ComparEdge's picture
Remove hardcoded counts
2eec534 verified
---
license: mit
task_categories:
- tabular-classification
language:
- en
tags:
- saas
- pricing
- market-analysis
- software-comparison
- sqlite
size_categories:
- n<1K
---
# SaaS Market Intelligence Dataset
331 software products. dozens of categories. Actual pricing data — not estimates.
## Files
| File | Description |
|------|-------------|
| `comparedge_database_schema.sql` | Full DDL. 5 tables, 3 views. Import: `sqlite3 db.sqlite < schema.sql` |
| `SaaS_Pricing_Transparency_Report_Q2_2026.pdf` | Quarterly pricing analysis — category breakdowns, LLM token costs, free tier availability |
## Schema overview
Tables: `categories`, `products`, `pricing_plans`, `features`, `price_history`
Views:
- `v_category_stats` — avg rating, avg price, free tier % per category
- `v_llm_comparison` — LLM products with pricing plan details
- `v_free_tier_products` — all 246 products with perpetual free tiers
## Billing model distribution
Our pipeline classifies each plan into one of four billing models: per-seat (most common in project management and CRM), flat rate (typical for individual tools), usage-based (LLM APIs, cloud hosting), and freemium.
74.3% of tracked products offer a functional free tier. Not a 14-day trial — a permanent plan with real utility.
## Methodology
Pricing scraped from public vendor pages. Where a vendor hides pricing behind "Contact Sales", we record null. Ratings from G2 and Capterra — averaged when both exist, used as-is when only one source is available.
Full SQLite database (864 KB): [Kaggle](https://www.kaggle.com/comparedge/saas-ai-tools-market-2026)