How to Analyse Your Competitors' Google Reviews to Find Their Weak Spots

Short answer: collect the Google Maps URLs of your top 5 to 10 competitors, scrape all their reviews with Google Maps Reviews Scraper, then run every review through an AI prompt that tags its topic and sentiment with the Bulk LLM Runner. Count the negative tags per competitor, and you have a ranked list of what customers dislike about them. Scraping costs $0.0006 per review, so 5,000 reviews cost $3.

Your competitors' reviews are the most honest market research you will ever get for free. Their unhappy customers tell you, in their own words, what they wanted and did not get: slow replies, hidden fees, rude staff, missed appointments, parking. Every repeated complaint is a promise you can make in your ads, on your landing page and in your Google Business Profile. The work is reading thousands of reviews, and that is exactly the part to automate.

Google Maps Reviews Scraper is built by Compass, not by me. It is one of the busiest actors on Apify, with more than 160 million runs, and it is on my list of the best Apify actors for local business data.

What data do you get from each Google review?

• Star rating and the review text, plus an automatic translation for reviews in other languages
• Published date
• The owner's response and when it was posted
• Detailed ratings per aspect where Google shows them, such as food, service and atmosphere
• Review context tags, attached photos, review ID and URL, and likes on the review
• Reviewer name, number of reviews they have written and Local Guide status

The owner response fields are worth a separate look: how often a competitor answers negative reviews, and how fast, tells you a lot about how they run the business.

How do you analyse competitor reviews step by step?

1. List your competitors

Search Google Maps the way your customers do ("dentist near me", "plumber in Leeds") and take the top 5 to 10 places from the map pack. Copy each place URL. If you have many locations or want the full list of businesses in an area, the Google Maps Scraper returns place URLs and IDs for a search.

2. Scrape the reviews

Paste the place URLs into Google Maps Reviews Scraper. Set Sort reviews by to newest and add a start date, for example the last 12 months, so you analyse the business as it is now, not as it was five years ago. For hotels, set Reviews origin to Google to get full coverage of Google-native reviews.

3. Tag every review with AI

Export the reviews and upload the file to the Bulk LLM Runner with a prompt like:

Review ({{stars}} stars): {{text}}
Return: sentiment (positive, negative or mixed), main_topic
(one of: price, wait time, staff attitude, quality of work,
communication, booking, cleanliness, location or parking, other),
and complaint (one short sentence, or none).

Fixing the topic list in the prompt matters. It makes every row use the same labels, so you can count them. The actor adds sentiment, main_topic and complaint as new columns next to each review.

4. Count and compare

Make a pivot table: competitors down the side, topics across the top, counting negative reviews. Patterns jump out immediately. One competitor has a wait time problem, another has a pricing transparency problem, and all of them are weak on communication.

5. Turn the weak spots into your message

If the biggest complaint across your market is "never called back", then "we reply within one hour" is your headline, and it needs to be true. Use the exact words customers use in reviews in your ad copy and landing pages: those are the words people search with.

How much does it cost?

• Reviews: $0.0006 each. 10 competitors x 500 reviews = 5,000 reviews, about $3.00.
• AI tagging: $0.002 per row in the Bulk LLM Runner plus the model's token cost. With a small fast model, tagging short reviews is cheap; 5,000 rows is $10 in row fees plus model usage. Test on 50 rows to see the exact cost.

Apify's free plan includes $5 of monthly credit, enough to scrape around 8,000 reviews. Note that the Bulk LLM Runner needs a paid Apify plan for anything beyond a 3-row test.

What else can you learn from review data?

• Review velocity: how many reviews each competitor gets per month. A sudden jump often means a review campaign, and possibly fake reviews.
• Response rate: the share of negative reviews that got an owner response. Low response rates are an easy way to look better by comparison.
• What people love: positive tags show what customers value most, which is what your service must not fall short on.

Reviews are one side of local search. The other is where you actually rank on the map in different parts of town, which the Geo-Grid Local Rank Tracker measures point by point.

Is it legal to scrape Google reviews?

Reviews are public, and analysing them in aggregate for competitive research is common practice. Reviewer names and profile links are personal data under GDPR and similar laws, so drop those columns if you do not need them, and never contact reviewers or republish their reviews as if they were about you. Google's terms restrict automated collection, so use the data internally for analysis. This article is independent of Google.

Frequently asked questions

More local and SEO tools are in the SEO and AI visibility hub.