如何從競爭對手的 Shopify 商店找出客戶投訴(分析 1-3 星評價)
How to Find What Customers Complain About in Competitor Shopify Stores (Mining 1-3 Star Reviews)
分析競爭對手 Shopify 商店的低分評價,可找出產品缺陷、配送及客服等反覆出現的投訴,並整理成可採取行動的主題。Apify 上的 Shopify Reviews Scraper 可偵測商店使用的評價應用程式,並以統一格式匯出評價,再用 Python 篩選 3 星或以下的評價。
The fastest product research you can do is reading a competitor's bad reviews. Five-star reviews tell you what a brand already does well. One- to three-star reviews tell you exactly where buyers felt let down, in their own words, and that is where your product, listing copy or ad angle can win.
This guide shows a simple workflow: export a competitor Shopify store's reviews, keep only the low ratings, and group the complaints into themes you can act on.
Why low-star reviews are worth the effort
- Product gaps: repeated complaints ("runs small", "zipper broke", "smell faded") are a spec sheet for a better version.
- Ad and landing page copy: if buyers of a rival product complain about slow shipping, "ships in 24 hours" becomes your headline.
- Objection handling: the worries people had after buying are the ones your FAQ should answer before they buy.
- Supplier briefs: concrete defect language is easier to hand to a manufacturer than "make it better".
Step 1: Export the competitor's reviews
Most Shopify stores show reviews through a third-party app (Judge.me, Yotpo, Okendo, Loox or Stamped.io), so there is no single export button for a store you don't own. I built the Shopify Reviews Scraper on Apify to handle that: give it store homepages and it detects the review app and returns every review in one schema.
{
"storeUrls": [
{ "url": "https://competitor-one.com" },
{ "url": "https://competitor-two.com" }
],
"maxReviewsPerStore": 2000,
"maxProductsPerStore": 250
}
To focus on a single product line, use productUrls with specific /products/... links instead. When the run finishes, download the dataset as JSON.
Step 2: Keep only the low ratings and find the themes
Each record has rating, title, body, productTitle and date, so filtering is a few lines of Python:
import json, re
from collections import Counter
reviews = json.load(open("dataset.json"))
low = [r for r in reviews if r.get("rating") and r["rating"] <= 3]
print(f"{len(low)} of {len(reviews)} reviews are 3 stars or lower")
# Complaints per product
print(Counter(r.get("productTitle") for r in low).most_common(10))
# Recurring two-word phrases in complaint text
stop = {"the", "a", "and", "it", "i", "to", "was", "is", "of", "my", "for", "this", "but", "not", "in", "on", "that", "they", "with"}
pairs = Counter()
for r in low:
words = [w for w in re.findall(r"[a-z']+", f"{r.get('title') or ''} {r.get('body') or ''}".lower()) if w not in stop]
pairs.update(zip(words, words[1:]))
for (a, b), n in pairs.most_common(25):
print(n, a, b)
Phrases like "too small", "stopped working", "took weeks" or "customer service" jump out quickly. Read 10-20 full reviews behind each top phrase before drawing conclusions. Counts point you to the theme, and the full text tells you what actually went wrong.
Step 3: Turn themes into actions
Make a simple table: theme, how often it appears, example quote, what you'll do about it. For example:
| Theme | What to do |
|---|---|
| Sizing runs small | Add a size chart and "size up" note, or adjust the spec |
| Slow shipping | Lead your ads with your delivery time |
| Broke after a few weeks | Upgrade the weak part and say so on the product page |
| Hard to contact support | Show a real support promise near the buy button |
Step 4: Keep watching
Competitors fix problems and new complaints appear. Schedule the same input monthly in Apify Console (Schedules, cron 0 8 1 * *) and compare the themes over time. If you would rather feed the text to an LLM for summaries, the clean one-schema output works well for that too.
A note on fair use: these are public reviews, so use them for research and don't republish other people's reviews as your own content.
For the full walkthrough of the export itself, see How to Export Judge.me, Yotpo, Okendo, Loox and Stamped Reviews from Any Shopify Store.
Want more data sources? See 13 Free Public Data Sources for Fresh B2B Leads.
來源:DEV Community · dev.to