Methodology.

How Cannabest collects, verifies, and publishes cannabis retail information. Last updated: August 3, 2026. Version 1.0.

Cannabest is a cannabis retail directory built on live menu data. We track what dispensaries actually carry, what they charge, and when we last confirmed it, then publish that with the date attached. We do not sell placement, we do not sell inclusion, and we do not sell removal.

This page explains exactly how that works, including what we do not yet do. If you find something here that does not match what you see on the site, that is a defect and we want to hear about it. See Corrections.

What we publish today

Cannabest currently publishes directory information: dispensary profiles, brand profiles, and the geographic and category pages that organize them. Every fact on those pages comes from one of the sources listed below and carries a verification date.

We do not currently publish first-hand product reviews, scored rankings, or best-of lists. When we do, they will be labeled by basis and will carry the name of the person who did the work along with the date they did it. We would rather launch with less than imply experience we have not had.

Where our data comes from

Menu and availability data. We license live menu data covering dispensary inventory, pricing, and product availability. That data originates from the menu platforms dispensaries themselves use, including Dutchie, Weedmaps, iHeartJane, Leafly, Sweed, Dispense, and Treez. We ingest it daily.

Dispensary details. Names, addresses, coordinates, phone numbers, websites, and license numbers where available come through the same licensed feed, which sources them from the dispensary's own menu platform record.

Legality and regulatory status. State-level cannabis legality, medical program status, and whether licensed retail actually operates in a given state are verified against state regulators and licensing authorities, with sources recorded. We treat "legal" and "has operating stores" as different facts, because they are: some states have legalized cannabis without licensing a single retailer.

How we verify

Freshness is structural, not promised. Every menu-derived fact carries a timestamp from the moment we last observed it. When a page says a price was verified on a date, that date comes from our ingestion record, not from an editor's intention. If we have not seen a dispensary's menu in more than 60 days, we flag it internally for status review rather than continuing to present stale inventory as current.

Items we have not seen in 30 days are treated as unavailable, not as available. We would rather understate a menu than send someone to a store for something that is gone.

Entity resolution. Menu data arrives messy. The same brand appears as a dozen different strings across different stores, with different punctuation, abbreviations, and misspellings. We resolve those to canonical records through a matching process: exact matches on known name variants, then normalized matching, then fuzzy similarity scored against every brand we know. High-confidence matches link automatically. Ambiguous ones go to a human review queue. Nothing is guessed silently, and every confirmed variant is recorded so the same string never needs deciding twice.

Completeness checks. When we ingest a market, we reconcile what we received against what our data provider reports should exist. A mismatch is treated as a failure of that run, not as a result. This exists because incomplete data that looks complete is worse than no data.

How pages earn publication

Not every page we could generate is a page worth publishing, and search engines are right to be skeptical of directories that publish everything. Cannabest pages are excluded from search indexes by default and become eligible only when they clear a substance threshold:

• Dispensary profiles need complete verified location data plus either linked live menu data or a meaningful set of unique structured facts beyond name, address, and phone.
• City pages need multiple qualifying dispensary profiles plus original content specific to that city.
• State pages need qualifying city pages or dispensary profiles, plus original content.
• Brand profiles need either multiple resolved products or verified availability across several retailers with current pricing, plus original content.
• Category pages need a meaningful number of qualifying brand profiles.

Pages that do not clear their threshold still exist and remain usable if you navigate to them. They are simply not offered to search engines as answers, because they are not yet good enough to be answers.

What we never do

• No paid placement. No dispensary, brand, or operator can pay to be listed, to rank higher, to be described more favorably, or to be removed. There is no rate card for any of that, and there will not be one.
• No fabricated content. We do not publish invented reviews, invented quotes, or synthetic user activity. We do not present editorially written questions as though users asked them.
• No unlabeled experience claims. If we have not visited a store or tested a product, we do not write as though we have.
• No health or medical claims. We describe what is available and what it costs. We do not tell you what cannabis will do for a medical condition, because we are not qualified to and because it would be illegal for us to.

Commercial relationships

Cannabest is owned by Heady, a cannabis marketing agency with clients in the industry. Some of those clients appear in this directory. We disclose that relationship on every page where such an entity appears, automatically and at the template level rather than at an editor's discretion. See Disclosures for the full explanation of why we include them rather than excluding them.

What is coming, and labeled as such

We are transparent about the roadmap because the gap between what we do now and what we intend to do is itself a fact readers deserve:

• First-hand tested reviews and rankings, with named reviewers, recorded test dates, and a published scoring rubric. Colorado first, because that is where we can physically do the work.
• License verification against state registries, which will let us confirm license status independently rather than relying on what a dispensary entered into its menu platform.
• Consumer-facing signals such as ratings, photos, and hours sourced from mapping providers.
• Additional markets, tiered by depth. National directory coverage, tracked markets with live menu data, and tested markets with first-hand work are three different things and will always be labeled differently.

Changes to this methodology

This page is versioned. When our methods change, we update it, note the version, and date the change. We do not quietly revise how we work.