Overview

This lesson is for admins and operations leads. It pulls together everything earlier in the curriculum into a single setup checklist, then walks through the tuning patterns that matter when you're managing Reputation across a large operation.

If you only manage a handful of properties, you can skim. If you're rolling Sparrow Intel's reputation features out to a team handling thousands of reviews a year across many channels, read carefully.

Part 1: Setup checklist

Connect your review sources

• [ ] PMS connected (brings most OTA reviews in by default)

• [ ] Direct Airbnb integration connected (recommended in addition to PMS pass-through for full review fidelity and host-to-guest reviews)

• [ ] Google Business Profile connected (high-leverage for direct bookings)

• [ ] Browser extension installed for any OTA without API review access (Expedia in particular)

• [ ] Verified at least one test review per source is visible in the inbox

The next lesson — Connecting Review Sources — covers each integration in detail.

Choose your review scale

• [ ] Decided on 5-Star or 10-Point scale

• [ ] Set it in Review Settings

Sparrow Intel normalizes ratings across OTAs into whichever scale you pick. See Review Scale Settings for details.

Configure brand voice

• [ ] Picked the preset guidelines you want (Focus on Positives, Authentic Emojis, Varied Responses)

• [ ] Wrote a short custom prompt (3–6 rules) if your voice needs more than the presets

• [ ] Sent yourself a manual reply on a sample review and confirmed it sounds like your team

The single highest-leverage thing you can do for review-reply quality. See Brand Voice & Multi-Language Replies.

Set up Feedback Requests

In the rule editor at Feedback Requests:

• [ ] Named the request so you can tell it apart in the list (e.g. "4 Hours After Check-In")

• [ ] Set the schedule — a number of hours, before or after, check-in or check-out. (The note on this field is your short-stay guardrail: an after-check-in request won't fire after check-out, and vice versa.)

• [ ] Chose a Request Method — OTA message (through your PMS), or Email if available, OTA message if not

• [ ] Added any conditions (optional) — toggle Any/All, then Add Condition to scope the rule to specific properties, channels, etc.

• [ ] Selected the Requested Fields — Overall Rating and Text are always on; check the rating dimensions you want to collect (Cleanliness, Check-In, Communication, Location, Accuracy, Value)

• [ ] Selected a Message Template from the dropdown ⚠️ empty in this screenshot — the request needs one before it'll send well

• [ ] Set the 5-star redirect — for a 5-star private rating, send the guest to Nothing, Google, or a custom link. (Yes, to the OTA is available only on after-check-out requests, so it's disabled for check-in rules.)

• [ ] Saved with Save Feedback Request

Separately, on Brand Settings — applies to every request automatically:

• [ ] Logo and brand colors set

Then:

• [ ] Tested by triggering a request to a test reservation and confirming it arrives correctly

See How to Use Feedback Requests.

Set up Review and Prediction Rules

• [ ] One starter rule for safe auto-respond (clean 5-stars only)

• [ ] One rule for escalating negative reviews to a supervisor

• [ ] One rule for prediction-based recovery touchpoints (if you're on the plan with Predictions)

• [ ] One rule for issue-driven task creation

• [ ] Tested each rule with a star-only action before enabling real-world side effects

Build these in Rules. See Rules for Reviews & Predictions for patterns.

Set up Host-to-Guest Reviews (Airbnb only)

• [ ] Reviewed the brand voice prompt to make sure host-to-guest output matches expectations

• [ ] Built a rule scoped to your Airbnb properties

• [ ] Designated a daily reviewer for the Scheduled queue for the first month

Connect your task system

• [ ] Operations integration connected (Breezeway, PMS-native tasks, Asana, or webhook)

• [ ] Tested a task created from a real review-detected issue

• [ ] Department mapping verified (cleaning issues land with cleaning, maintenance with maintenance)

See the Task Creation Guide.

Part 2: Tuning for high-volume operations

The above gets you running. The below gets you running well at scale.

Triage view discipline

In a high-volume review inbox, the default "everything" view is not workable. Build saved views your team uses by default:

• Awaiting reply — the daily queue

• Awaiting reply + Negative — the priority queue

• Marked for removal — Marked — the removal filing queue

• Property group X + last 30 days — for region- or brand-specific teams

• Has issues — for ops handoffs

Train every agent to start their day in one of these — never in the unfiltered firehose.

Assignment patterns

Two patterns scale; pick the one that fits:

Reviews don't have an explicit assignee field the way Conversations do today, but informal ownership combined with view filters works at most scales.

Brand voice hygiene

Your brand voice settings are not set-and-forget. The bigger your operation, the more your tone preferences drift.

• Monthly — review your brand voice prompt; refine based on patterns you saw in the Scheduled queue and your edits to AI suggestions

• Whenever you change the brand voice prompt — manually generate replies on 5–10 representative recent reviews and confirm the new voice lands

• Whenever your team's sign-off, naming, or policy language changes — reflect it in the prompt

A team running auto-respond at scale should have a designated brand voice owner.

Rules tuning rhythm

Once review and prediction rules are running, set a recurring review:

• Weekly for the first month — sample 20 auto-responded reviews; flag any you'd have written differently

• Monthly thereafter — broader sample, look for patterns

• Whenever a metric moves — increased Insights surfaced complaint themes, increased removal filings, sentiment shift

Adjust thresholds, conditions, and the brand voice prompt based on what you find. Most "auto-respond got it wrong" issues are actually "brand voice was missing context."

Onboarding new agents

Once Reputation is dialed in, you'll keep adding agents. A standard onboarding:

1. Read the Flight School Reputation Management track end-to-end (this curriculum)

2. Shadow a working agent for half a day

3. Observe how senior agents handle a range of reviews

4. Reply manually only — no auto-respond privileges for the first week

5. Full access from week 2

For a team handling thousands of reviews a month, an agent's ramp from "first day" to "full productivity" should take days, not weeks. The curriculum and the AI assistance both shorten that timeline if you let them.

When to ask for help

Before reaching out to support, ask yourself: was the brand voice off? If Chirp AI said something wrong on a review reply, there's a good chance a small brand voice tweak would fix it.

The in-product chat icon (bottom-right of the portal) reaches our support team. Things worth pinging us about:

• Persistent misclassification (Chirp keeps reading a phrase as negative when it isn't)

• Channel sync issues that don't resolve in a day (review missing from inbox, host-to-guest draft not generating)

• Auto-respond posting things that shouldn't have been posted (we want to know — this is how we improve)

• Removal pipeline tooling friction

• Feature requests — we read every one

Up next

Connecting Review Sources — the integration playbook for Google Business Profile, Airbnb, the browser extension, and the OTAs that aren't covered by your PMS.