Croma Product Reviews — Data & Analysis Guide

This actor collects customer reviews for a single Croma product page (identified by the /p/<product_code> segment in the URL), and returns a structured dataset that you can use for analysis.

---

Input

Only two inputs are required:

1. url
   A Croma product page URL that contains /p/<product_code> at the end.

   Examples:

   • https://www.croma.com/.../p/258433
   • https://www.croma.com/.../p/312251
   • https://www.croma.com/.../p/308631

2. max_reviews
   The maximum number of review records to collect before stopping.

   Examples:

   • 50
   • 250
   • 1000

---

Output

The output is a list of review records (a list of dictionaries), where each item represents one review.

Output fields you will typically have per review

A single review record usually includes:

Reviewer & review text

• username — reviewer alias/name
• comment — review text
• date — review date (YYYY-MM-DD)
• rating — numeric rating (e.g., 1–5)
• verifiedPurchase — whether purchase is verified

Purchase location (if provided)

• purchase_city_name
• purchase_state
• purchase_country

Product-level context (same value repeated for each review row)

• product_code
• averageRating
• review_text_count (or similar count field from the API, if available)

Rating distribution snapshot (same value repeated for each review row)

• 1 star_count
• 2 star_count
• 3 star_count
• 4 star_count
• 5 star_count

---

Example

Example Input

• url: https://www.croma.com/marshall-emberton-ii-20w-portable-bluetooth-speaker-ip67-water-resistant-multi-directional-control-knob-stereo-channel-black-brass-/p/258433
• max_reviews: 3

Example Output (illustrative)

You would receive a list containing up to 3 review records. Each record looks like:

• username: Surjit
• comment: Have had a harrowing experience with my online order…
• date: 2023-09-14
• rating: 1
• verifiedPurchase: True
• purchase_city_name: MUMBAI
• purchase_state: MAHARASHTRA
• purchase_country: IN
• product_code: 258433
• averageRating: 4.3
• 1 star_count: 2
• 2 star_count: 0
• 3 star_count: 0
• 4 star_count: 1
• 5 star_count: 14

---

Analyses you can do with this dataset

1) Rating distribution and quality score

• Star distribution: % of 1-star vs 5-star reviews
• Weighted average rating: confirm vs reported averageRating
• Top-box / bottom-box: (4–5 stars) vs (1–2 stars)

Why it matters:

• Quickly measures product satisfaction and whether negative reviews are concentrated.

---

2) Review volume trends over time

• Reviews per week/month
• Growth rate in review volume
• Spikes after promotions, releases, or price changes (if you track dates)

Why it matters:

• Helps correlate campaigns or launches with customer feedback and adoption.

---

3) Verified purchase vs non-verified comparisons

• Average rating: verified vs non-verified
• Negative review share: verified vs non-verified
• Theme differences in text (complaints vs praise)

Why it matters:

• Verified reviews often better represent actual product experience.

---

4) Geography-based insights (if location data exists)

• Ratings by city/state
• Concentration of negative feedback by region
• Outlier locations with unusually high complaint rates

Why it matters:

• Can reveal supply chain, store-level, or service issues linked to geography.

---

Reporting ideas (what to present)

You can summarize results in a simple dashboard/report:

• Total reviews collected (up to max_reviews)
• Average rating and star distribution
• Monthly review count trend
• Top negative themes (Top 5)
• Verified vs non-verified comparison
• Geography summary (if present)

---

Notes & good practices

• Some products may have thousands of reviews; use max_reviews to control time and data size.
• Not every review contains purchase location; treat missing values as normal.
• Use the data responsibly and comply with site terms and applicable policies.

---

Summary

With url + max_reviews, you can collect structured Croma review data and produce:

• Satisfaction metrics (ratings)
• Trends over time
• Theme-based insights from text
• Verified purchase quality checks
• Regional breakdowns (where available)