PYRAMYD Field Report · October 2026
Five findings drawn from the PYRAMYD Product Graph — 250K+ products across 2,546 categories, 2.4M+ aggregated reviews, and 2,000+ signal sources. Every figure comes from a read-only query run on 8 October 2026.
The headline finding
The enterprise software market is more crowded, more concentrated and faster-moving than most category maps show. Buyers write about a few products, most rivals go unreviewed, and the vendors you track change something every few days.
Five findings
Finding 01
34
vendors in the median category
Across the 2,546 enterprise software categories in the graph, the median category has 34 vendors with a product in it. One in twenty has more than 300, and 271 categories have 200 or more. A category map with a handful of named rivals describes the top of the market, not the market.
Finding 02
89.7%
of review volume on the top 10% of products
The top 10% of products by review count hold 89.7% of all review volume. Of the 106,868 products with any reviews, 97,288 have fewer than 100. Competitive intelligence that leans on reviews alone is thinnest exactly where most competitors live.
Finding 03
24%
of categories are vertical industry software
620 of the 2,546 categories (24%) are vertical industry software. They hold 40,684 products, 16.5% of the products in a category, but only 8.4% of review volume. If you compete in healthcare, financial services, legal or construction software, your rivals are less reviewed and harder to see.
Finding 04
75×
more reviews per vendor for 4+ product vendors than single-product ones
Vendors with four or more products average 1,198 reviews; single-product vendors average 16. Only 4,839 of 181,515 vendors sell four or more products, yet they are the ones buyers write about, so tracking a rival means tracking its whole portfolio, not one product page.
Finding 05
3 days
median gap between material signals for top vendors
For the 30 of the top 100 vendors (by review volume) with tracked product, pricing, partnership, leadership, release or integration signals, the median gap between one such signal and the next is 3 days, across 744 gaps. A battlecard refreshed once a quarter is out of date for most of its life.
Methodology
Every figure traces back to a read-only query against the Product Graph, run on 8 October 2026. The SQL is available on request.
Book a 30-minute demo. We'll filter the same five queries to your specific category and walk through what changes when CI is grounded in a live graph instead of a quarterly slide deck.