DEV Community

Cover image for Mine competitors' Google Play reviews for feature ideas (Python, no API key)
Fjord Labs
Fjord Labs

Posted on

Mine competitors' Google Play reviews for feature ideas (Python, no API key)

Your competitors' 1–3 star reviews are a free list of what users want and don't get. The problem is collecting them: the Play Store has no public reviews API. Here's how I pull them for a whole category in one call and rank the complaints.

Find the apps and fetch the reviews

I use a hosted Actor I maintain on Apify (Google Play Reviews Scraper). It accepts app IDs or search terms, so you don't have to look up IDs by hand. With pip install apify-client pandas:

from apify_client import ApifyClient
import pandas as pd

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("fjordlabs/google-play-reviews").call(run_input={
    "searchTerms": ["habit tracker"],
    "appsPerSearch": 5,
    "countries": ["us", "gb"],
    "languages": ["en"],
    "maxReviewsPerApp": 300,
    "sort": "newest",
    "minStars": 1,
    "maxStars": 3,
    "reviewsNewerThan": "90d",
    "includeAppDetails": False,
})

rows = list(client.dataset(run["defaultDatasetId"]).iterate_items())
df = pd.DataFrame([r for r in rows if r.get("type") == "review"])
Enter fullscreen mode Exit fullscreen mode

Each review row has appId, score, content, thumbsUpCount, appVersion, at and the developer's reply, if any. Reviewer names are off by default (includeReviewerNames), which keeps you clear of collecting personal data you don't need.

Rank the complaints

Count how often a few topic words show up, per app:

topics = ["sync", "widget", "subscription", "ads", "crash", "dark mode", "export"]
text = df["content"].fillna("").str.lower()
for t in topics:
    df[t] = text.str.contains(t)

print(df.groupby("appId")[topics].sum())
Enter fullscreen mode Exit fullscreen mode

Weight by thumbsUpCount if you want the complaints other users agree with. If "export" or "widget" tops the table for three of five competitors, that's a gap to fill.

Cost

Reviews are $0.0001 each, so 5 apps × 300 reviews is about 15 cents plus a tiny start fee. Use reviewsNewerThan to fetch only new reviews on repeat runs, and run it monthly on a schedule.

Disclosure: I built and maintain the scraper. Questions and feature requests welcome in the comments. Links to Apify include my referral code.

Top comments (0)