SaaS pricing barely moves

Why you will see a much bigger number elsewhere

The median rise here is 22.2%, and quoting that alone would be the easiest headline on this page. It would also be wrong. It is a median taken across the 522 plans that rose — a little over 0.9% of the catalogue. The other 98.2% of plans are not in it, and neither are the 631 that were cut.

Small vendors move in whole dollars off a small base, so a plan going from $5 to $7 books a 40% rise on $2. That is why the conditional medians are large in both directions and nearly symmetrical, and why the only figure that describes the market is the median across every change: -3.8%.

Method, and what it will not tell you

• The denominator is plans we actually re-read. A plan whose product was not scraped in the window could not have been seen to move, so it is excluded rather than counted as unchanged.
• A discount switching on or off is not a price change. An annual discount makes the monthly figure a fixed multiple of the annual one, so reading one and then the other records a move that never happened. Those are excluded by ratio; before they were, they were about a third of everything here.
• Moves we later corrected are excluded. If we reread a page and revised our own extraction, the difference was ours, not the vendor's, and it does not appear.
• There is no year-on-year figure. Change records begin in May 2026 and only reach useful density in July, so this is a snapshot of one window, not a trend. Widening the window past 90 days returns almost the same counts because there is no earlier history to add.
• The catalogue is mostly small software. It is built from launch sites and indie directories, so it is not a read on enterprise contract pricing, which is negotiated and never published.