Overview

For most teams, this is the lesson that changes the math. Once you trust Chirp AI to write the first draft of every review response, the time it takes to clear a review queue drops dramatically — and the quality stays consistent across hundreds of replies a month.

This lesson is about how AI suggestions work, what to look at before sending, and when to send a draft as-is, edit it, or override.

How a suggestion gets generated

When you open the reply composer, Chirp AI:

1. Reads the review body and the AI analysis layer (sentiment, opinions, issues)

2. Looks up the reservation and property

3. Applies your team's brand voice settings (covered next lesson)

4. Drafts a response in the review's language

5. Surfaces the draft in the composer

To generate an AI reply click: GENERATE RESPONSE WITH CHIRP AI

If you want a different angle, click the button again to ask Chirp for a fresh draft.

Where to find the suggestion

Open any review awaiting a reply, then click Reply. The suggested draft will appear in the composer text area, with controls to:

• Send as-is — click Respond and the reply will post via the OTA

• Edit then send — click into the draft, modify it, then Respond

• Regenerate — ask Chirp for a different draft by clicking GENERATE RESPONSE again

• Cancel — exit and wipe the suggestion

You can refrain from generating an AI section if you'd rather write from scratch on a particular review or opt to utilize the template feature.

What Chirp uses to draft a response

Chirp's draft isn't generic. It draws from:

• The review itself — what the guest specifically praised or criticized

• Sentiment and opinions — so the reply addresses the actual sentiment, not just the star rating

• Full conversation context — so the reply acknowledges elements of the stay discussed via conversations during the trip

• The reservation — guest name, dates, property, any other context Chirp has about the stay

• Your brand voice settings — preset (Focus on Positives, Authentic Emojis, Varied Responses) or your custom prompt

• The OTA's character limits and tone conventions — Google responses read differently from Airbnb responses

The more brand voice tuning you give it, the better the drafts get over time.

A workflow that scales

The pattern that works for most teams:

1. Open the next awaiting-reply review

2. Glance at the star rating and the opinions Chirp surfaced

3. Positive review, no issues, high rating → read the draft once, send

4. Mixed or negative review → read more carefully, address the specific complaint in your edit (as or if needed), send

5. Severely negative or policy-violating → consider human escalation or, if applicable, mark for removal

With a few weeks of tuning your brand voice, most positive and neutral reviews land in bucket 3, which means you can blow through them.

When you disagree with the AI

Editing a draft is fine and expected. Chirp gets better the more your team uses it — your edits and patterns inform future drafts.

A few common cases for editing:

• The draft is too formal or too breezy — adjust the tone, or refine your brand voice settings

• The draft missed a specific opinion — call out the praise or address the complaint explicitly

• The draft is technically correct but you want to add a personal touch — go for it

• The draft mentioned something Chirp shouldn't have — strip it; consider whether your brand voice prompt should call out the boundary explicitly

If you find yourself rewriting most drafts in a consistent way, that's a signal to refine your brand voice — either pick a different preset or tighten your custom prompt.

What about auto-reply?

You don't have to send each response manually. With a Rule for Reviews, Chirp AI can send the response automatically when conditions you set are met — for example: 5-star overall rating + 5-star cleanliness + positive sentiment on the review text. That's covered in Rules for Reviews & Predictions.

Up next

Brand Voice & Multi-Language Replies — how to teach Chirp the way your team sounds, and how to handle reviews in any language.