How to scrape Yelp reviews
The catch most people hit first: Yelp's Fusion API returns only up to three review excerpts per business, and each one is truncated to a preview, not the full text. For customer research that is close to useless, you get a sentence fragment from three reviews when the page shows hundreds. And Yelp is aggressive about blocking scrapers that hit the page directly, so the naive script route runs into rate limits and challenges fast.
That leaves a few realistic paths, from copying reviews by hand to reading the public page through your own browser. Here is how each one holds up, and the fastest way to turn a business page into usable customer language.
★ 4.5 on the Chrome Web Store · no card needed
Manual copying: fine for a handful, painful past that
Open the business page, sort by rating, and copy what you need. For a quick read on one competitor this works. It falls apart when you want the whole picture: Yelp paginates reviews, hides some behind a 'not recommended' filter, and the language you are hunting for is spread across pages. You end up reading the same top reviews everyone reads and missing the pattern deeper in the list.
The Fusion API: three snippets and a dead end
Yelp's official Fusion API is built for showing business info and star ratings inside apps, not for research. It caps review data at three excerpts per business, and those excerpts are truncated previews rather than full reviews. There is no endpoint that returns the complete review body for a business you do not own. If your question is 'what are all these customers actually saying', the API cannot answer it, and writing your own scraper against Yelp's page invites blocks and CAPTCHAs.
The paste-a-URL route
Adlicio reads the reviews on a public Yelp business page through your own browser session, the same content you see as a visitor, with no API key. Paste the business URL and it captures the reviews in about 60 seconds.
Then it does the part that matters: the reviews come back clustered into ranked angles, the recurring praise, the complaints, and the objections, each with the verbatim quotes behind it. Local-service reviews are especially rich because they carry hire or no-hire decision language, the exact reasons someone chose this business or walked away, which is the raw material your next ad is built from.
The 60-second version
- 01Pick the business pages your buyers weigh
A competitor you lose customers to, the category leader in your market, and your own listing. Their reviews hold different halves of the story.
- 02Paste each business URL into Adlicio
The scrape reads the public review pages and finishes in about 60 seconds per business.
- 03Work the clusters, keep the quotes
The output ranks the recurring reactions into angles with hook lines. A repeated complaint becomes your promise, the verbatim quote becomes your proof line.
Questions people also ask
Why does the Yelp API only give me three reviews?+
The Fusion API is designed to surface business details and ratings, not full review corpora, so it returns up to three truncated excerpts per business. There is no official endpoint for the complete review text, which is why any real research reads the public page instead.
Will scraping Yelp get me blocked?+
Yelp blocks automated traffic that hammers its pages. Adlicio avoids that by reading through your own browser session at human pace, the same way a visitor loads the page, rather than firing a bot at the site.
What makes local-service reviews useful for ads?+
Reviews for plumbers, clinics, contractors, and the like are full of hire or no-hire decision language: why someone picked this business, what almost stopped them, what sealed it. That is direct objection and desire material for your copy.
Can I export the Yelp reviews I collect?+
Yes. Every scrape lands in your history with a CSV export, so you can filter the raw reviews alongside the ranked angles.
More guides
Yelp scraperAll scrapersSee pricing
Run this play on your own Yelp page.
Paste one public URL. Adlicio returns the angle, hook, and proof to test next.