Awaiting info, feedback, or approval from client.

Send auto-reminder to client every 48h until response; if overdue 5 days, notify AM.

Resolved

Issue addressed, pending client confirmation.

Send resolution summary + draft closing message; auto-close after 48h if no response.

Closed

Ticket completed and archived.

Move to archive; trigger satisfaction survey (optional).

There are no rows in this table

⁠

Stages Included:

New → notify AM, auto-prioritize, assign owner

In Progress → daily reminders, escalate if stuck

Waiting on Client → auto-reminders, escalation at 5 days

Resolved → send summary, auto-close if no client response

Closed → archive + optional satisfaction survey

Want to print your doc?
This is not the way.

Try clicking the ··· in the right corner or using a keyboard shortcut (

CtrlP

) instead.