The figures are out there
Articles that put a number on "SaaS prices go up" are not hard to find. Some break it down by category.
PricePulse, a service that tracks SaaS pricing, publishes this:
Design & Dev Tools +57% avg
Canva and GitHub Copilot both effectively doubled their prices. Figma raised by 67%.
And for another category:
Airtable +100% / Linear +60% / ClickUp +58% / Monday +30% / Notion +25% / Asana +23% / Miro +20%
The article says it draws on "54 verified entries".
No individual figure carries a date. None links to a pricing page. The article body contains no outbound links at all. What "verified" means is not stated.The two that could be checked did not reproduce
This cat saves Figma's and Linear's pricing pages. Here is what they say.
Figma, "+67%"
Between June and July 2024, Figma's Professional went from displaying $12 to displaying $15. That looks like a 25% rise.
It was not a rise. The earlier page carries this footnote:
\*Professional also available for $15 /seat, billed month-to-month.
The $12 was the annually-billed monthly rate; month to month it was $15. The later page says "Save 20% when billed annually", and 20% off $15 is $12. Neither axis moved. (full working)
Figma moved to seat types in 2025, but plans and seats are different units and cannot be compared.
There is no way to get +67% out of the archived pricing pages.Linear, "+60%"
The upper plan went from $12 in early 2022 to $16 in November 2025 — +33%. The lower plan went $8 to $10 — +25%. Both are the annual-billing display. (four years of it)
There is one combination that gives 60%. Today's lower plan ($10) against today's upper plan ($16) is +60%. But that is the gap between two tiers, not a price rise.That is a guess. The method behind the published figure is not stated, and not knowing is the point of this article.
The claim is not that they are wrong
To be clear: this is not an argument that those figures are incorrect.
It is that they cannot be checked.Without dates, there is no way to know which repricing is meant. Without plan names, no way to know which tier was compared. Without a billing cycle, no way to know whether it is monthly or annual. Without a source URL, no way to open the page.
This cat has made the same mistake eight times. Every one was the same shape: the figure read correctly, and what the figure was a price of read wrongly. With Oh Dear it reached publication — +95-123%, comparing an annually-billed monthly rate against a monthly-billed one. The correct figures are +78-98%.
That correction was possible because the sources were printed. The archived URLs were in the article, so the working could be reopened.A figure with no source cannot be corrected when it is wrong, and cannot be confirmed when it is right.
What an average cannot contain
A category average has a structural blind spot.
Of nine repricings checked one at a time here, four move in a direction you cannot read off the price column.
| Product | Price column | What happened |
|---|---|---|
| Plausible | unchanged | $9 went from ten sites to one. Keeping ten costs +111% |
| Postmark | −73% | Included volume fell to a fifth; cost per 1,000 messages rose 36-63% |
| Netlify | free stayed free | Units became credits; bandwidth ceiling 100GB → about 30GB |
| Healthchecks.io | unchanged | Every plan's limit left the page. Lifted or unpublished cannot be told |
A repricing that does not move the price column does not enter a price-increase average. Plausible is counted as a company that did not raise prices. Its customers' bills doubled for the same usage.
The same applies downward. Obsidian Sync halved. Average "the companies that raised prices" and that row was never eligible.
The averages are not wrong. The question an average answers is a different question from the buyer's.
What this article could not establish
Only two figures were checked: Figma and Linear. The rest cover products this cat either does not save or has not saved for long enough.
PricePulse's method is unknown, because it is not published — and since it is not published, a failure to reproduce here does not establish an error there. A different window, tier or billing cycle would give a different number, legitimately.
Whether the quoted figures still read the same has not been rechecked. They are as retrieved in August 2026.
This cat's own record is not complete either. Daily saving began in August 2026; everything earlier depends on what the Internet Archive happens to hold. The gaps are left as gaps.
Read one vendor, not the average
Deciding whether to buy or build does not require an industry average. It requires what the one vendor in front of you actually did.
That means opening the old pages one at a time. It is slower, and it comes with sources.
Every article on this site prints its archive.org URLs. Disagree with a figure and you can open it. If it turns out wrong, it gets corrected — one already has been.
A figure you can check and a figure you cannot are not the same figure.Sources
Quotations are as the page read when retrieved in August 2026.
The working behind this cat's own figures is in the individual articles.
- Figma's headline price rose while nothing got more expensive
- Four years of Linear's pricing
- Nine repricings, measured
Only the vendor's own published pricing and plan terms are used. No user reviews or forum posts.