Scrape G2 Reviews for B2B SaaS Research (2026)

Published April 28, 2026 · 1614 words · For researchers

Thirdwatch's G2 Software Reviews Scraper returns B2B SaaS reviews + ratings + comparison data at $0.008 per record — review text, pros/cons, reviewer demographics, product metadata, comparison pages. Built for SaaS competitive-research, content-strategy teams, M&A diligence analysts, and B2B-buyer-intent platforms.

Why scrape G2 for B2B SaaS research

G2 is the canonical B2B SaaS review platform. According to G2's 2024 Software Buyer Behavior report, 86% of B2B buyers reference G2 during purchase decisions, and category-leader rankings on G2 correlate with 30-50% higher organic-traffic conversion than non-G2-ranked competitors. For SaaS competitive-research, positioning analysis, and content-strategy work, G2's data depth is materially better than Capterra or Trustpilot.

The job-to-be-done is structured. A SaaS competitive-research function maps category-level positioning across 50 SaaS verticals quarterly. A content-strategy team builds "X vs Y" comparison pages from G2 comparison-page data. An M&A analyst tracks target SaaS rating trajectory + reviewer-sentiment shifts. A B2B-buyer-intent platform surfaces "in-market" signals to vendor users. All reduce to category + product queries + per-review aggregation + per-product comparison data.

How does this compare to the alternatives?

Three options for G2 review data:

G2's first-party product offers authoritative reports for enterprise; the price gates out small/mid-market consumers. The G2 Scraper actor page gives you raw review-level data at the lowest unit cost.

How to scrape G2 in 4 steps

Step 1: How do I authenticate against Apify?

Sign in at apify.com (free tier, no credit card), open Settings → Integrations, and copy your personal API token:

export APIFY_TOKEN="apify_api_xxxxxxxxxxxxxxxx"

Step 2: How do I pull category-level reviews?

Pass G2 product slugs.

import os, requests, pandas as pd

ACTOR = "thirdwatch~g2-software-reviews-scraper"
TOKEN = os.environ["APIFY_TOKEN"]

PRODUCTS = ["stripe", "adyen-payments", "checkout-com",
            "salesforce", "hubspot", "pipedrive",
            "slack", "microsoft-teams", "zoom",
            "notion", "asana", "linear-app"]

resp = requests.post(
    f"https://api.apify.com/v2/acts/{ACTOR}/run-sync-get-dataset-items",
    params={"token": TOKEN},
    json={"queries": PRODUCTS, "maxResults": 100},
    timeout=3600,
)
df = pd.DataFrame(resp.json())
print(f"{len(df)} reviews across {df.product_name.nunique()} products")

12 products × 100 reviews = up to 1,200 records, costing $9.60.

Step 3: How do I extract pros/cons themes per product?

Cluster pros/cons text per product for theme discovery.

from collections import Counter
import re

def extract_themes(texts, min_word_len=4):
    counter = Counter()
    for t in texts:
        if not isinstance(t, str): continue
        words = re.findall(r"\b[a-z]{" + str(min_word_len) + ",}\\b", t.lower())
        counter.update(words)
    return counter

themes_per_product = {}
for product, grp in df.groupby("product_name"):
    pros_themes = extract_themes(grp.pros.dropna())
    cons_themes = extract_themes(grp.cons.dropna())
    themes_per_product[product] = {
        "top_pros": pros_themes.most_common(10),
        "top_cons": cons_themes.most_common(10),
    }

for p, t in themes_per_product.items():
    print(f"\n=== {p} ===")
    print(f"  PROS: {[w for w, _ in t['top_pros'][:5]]}")
    print(f"  CONS: {[w for w, _ in t['top_cons'][:5]]}")

Top pros/cons themes reveal what each product wins on (and what it loses on) at the category-leader level — direct input for positioning + content-strategy.

Step 4: How do I detect rating drift over time?

Persist (product, snapshot_date, rating, review_count) tuples for trajectory analysis.

import datetime, json, pathlib

ts = datetime.datetime.utcnow().strftime("%Y%m%d")
per_product = (
    df.groupby("product_name")
    .agg(
        avg_rating=("rating", "mean"),
        review_count=("review_id", "count"),
    )
    .reset_index()
    .assign(snapshot_date=ts)
)
out = pathlib.Path(f"snapshots/g2-{ts}.json")
out.parent.mkdir(parents=True, exist_ok=True)
out.write_text(per_product.to_json(orient="records"))
print(f"Persisted {len(per_product)} per-product aggregates")

Weekly snapshots build the time-series needed for rating-trajectory dashboards + M&A diligence.

Sample output

A single G2 review record looks like this. Five rows weigh ~10 KB.

{
  "review_id": "12345",
  "product_name": "Stripe",
  "product_slug": "stripe",
  "category": "Online Payment Processing",
  "rating": 4.6,
  "title": "Best payment infrastructure for SaaS",
  "review_text": "We've used Stripe for 4 years across 3 startups...",
  "pros": "Excellent API documentation, predictable pricing, strong fraud detection",
  "cons": "Pricing climbs aggressively at scale, support response can lag",
  "recommendations": "Best for startups + mid-market with technical teams",
  "reviewer_role": "Senior Engineer",
  "reviewer_industry": "Internet & Software Services",
  "reviewer_company_size": "201-500 employees",
  "review_date": "2026-04-12",
  "verified_buyer": true
}

pros and cons are the killer fields for theme extraction. reviewer_industry + reviewer_company_size enable segmentation analysis (does Stripe rate higher with mid-market vs enterprise reviewers?). verified_buyer: true filters out unverified reviews — important for trust-weighted analysis.

