Hospitality Reviews AI Analyzer

Analyze and compare customer reviews from Google Reviews and TripAdvisor for restaurants, hotels, and tourist attractions.

This Actor turns review data into a clear business report showing recurring complaints, positive themes, platform-specific perception gaps, and recommended actions.

It is designed for hospitality businesses, agencies, consultants, tourism professionals, and anyone who wants to understand customer perception across review platforms.

---

What this Actor does

This Actor analyzes review data and generates a structured reputation report including:

• Key review metrics
• Average rating by source
• Sentiment overview
• Most frequent positive themes
• Most frequent negative themes
• Google vs TripAdvisor comparison
• Platform-specific perception gaps
• Business interpretation
• Recommended actions
• Suggested replies to common negative reviews

The main goal is not only to summarize reviews, but to explain what each platform reveals.

For example:

• Google may highlight practical or operational issues such as service, waiting time, parking, or accessibility.
• TripAdvisor may highlight tourist experience, perceived value, price, atmosphere, or overall travel expectations.

---

Supported sources

The Actor supports:

• Google Reviews data
• TripAdvisor Reviews data
• Google + TripAdvisor combined analysis

You can provide only one source, or both sources for a cross-platform comparison.

---

How reviews are collected

You can either provide review data yourself or let the Actor collect it automatically from a public URL:

• Provided data — paste JSON, link a CSV URL, or pick an Apify dataset. The Actor analyzes exactly what you give it.
• Automatic scraping — provide a Google Maps URL and/or a TripAdvisor URL. The Actor then delegates the scraping to specialized, battle-tested community Actors (via Actor.call), waits for them to finish, reads their dataset and feeds the reviews into the same analysis pipeline.

Typical automatic flow:

Google Maps URL + TripAdvisor URL
        → automatic scraping (child Actors)
        → normalization + sentiment/theme analysis
        → cross-platform comparison
        → business report (JSON + Markdown)

The child Actor identifiers are configurable through the GOOGLE_REVIEWS_SCRAPER_ACTOR_ID and TRIPADVISOR_REVIEWS_SCRAPER_ACTOR_ID environment variables. Automatic scraping requires the Actor to run with permission to call other Actors; when it runs under limited permissions the scraping step fails gracefully and the run continues with any other available source.

Users are responsible for ensuring they have the right to use and process the review data they collect or provide.

---

Input options

You can provide review data in several ways. For each platform the highest-priority source available is used, in this order: inline JSON → CSV URL → Apify dataset ID → automatic scraping from a URL. A scraper is never launched when data is already provided for that platform.

1. Inline JSON

You can paste review objects directly into the Actor input.

2. CSV URL

You can provide a public CSV URL containing review data.

3. Apify Dataset ID

You can provide a dataset ID from another compatible Apify Actor (for example the output of a Google Maps or TripAdvisor reviews scraper). Use the dataset picker on the apifyDatasetIdGoogle and apifyDatasetIdTripadvisor fields to select your datasets: this automatically grants the Actor read access at runtime, even when it runs under limited permissions.

4. Automatic scraping from a URL

Provide a public googleMapsUrl and/or tripadvisorUrl. Control the volume and language with maxGoogleReviews, maxTripadvisorReviews, scrapingLanguage (fr, en, es, or all) and cap the spend per source with maxScrapingCostUsdPerSource. The report's scrapingMetadata field and the Data collection section of the Markdown report record how each source was collected.

---

Recommended input fields

Each review should ideally contain:

• rating
• review_text
• review_title
• published_date
• visit_date
• language
• owner_response
• reviewer_name

The Actor will try to normalize different field names automatically.

---

Business types

The Actor supports different theme sets depending on the type of business:

• Restaurant
• Hotel
• Tourist attraction
• Generic business

Restaurant reports can detect themes such as service, waiting time, food, price, atmosphere, noise, cleanliness, location, parking, authenticity, and tourist experience.

Hotel reports can detect themes such as cleanliness, bedding, noise, reception, breakfast, room, bathroom, location, parking, value for money, air conditioning, view, service, and tourist experience.

Attraction reports can detect themes such as price, waiting time, accessibility, organization, visit interest, guide quality, crowding, cleanliness, reception, tourist experience, and value for money.

---

Output

The Actor generates:

• A structured JSON result in the default dataset
• A Markdown report saved as REPORT.md
• A full JSON report saved as REPORT.json

The Markdown report includes:

• Executive summary
• Google Reviews analysis
• TripAdvisor Reviews analysis
• Cross-platform comparison
• Important perception gaps
• Business interpretation
• Strengths and weaknesses
• Recommended actions
• Suggested review replies
• Limitations

---

Example use cases

Restaurants

Compare what local customers say on Google with what visitors say on TripAdvisor.

Useful for detecting whether complaints are mostly operational, price-related, or linked to tourist expectations.

Hotels

Analyze differences between practical guest feedback and travel-experience feedback.

Useful for identifying issues around cleanliness, bedding, breakfast, noise, staff, or perceived value.

Tourist attractions

Understand what visitors praise or criticize most often.

Useful for detecting problems related to waiting time, organization, crowding, pricing, guide quality, or accessibility.

---

Example insights

The Actor can generate conclusions such as:

• “TripAdvisor concentrates complaints related to price and value for money, suggesting a perception issue among tourists.”
• “Service appears more often in Google reviews, suggesting a local or operational friction point.”
• “Waiting time is visible on both platforms, but sentiment is more negative on TripAdvisor.”
• “Google and TripAdvisor reveal different parts of the customer experience.”

---

Pricing suggestion

This Actor is designed to support pay-per-report or pay-per-result pricing.

A combined Google + TripAdvisor report generally provides more value than a single-source report because it compares platform-specific perception gaps.

---

Limitations

• The current version uses rules, keyword detection, and structured analysis.
• It does not use generative AI by default.
• It does not scrape Google or TripAdvisor itself; automatic scraping is delegated to specialized child Actors and requires permission to call other Actors.
• The quality of the report depends on the quality and volume of the input reviews.
• Small datasets may produce less reliable conclusions.
• For best results, use at least 50 reviews per source when possible.

---

Best for

• Restaurants
• Hotels
• Tourist attractions
• Hospitality consultants
• Reputation management agencies
• Local SEO specialists
• Tourism businesses
• Customer feedback analysts

---

Output language

The report can be generated in:

• French
• English
• Spanish