G2 Software Categories & Product Pages

A map of how the B2B software market sorts itself. G2, one of the two dominant software review marketplaces, publishes 2,239 distinct category pages - and no other reachable source describes the software landscape at that granularity. The All Categories list view organizes them under featured groups such as CRM & Marketing, Cloud Computing, ERP & Commerce and HR & Office; the hierarchy view nests 2,238 categories beneath 43 top-level groups - Artificial Intelligence, Sales Tools, Marketing Services, Security, Analytics Tools & Software, Collaboration & Productivity, Development, ERP and Vertical Industry among them - with parent-child relationships expressed directly in the markup.

Each category resolves to a product index meant to carry a composite satisfaction score out of five, verified review counts, feature sub-scores, vendor-listed pricing signals, alternatives and integration lists. Datadory captures the full tree as rows and holds the product-page layer as documented fields confirmed at sample time.

The taxonomy also moves. A Recently Added Categories rail tracks new segments as they launch - Agentic GTM Platforms and MCP Server Infrastructure Platforms both appeared as fresh entries during the August 2026 verification pass - which makes the dataset a running record of where the software market is inventing new shelves.

Three rows exactly as captured during the August 2026 verification pass:

category slug : ai-chatbots
category name : AI Chatbots Software

category slug : large-language-models-llms
category name : Large Language Models (LLMs) Software

category slug : agentic-gtm-platforms
category name : Agentic GTM Platforms Software
parent group  : Artificial Intelligence

Read the third row twice. Agentic GTM Platforms did not exist as a buying category a few years ago; now it sits inside the Artificial Intelligence group with its own page indexing its own vendors. Every row carries that shape - a stable slug usable as a join key, a display name, and a parent group that places it in the 43-group tree - so the whole taxonomy loads as one flat, well-behaved table.

Get a sample of this dataset filtered to your groups or subcategories, and it arrives in this exact shape.

Four fields verify against the taxonomy itself, every definition checked during the August 2026 pass and every example captured from the source:

Three more fields belong to the product pages each category indexes. They are documented in the field layout but were not observable during verification, so Datadory carries them as additional fields on request - confirmed against your actual sample before any scheduled delivery starts rather than promised blind:

• Geography: global vendor scope on the English-language storefront, which is the spine of the dataset. Localized G2 editions exist for other markets and are scoped separately on request.
• Temporal: a continuously refreshed present. The Recently Added Categories rail surfaces new segments within days of launch, and no archival snapshots are published behind it. History is something you accumulate from scheduled deliveries, which Datadory handles.
• Granularity: one category page per software segment - 2,239 total - nested under 43 top-level groups, with each category page indexing its products. Both the segment-level cut and the group-level rollup are native to the structure.

API, files, or your warehouse. Daily, weekly, or hourly.

Samples precede any commitment, and full deliveries land keyed on category slug and parent group, so each pull joins cleanly to the last and to whatever else you already hold.

Ranked by fit recorded in Datadory's persona index:

1. Sales & Growth Teams - map ICP segments and locate competitor sets across the 43 groups before outreach (sales growth teams x application software).
2. Market Researchers & Consultants - frame software-market segmentation on a citable 43-group, 2,239-category structure (market researchers x application software).
3. Competitive Intelligence & Product Teams - watch the Recently Added rail and category membership for rivals repositioning into adjacent segments (competitive intel product teams x application software).
4. Data Scientists & ML Engineers - train SaaS classification and market-mapping models on a human-curated labeling tree four levels deep (data scientists x application software).
5. Investors & Quant Researchers - read taxonomy growth and group density as a coarse, bottom-up signal of where software vendors crowd (investors quants x application software).
6. Journalists, Academics & Students - cite a named taxonomy when describing how the software market organizes itself (journalists academics x application software).
7. E-commerce Operators - shortlist storefront-adjacent SaaS by browsing the relevant subtree instead of the entire market (e-commerce operators x application software).
8. Developers & Data-Product Builders - ship marketplace navigation and category pickers on a tree that already agrees with itself (developers builders x application software).

Capterra software directory is the deeper catalog by row count - 86,267 enumerated product profiles across 1,025 categories with ratings, review counts and starting prices - but its taxonomy stops at category level, while G2 nests 2,238 subcategories under 43 groups. SourceForge software directory covers a different population entirely: 500,000+ open-source projects with download counts rather than commercial buyers rating paid products. Product Hunt launches capture the moment of arrival, not the settled verdict. The common pattern is segmentation from this taxonomy and metrics from a directory - which is precisely why the two land in one warehouse cleanly here. The industry hub lays the whole stack out.

Stated plainly, because they shape the analysis:

• No archive. Coverage is the current state of the taxonomy plus the Recently Added signal. A history of when a category launched means accumulating deliveries over time, which is a scheduling decision, not a research project.
• No public product count. G2 never publishes how many products sit behind the 2,239 categories; the figure commonly cited runs to the hundreds of thousands, but only the category count is verifiable.
• Three product-page fields ride on request. Score, review count and pricing are documented in the product layout but were not observable during the verification pass, so they ship as additional fields confirmed at sample time.
• One storefront spine. Figures reflect the English-language edition; localized markets exist as separate editions and need separate scope.

Because the raw artifact is a pair of very large hand-rendered pages and most questions want a table. Datadory normalizes the tree into the four-field dictionary above, reconstructs the 43-group hierarchy as explicit parent pointers, cuts samples to your segment set, and schedules deliveries so taxonomy change becomes a longitudinal asset instead of a screenshot you re-take. Browse the rest of the shelf on the application software data hub, the best application software datasets ranking, or the full catalog.