Of five pricing pages we have been reading since the second of October, Linear’s and monday.com’s are the ones whose control counts are still the numbers they were. The other three we will not say that about, and the reason is different for each of them — including one whose count is not the number it was when we started, which is how a reader can tell that the instrument behind this page is able to see a difference at all.
Five pricing pages, fetched and counted by code with no model in it — no buyer, no decider, nobody reading the page — each read until its own list of controls stopped changing, and then read the same way again, and again, since the second of October. Linear’s pricing page has not moved across 40 readings and 519 hours of clock. Every one of those readings counted 64 controls on it, and repeated fetches of it, in sitting after sitting, have established that the page hands that count to every fetch rather than serving more than a single version of itself. monday.com’s pricing page has not moved either: 169 controls across 9 readings and 144 hours of clock. Repeated fetches of monday.com’s page have established the same thing about it — it hands that count to every fetch rather than serving more than a single version of itself. Its window is shorter than Linear’s for a reason worth printing: it starts where its count stopped being the number it was when we started watching, and the edit itself is published, with the control names, on our page about what changed in seven days. Airtable’s page was on course to carry the same sentence and does not: our earlier reading of it counted 103 controls where this morning’s counted 99, so that count has changed. Counts that held and a count that stopped are what this page is for — Linear’s and monday.com’s against Airtable’s — counted by the same code, on the same pages, on the same mornings.
Nobody was in the loop. No ghost user walked these pages, no model decided anything, and not one sentence or number here was written by one — which is why this is a reading and not a run, and why there is nothing on it about where a person got stuck.
- The instrument
- Each page was opened in a real browser, watched until its own list of controls stopped changing, and then counted — the same way, by the same code, on every reading since the second of October. Nothing was signed up for, nothing was pressed, and no sentence and no number on this page was written by a model. A page may only be said here to hold a number if every reading we have taken of it agrees about that number, and if repeated fetches of it have established that the page hands the same number to every fetch rather than serving more than a single version of itself. That second condition is the one that does the work, and it is why Notion AI’s and Retool’s counts carry no claim here in either direction.
- When
- What is published here is the window we read across and not a window we infer. The hours are the distance from the last clock of the earliest agreeing reading to the first clock of the latest, which throws away every minute inside both, so the figure is a floor rather than an estimate: the count has held at least that long and possibly longer. It is not an edit frequency — nothing here says how often these companies touch these pages, and a count that has not moved while we watched may change the moment after this page goes up. The count in the table is this morning’s own.
- Pages read
- monday.com/pricingairtable.com/pricingretool.com/pricinglinear.app/pricingwww.notion.com/pricing
Counted on the pricing pages themselves.
We opened each pricing page, waited until its own list of controls stopped changing, and counted the controls that send a reader somewhere else in a new tab — recording, for each one, whether it goes to material of that company's own on a separate site or to somebody else's. These are a fact about these pages on 20 October 2026 and nothing else, and anyone can open the same pages and count them again.
| Pricing page | Controls on the page | Open a new tab | …to its own material on a separate site | …to somebody else's site |
|---|---|---|---|---|
| monday.commonday.com/pricing | 169 | 32 | 8community.monday.com, ir.monday.com, support.monday.com | 13 |
| Airtableairtable.com/pricing | 99 | 17 | 5community.airtable.com, support.airtable.com | 7 |
| Retoolretool.com/pricing | 231 | 10 | 4docs.retool.com | 2 |
| Linearlinear.app/pricing | 64 | 6 | 0 | 4 |
| Notion AIwww.notion.com/pricing | 73 | 0 | 0 | 0 |
What we read and are not publishing.
- Which controls left Airtable’s pricing page
- This page prints both counts of that page and names nothing on it. Going from a count to a name means pairing a control in the earlier reading with a control in this one, and a control whose label was rewritten at an address that stayed put is an edit a careless pairing prints twice — a demolition and a construction. That pairing is published, with the named controls, on our page about what changed in seven days, and it was not done for this reading. So the honest form of what we know is the pair of counts and nothing more.
- Anything at all about Notion AI’s and Retool’s control counts
- Our own fetches disagree with each other about Notion AI’s control count inside a single sitting, and about Retool’s in sitting after sitting. A page that does not hand the same count to every fetch of it cannot be told, from a run of readings that agreed, to have been left alone — the agreement would be our own sampling published as their restraint. That is not a claim that their pages are as they were, and it is not a claim that they edited them; it is a claim we are not entitled to make in either direction, and no later reading lifts it. Both pages are in the table below, because leaving them out would be this page with its awkward half deleted.
- That monday.com’s page is as settled as Linear’s
- Both counts are in the same column of this page and the windows behind them are not the same size. Linear’s reaches back to the second of October; monday.com’s only reaches back to the morning after its count stopped being the number it was when we started watching, and the readings before that are about a different number. The edit itself is published, with the control names, on our page about what changed in seven days. So the two sentences are the same kind of sentence and nothing like the same weight of evidence, and this page prints both windows rather than averaging them.
- How often any of these companies edits a pricing page
- Nothing here is a rate. A handful of pages watched for a fortnight does not license a sentence about how often anybody ships; it licenses sentences about these readings. A reader who wants a rate is owed many more windows than this.
- A price reading
- This page publishes no reading of what anything costs. It is about whether the number of controls on a page is the number it was, and the price readings of these same pages are published, with their records, on the pages that compare them.
- Anything a person did
- Nobody walked these pages. There is no buyer here, no model decided anything, and no sentence on this page was written by one — so there is nothing here about where somebody got stuck, and these numbers are not evidence about what a reader would do.
What is worth anything here is not the size of a number but that it is the number it was, and it cannot be produced by pointing a tool at a competitor’s pricing page today. It takes having read everybody’s pages, over and over, for weeks — and it takes having caught somebody changing theirs, because readings that all agree prove nothing if the instrument that took them could not have seen a difference. Airtable’s page is that catch, and it is a catch with a length to it: its count had held at 103 across 32 readings and 376 hours of clock before this morning’s reading counted 99. The pair of readings is printed here rather than summarised: where our earlier reading counted 18 controls opening a new tab, this morning’s counted 17; where it counted 8 of those going to somebody else’s site, this morning’s counted 7. What we are not saying is which of its controls are gone. Pairing a control in the earlier reading with a control in this one is different work from counting, and a control whose label was rewritten at an address that stayed put is an edit a careless pairing prints twice; that pairing is published, with the named controls, on our page about what changed in seven days, and it was not done for this reading. Nothing here asks a reader to trust a subtraction — both readings are printed, and anyone can open the page and count it.
← Everything we have published, including the runs with a buyer in them