Clief Notes Competition #7: The Operator

Stop bleeding money
on complaints.

A complaint resolution operator that investigates before it apologizes, protects your cleaners, and turns complaints into long-term clients.

10,000+

Cleans completed

5

Possible resolutions

Live demo

Try it yourself

Pick a complaint. Watch the operator investigate, decide, and draft every response in real time.

Complaint Resolution Operator

Online
OP

Hi, I'm the Complaint Resolution Operator. Paste a client complaint and I'll tell you what to ask, what to check, and what to do. I'll draft the client response and the cleaner notification for you. Pick a scenario below or type your own. Nothing is saved or sent anywhere.

Try the pre-built scenarios or type your own complaint. Nothing is saved.

The problem

The industry's default is broken

What most companies do

  • Companies over-promise what one visit can do
  • Cleaners can't deliver on it and start cutting corners
  • Good cleaners burn out and their hourly wage tanks
  • Client complains, you apologize and throw money at it
  • When the client complains, nobody checks the evidence

What this operator does

  • Check the checklist before responding
  • Check the timestamps. Did we go overtime?
  • Read conversation history for expectation gaps
  • Assess the client's risk profile
  • Protect the cleaner when the evidence backs them
  • Guarantee quality, not completion of everything

Every complaint ends in one of five places

Five resolutions. No guessing.

1

Education

Not in their package. Explain what's included, recommend the right one.

2

Another Appointment

Task wasn't checked off. Schedule a follow-up. No discount.

3

Reclean

Task was checked off but quality wasn't there. Send someone back.

4

Gift Card

Expectation gap on our side. $100 to $300 depending on severity.

5

Escalate

Client wants a full refund. Goes to a human, every time.

Decision logic

How the operator thinks

Six rules. No guessing. Every complaint runs through the same evidence-based process.

The checklist is the evidence

If it wasn't checked off, the cleaner is telling you they didn't do it. That's a follow-up appointment, not a refund.

Check the timestamps

If the cleaner went overtime, the client got free service. If they left early, that's a red flag. The clock doesn't lie.

Never hunt for problems

Don't ask for photos. Clients who go looking will find issues that weren't there. But if they send photos on their own, use them.

Gift cards are for our mistakes

Not for when the client is disappointed with the nature of the service. If the system worked and they're just unhappy, they need another appointment, not money.

Protect the cleaner

If the checklist shows they did the work, defend the work. Bad complaint handling burns out good cleaners faster than bad clients do.

Read the risk signals

First-time one-offs want perfection. "We'll see how it goes" means they're already skeptical. Clients who skipped the Special Package will complain Maintenance wasn't thorough enough.

Why I built this

Every rule was learned the hard way

I'm Ruby Sparks. I run a luxury house cleaning company. 10,000+ cleans and counting.

Early on, I handled complaints the way everyone does. Client says something's wrong, you apologize, you throw money at it, you move on. I gave away refunds I didn't owe. I burned out good cleaners because a client complained and I took the client's word without checking the evidence.

The turning point was the checklist. It's timestamped. If the cleaner didn't check it off, they're telling you they didn't do it. If they did check it off and the client disagrees, that's a completely different conversation. Two paths. Two resolutions. No guessing.

This operator encodes that system. Complaints are rare now. When they come in, the system handles them. No panic, no bleeding money, no losing good cleaners over bad information.

Get started

Use it yourself

Drop the folder into a Claude project. Paste a complaint. The operator handles the rest.

complaint-resolution/
├── identity.md          Who the operator is
├── rules.md             Full decision logic
├── examples.md          4 real scenarios
├── reference/
│   ├── package-overview.md
│   ├── gift-card-rubric.md
│   ├── client-risk-signals.md
│   └── response-templates.md
└── README.md            How to use it
View on GitHub