Research a competitive set
Compare products within a defined category. Keep ratings, review volumes and observation dates together for like-for-like research.
All use casesG2 software
SOFTWARE RESEARCH & REVIEW INTELLIGENCE
G2 software
review data.
Read the rating.
Keep the context.
Product profiles, review evidence and category context, structured for clearer software-market research.
FROM SCORES TO BETTER QUESTIONS
Compare software.
Understand the feedback.
Connect the product, its review evidence and the market it belongs to. Start with the research question, then define the data you need.
Investigate customer feedback
Organize permitted review content by product and segment to examine reported strengths, friction and unmet needs.
Track reputation over time
Follow changes in published ratings and review counts with dated snapshots, without mistaking one page for a complete history.
THREE CONNECTED RECORD TYPES
The product, the review
and where it was found.
A useful dataset separates aggregate signals from individual feedback. Source identifiers connect the records without flattening their meaning.
A product with a clear identity.
Connect the product, category and aggregate rating. Keep a vendor or service provider distinct from the software it sells.
- product_id / entity_type
- Source identifier and the kind of entity being compared.
- name / categories
- Product label and published category memberships.
- aggregate_rating / review_count
- The score, its scale and published review volume.
- source_url / observed_at
- The source reference and the time of collection.
Illustrative schema / not a collected G2 result
{
"example_only": true,
"record_type": "product",
"sample_id": "example-crm",
"product_id": null,
"source_url": null,
"observed_at": null,
"name": "Example CRM",
"entity_type": "software",
"categories": [
"CRM"
],
"aggregate_rating": {
"value": 4.4,
"scale": 5
},
"review_count": 240,
"reviews_collected": 0,
"review_text_included": false
}Fictional records and values. No live data request.
Fictional product and values. The aggregate count is not the number of reviews delivered; individual review access is scoped separately.
Illustrative delivery schemas, not collected G2 records or a built-in API response. Field availability, review access and permitted use are confirmed during scoping.
DISTINCTIONS THAT PROTECT THE ANALYSIS
A score alone
does not tell the story.
Normalize the format while preserving what the source actually says, how it was selected and what remains unknown.
Keep the denominator
A 4.5 rating from 12 reviews and the same rating from 1,200 are different observations. Preserve the review count and rating scale.
Separate source from interpretation
Keep original feedback, published incentive or verification labels, and derived themes distinct. Do not infer missing labels or reviewer identities.
Position is not a permanent rank
Store filters and sorting alongside the result. A category position is not automatically a G2 Grid score, market share or an endorsement.
BUILD THE COLLECTION AROUND THE QUESTION
Match once.
Refresh what matters.
Validate source access and a small sample before scaling. Avoid repeated product discovery and unnecessary processing.
- 01 / SCOPE
Define coverage and access
Agree products, categories, review period and intended use. Confirm a licensed API, permitted export or other authorized collection route.
- 02 / MATCH
Connect products and reviews
Resolve source identifiers, preserve entity types and deduplicate records. Keep missing values and failed requests explicit.
- 03 / MAINTAIN
Deliver useful updates
Validate counts and schema, retain observation dates and agree a refresh schedule. Scope optional analysis only where usage rights allow it.
Start with the G2 API
G2 documents product, category and review access with endpoint permissions. Licensed structured data can avoid repeated page rendering and reduce parsing work.
G2 API documentationParse before adding processing
When page access is appropriate, use available structured listing data before adding rendering or AI. Validate the content: aggregate-rating markup is not a collection of individual customer reviews.
Scraper API referenceG2 restricts automated collection and reuse without authorization. We review source terms, access and intended processing before agreeing a project. No login-protected reviews, private profiles or buyer-intent data are included by default.
THE RIGHT SERVICE FOR YOUR WORKLOAD
Research across websites.
Or build a review dataset.
Our self-service API handles your page workflow. A managed project adds source scoping, product matching and a maintained delivery schema.
Scraping API
Collect accessible public pages across permitted sources, with optional AI processing.
- Content-only collection without AI
- Your URLs, parsing and result handling
- No login or CAPTCHA bypass
Managed G2 data
A tailored product and review dataset, subject to appropriate access and agreed usage rights.
- Access and sample validation
- Product matching and quality checks
- Scoped JSON, CSV or API delivery
BEFORE YOU SCALE
Review coverage.
Clear expectations.
Does a category page contain all the underlying reviews?
No. A category listing can show names, aggregate ratings and segment information without including individual reviews. We scope category discovery, product detail and review collection separately, and report the records actually delivered rather than using the published review count as a collected total.
Can software products and service providers be compared together?
They need separate entity types. For example, a consulting-partner category describes providers, while a CRM category describes software. We retain the category and entity context so a provider listing is not silently treated as a software product or a review.
How do you handle rating scales and company segments?
Each rating keeps its scale, and each segment keeps its source label. Aggregate ratings, individual review ratings and other scores remain separate. Missing percentages stay unknown; inconsistent segment totals are flagged rather than silently adjusted to make them add up to 100%.
Will we receive every review and the original review text?
Coverage depends on your authorized access, date range, filters and agreement with G2. Full review text, distribution rights and reviewer metadata are not included automatically. G2 documents a separate review-syndication access route for partners. We confirm the permitted fields and use before promising a review archive.
How are edits, new reviews and historical trends handled?
Match observations using product and review identifiers. Keep publication, update and collection times separate, and compare successfully collected records. A historical series needs an agreed baseline and collection window; a current page is not a complete history. A missing or failed page is not proof that a review was deleted.
Can you classify feedback or summarize recurring issues?
Where the source agreement permits that processing, analysis can be scoped separately from collection. Derived themes retain links to their supporting records and remain distinct from the original review text. A model-generated summary is not a verified customer fact, and reviewer feedback is not a buyer-intent signal.
What affects cost, refresh frequency and delivery?
Product coverage, review volume, access or licensing costs, fields and refresh cadence determine the workload. Deduplicating products and requesting only needed changes can reduce repeated work. JSON preserves nested feedback; CSV can use linked product and review tables. We quote after validating a sample and the access route, not from an assumed price per page.
START WITH YOUR RESEARCH QUESTION
Which software market
are you studying?
Share the products, categories and review fields you need. We will review the access route and a sample scope before proposing delivery.
- Product URLs or target categories
- Review period, segments and fields
- Existing data access and intended use
Projects are accepted individually, subject to feasibility and available capacity. An enquiry does not reserve a solution or confirm delivery. Pricing and terms are agreed directly with our team.
We review access and scope first. This form does not run a scrape or create a paid task.
ScrapingBuddy is not affiliated with G2. The research image and dataset examples are illustrative, not actual customer reviews. Collection, analysis and delivery depend on appropriate source access and agreed usage rights.