What Our Amsterdam Coffeeshop Listings Include (and What They Don't)

So what are you actually looking at when you open one of our listings? A dated menu record—not a live stock check, a shop review or a promise about what’s behind the counter. Let’s draw that line clearly.

What the catalogue holds

At the snapshot generated on 21 October 2026 the site lists 251 Amsterdam coffeeshops. Of these, 212 have captured menus and 186 have dated menus: 28 fresh within the observation window from 21 September to 21 October, and 158 last known. The remaining groups are 26 shops with captured material without a usable date and 39 without captured menus. Across the archive, 10,743 price lines have been observed, of which 10,690 are trusted; the cheapest leaderboard draws from 4,061 entries. By month of last-seen menu, the snapshot counts 22 shops in October 2026, 18 in September, 15 in August and 23 in July. Older archives remain labelled as historical.

Where the lines come from

The source material is photos of menus. Find a menu, read its date from the page itself, transcribe the lines into price records. Lines that fail the checks aren’t deleted - they stay untrusted, which is why 10,743 observed lines and 10,690 trusted lines are two different numbers on one page. A doubtful line gets flagged, never smoothed over, because a silent deletion would look exactly like a correction.

How a shop page is assembled

Each shop page shows its captured menus with their dates, in blocks marked fresh or last known. Fresh means the menu’s date falls inside the last 30 days, and only those prices feed the city medians. Last-known prices stay on the page with their dates, and the difference to the current median is shown as indicative rather than as a saving. A daily rebuild refreshes the site without rewriting a single record: a price stays exactly as old as its menu. Neither label judges the shop - fresh and last known describe the menu’s age, and both blocks stay on the page because both carry information of a different kind.

What every price line carries

Read a price line through its product name, category, format, printed price and available source date. A trusted flag indicates a transcription check, not a live stock check. A missing field must not be supplied by guesswork: if the source gives no usable weight or date, that uncertainty matters. The last-menu-seen guide explains the distinction between a captured menu and a dated observation in detail.

Read together, a line makes one narrow statement: this shop printed this name, in this format, at this price, on this date. That is the entire claim, and every page on the site is built out of statements of exactly that size. When a field is absent, it stays absent. A line without a printed weight keeps a unit price instead of a per-gram one, and a name without a variant keeps its bare form. The gaps are information too: they show where menus stopped printing.

What a listing never contains

No live till readings, ever. A price on this site is what a shop wrote on a menu on a day, not what the counter charges at this moment. No stock information: the archive cannot tell you whether a line is on the shelf tonight. No quality judgments: nothing here rates products or tells you what anything is like. No booking, no ordering, no alerts. And nothing is for sale: the site is 18+, info only, no sales, and the only counter that matters is inside the coffeeshop.

How the numbers on a listing are built

Weed, hash and extracts are compared per gram, computed only when the menu prints a weight. Joints and other pieces are compared per unit and never enter a per-gram median. Medians come from fresh menus only, so a €13 city median for weed 1g is a statement about 139 listings across 27 shops in the 30-day window, not about the whole archive. Comparisons stay like with like: formats against the same formats, variants against the same variants, so a Cali gram is never placed against a Dutch gram and a pack total never lands in the single-gram column.

What the numbers are for

Used properly, a listing answers three questions. What did this shop print, and when? Where does that sit against the fresh medians in the price overview? How does it compare with other observations in the same format? Used as a live price list, it cannot answer the question being asked. The pages help you prepare a specific query for a visit, while the source menus show what was last captured.

The limits, stated plainly

The catalogue is an archive of printed menus. In the October observation window, 28 of 251 listed shops supplied fresh evidence, so medians rest on a minority of the directory. Per-format samples differ: weed 1g has 139 listings, hash 1g has 93 and extracts 1g has 40 in that snapshot. Read those counts beside the figures rather than using the entire archive total to describe a recent sample. Prices are observed, not live, so always verify on site. The method page explains capture and the shop directory exposes individual records. A shop that changed its menu yesterday may still show older evidence until another menu is captured. The archive cannot infer the missing update from the date of its own rebuild.

What happens when a line is incomplete

The site can preserve a printed line even when it cannot turn that line into a comparable per-gram value. If a menu gives a product name and total but no weight, the record remains a unit price; no weight is guessed. If a date cannot be established, the image does not silently become fresh. If wording is ambiguous, transcription does not resolve it into a stronger claim than the source supports. These limits may leave fewer comparable rows, but they keep the published result tied to what the menu actually showed. Check the menu-reading guide for category and unit, and the coverage report for how missing or undated evidence is counted.