Common pitfalls

Three things go wrong in G2 scraping pipelines. Verified-vs-unverified mixing — about 70-80% of G2 reviews are verified buyers; the rest are unverified. For high-trust analysis, filter to verified_buyer: true exclusively. Category-leader bias — top-ranked products in any G2 category have disproportionately positive reviews because high-rated products attract more reviews. For balanced research, sample reviews proportionally by review-count rather than just top-N. Review-bombing detection — G2 occasionally has coordinated negative-review attacks (typically post-pricing-change, post-acquisition); cross-check rating drops against new-reviews velocity to distinguish genuine sentiment shifts from attacks.

Thirdwatch's actor uses Camoufox + residential proxy at $5/1K, ~36% margin. The 4096 MB memory and 3,600-second timeout headroom mean even 5,000-review batches complete cleanly. Pair G2 with Capterra Scraper for SMB-tier coverage (G2 skews mid-market+) and Trustpilot for B2C-overlap visibility on consumer-facing SaaS. A fourth subtle issue worth flagging: G2's review-recency weighting in headline ratings means very recent rating shifts (last 30 days) have disproportionate weight; for stable trajectory analysis, compute trailing-90-day average rating separately from the headline G2-displayed rating. A fifth pattern unique to G2: many SaaS products have multiple G2 entries (separate listings per region or per product-tier — e.g., "Salesforce" + "Salesforce Sales Cloud" + "Salesforce Service Cloud"); for company-level analysis, manually map related G2 entries to a canonical company-name and aggregate reviews. A sixth and final pitfall: G2's "Grid" category-leader visualizations are based on G2-internal scoring formulas that aren't exposed in the actor's data; for replicating Grid positioning, you need G2's first-party data product. A seventh and final pattern worth flagging for production teams: data-pipeline cost optimization. The actor's pricing scales linearly with record volume, so for high-cadence operations (hourly polling on large watchlists), the dominant cost driver is the size of the watchlist rather than the per-record fee. For cost-disciplined teams, tier the watchlist (Tier 1 hourly, Tier 2 daily, Tier 3 weekly) rather than running everything at the highest cadence — typical 60-80% cost reduction with minimal signal loss. Combine tiered cadence with explicit dedup keys and incremental snapshot diffing to keep storage and downstream-compute proportional to new signal rather than total watchlist size.

An eighth subtle issue: snapshot-storage strategy materially affects long-term economics. Raw JSON snapshots compressed with gzip typically run 4-8x smaller than uncompressed; for multi-year retention, always compress at write-time. Partition storage by date prefix (snapshots/YYYY/MM/DD/) to enable fast date-range queries and incremental processing rather than full-scan re-aggregation. Most production pipelines keep 90 days of raw snapshots at full fidelity + 12 months of derived per-record aggregates + indefinite retention of derived metric time-series — three retention tiers managed separately.

A ninth pattern unique to research-grade data work: schema validation should run continuously, not just at pipeline build-time. Run a daily validation suite that asserts each scraper returns the expected core fields with non-null rates above 80% (for required fields) and 50% (for optional). Alert on schema breakage same-day so consumers don't degrade silently. Most schema drift on third-party platforms shows up as one or two missing fields rather than total breakage; catch it early.

Related use cases

• Build G2 vs competitor comparison pages
• Track G2 rating changes for SaaS companies
• Scrape Capterra reviews for SMB SaaS research
• The complete guide to scraping reviews
• All Thirdwatch use-case guides

Frequently asked questions

Why G2 specifically for B2B SaaS research?

G2 is the canonical B2B SaaS review platform — 100K+ products across 2,000+ categories with verified-buyer reviews, ratings, and detailed pros/cons analysis. According to G2's 2024 Buyer Behavior report, 86% of B2B buyers reference G2 during purchase decisions. For SaaS competitive research, positioning analysis, and content strategy, G2 is materially deeper than Capterra or Trustpilot for technical buyers.

What review fields does the actor return?

Per review: rating (1-5 stars), title, review text, reviewer industry/role/company-size, pros, cons, recommendations, date. Per product: aggregate rating, review count, market segments served, integrations list, pricing range, comparison data. About 70-80% of G2 products with 50+ reviews have comprehensive metadata; smaller products may have sparse fields.

Can I scrape comparison pages (X vs Y)?

Yes. G2's comparison pages (e.g., Salesforce vs HubSpot) aggregate per-product ratings + feature comparisons. The actor pulls these pages directly. For competitive-positioning research, comparison-page scraping is the highest-signal source — features that differ between products + how reviewers rate each on those features.

How fresh do G2 reviews need to be?

For brand-monitoring and rating-drift detection, weekly cadence catches signals. For competitive-research and positioning, monthly is sufficient. For longitudinal tracking (12-month rating trajectory), quarterly snapshots produce stable trend data. Most G2 products see 5-50 new reviews per month; daily cadence is over-frequent for most use cases.

How does G2 handle anti-bot defenses?

G2 uses DataDome aggressively. Thirdwatch's actor uses Camoufox stealth-browser + residential proxy. Production-tested with 95%+ success rate. Failed queries trigger SessionError which Crawlee auto-retries with fresh proxy. Rate-limit guidance: stay under 100 product-page fetches per hour per proxy IP for sustained reliability.

How does this compare to G2's first-party data product?

G2 sells aggregate reports + Tableau/Looker connectors at enterprise price points ($25K-$200K/year). Their data is authoritative but the price gates out small/mid-market consumers. The actor delivers raw review-level data at $8/1K — for cost-optimized SaaS competitive research, the actor is materially cheaper. For boardroom-grade reporting on G2-curated category leaders, G2's first-party product wins.

Run the G2 Scraper on Apify Store — pay-per-record, free to try, no credit card to test.