Rakuten Ichiba Reviews Scraper ⭐🇯🇵

Extract individual customer reviews for Rakuten Ichiba (楽天市場) products — the actual review text, star rating, reviewer, date, helpfulness votes and purchase context — as clean, structured JSON. Supply keywords (reviews are pulled for the top matching products) or direct item URLs.

The only way to get Ichiba product-review content on Apify. Other Rakuten actors return listings and aggregate scores; this one returns the reviews themselves — what customers actually wrote.

---

Who is this for?

You are…	You use it to…
🏷️ Brand / product manager or seller	Monitor sentiment, complaints and feature requests on your and rivals' SKUs.
🔬 Market / consumer researcher	Mine Japanese-language opinion at scale across a category or brand.
🤖 AI / NLP team	Build sentiment & summarization datasets from real Japanese reviews.
📈 E-commerce analyst	Correlate rating distribution, verified-purchase share and helpfulness with sales.

---

What you get (output fields)

Field	Type	Description
rating	integer	This review's star rating (1–5).
body	string	Full review text.
title	string	Review theme/title, when present.
nickname	string	Reviewer nickname, when public.
reg_time	string	Submission timestamp (ISO-8601, JST).
verified_purchase	boolean	Whether the reviewer bought the item.
helpful_count	integer	"Helpful" votes.
variant_info	string	Which variant/SKU was purchased.
usage	string	Who/what it was bought for (e.g. 自分用).
frequency	string	Purchase frequency (e.g. はじめて = first time).
sex / age	string	Reviewer demographics, when disclosed.
shop_replied	boolean	Whether the shop replied.
Item context (on every row)
item_id / shop_id	integer	Product & shop ids.
item_name / item_url / item_price / item_image	–	Product identity.
item_review_average / item_review_count	number	Aggregate rating & total.
item_rating_distribution	array	{stars, count} histogram.
item_ai_summary	string	Rakuten AI review summary (optional).
source / scraped_at	string	Provenance & timestamp.

Full definitions in the dataset schema tab.

---

High-value use cases

• Voice-of-customer / sentiment analysis — pull hundreds of reviews per SKU and run NLP to surface complaints, praise and feature requests.
• Competitive product intelligence — compare rating distribution and verified-purchase share across rival products.
• Review-driven merchandising — find which variants (colour/size) buyers love or return, via variant_info + rating.
• Training data — assemble a labelled Japanese review corpus (text + rating).
• Quality monitoring — schedule runs and watch for a drop in new-review ratings or a spike in low-star reviews.

---

Input parameters

Field	Type	Default	Description
searchQueries	array	[]	Keywords; reviews collected for the top products of each.
itemUrls	array	[]	Direct item URLs or shop/item shorthand.
maxProductsPerQuery	integer	20	Products per keyword to review.
maxReviewsPerItem	integer	100	Reviews per product.
maxItems	integer	1000	Global review-row cap.
sort	enum	newest	newest, rating, helpfulness.
minRating	integer	–	Only reviews ≥ this rating (1–5).
buyerOnly	boolean	false	Verified purchasers only.
withImagesOnly	boolean	false	Reviews with photos only.
includeAISummary	boolean	false	Attach Rakuten AI review summary.
proxyConfiguration	object	Apify Proxy	Egress proxy.

Example input

{
  "searchQueries": ["マグカップ"],
  "maxProductsPerQuery": 10,
  "maxReviewsPerItem": 100,
  "sort": "helpfulness",
  "minRating": 1,
  "maxItems": 1000
}

Example output (one row)

{
  "item_id": 10000832,
  "shop_id": 376432,
  "item_name": "ディーンアンドデルーカ モーニングマグ …",
  "item_url": "https://item.rakuten.co.jp/deandeluca/dd0050/",
  "item_review_average": 4.63,
  "item_review_count": 73,
  "rating": 5,
  "body": "アーモンドベージュめちゃくちゃ可愛いお色でした！家族の分も購入しました。",
  "reg_time": "2026-07-26T17:40:06",
  "verified_purchase": true,
  "helpful_count": 0,
  "variant_info": "カラー:アーモンドベージュ",
  "usage": "自分用",
  "frequency": "はじめて",
  "source": "url:https://item.rakuten.co.jp/deandeluca/dd0050/",
  "scraped_at": "2026-08-03T09:20:11Z"
}

---

Scheduling & integrations

• Schedule runs to track new reviews and rating trends over time.
• Webhooks fire on completion — pipe new reviews into your data warehouse.
• Export to JSON, CSV, Excel or Google Sheets; stream to S3 or a database via the Apify API and integrations (Make, Zapier).

Use with AI assistants (MCP)

This Actor is callable as a tool by AI agents through the Apify MCP server, so a Claude/ChatGPT-style assistant can pull Rakuten reviews for a product on demand and summarize sentiment. Point your MCP client at Apify and call this Actor.

---

FAQ

How many reviews can I get per product? As many as the product has — paginated 30 at a time up to your maxReviewsPerItem.

Can I filter reviews? Yes — by minimum rating, verified purchasers only, reviews with images only, and sort by newest / rating / helpfulness.

How do I target specific products? Paste their item URLs, or use keywords to auto-discover the top products and review them.

Do I need a Rakuten account or API key? No. The Actor uses a built-in guest token and Rakuten's public search — no credentials required.

What language are reviews in? Japanese (UTF-8), preserved exactly as written. Pair with any translation/NLP step downstream.

Is scraping this legal? You are responsible for how you use the data. This Actor collects only publicly displayed reviews; review Rakuten's terms and your local regulations (and privacy rules for personal data) before commercial use.

A field is empty for some reviews. Reviewers can leave nickname, age or sex blank — those come through as null. The rating and body are always present.