When we launched this site we put up a page called How We Rank Ourselves and made a specific promise on it: we would publish our own search numbers every month, including the flat months, because our clients cannot buy their way to traffic and neither can we.
The first update was due in early August. It is now the twenty-fifth, and the four figures that page actually promised, keyword positions, organic impressions, referring domains and audit calls, are not in this update either. That is the least flattering way to open a progress report and it is the accurate one.
What follows is the whole account: what we shipped in eight weeks, the rule that shaped all of it, the pages we refused to publish, the mistakes we caught, and why you are not looking at ranking data today.
What we shipped
On 1 July the site had 18 pages in its sitemap. Today, counting this post, it has 68. Thirty of those are market guides: state and city pages covering cannabis retail in California, Washington, Massachusetts, Connecticut, Nevada, Oregon, Maryland, New York, Michigan, Missouri, Arizona, Minnesota, Illinois and Colorado, hemp in Texas and Florida, and metros including Detroit, Kansas City, Phoenix, Minneapolis, Denver, Boston, Baltimore, Portland and the New York boroughs.
Alongside those, 56 pages now carry FAQ structured data and 30 publish a sourced dataset with citations attached. The homepage sits at 49.8 KB, down from the 2.5 MB it weighed before the rebuild.
Those are all counts you can verify yourself in about a minute by opening the sitemap. That is the point of publishing them.
One rule did most of the work
Every number on every market guide comes from the responsible agency's own file, and carries three dates: what the file says it covers, when the agency published it, and when we read it.
That rule sounds like bookkeeping. In practice it is the difference between a page that repeats what everyone else repeats and a page that says something true and specific. A few examples of what it produced:
- Michigan. The Cannabis Regulatory Agency breaks monthly sales into five regions. Southwest Michigan sold $89.9 million in July 2026 against Wayne County's $30.6 million, nearly three to one, and by weight moved 210,653 pounds against Wayne's 54,787. The biggest volume in the state moves nowhere near its biggest city. The five regions sum to the statewide total to the cent, which is how we knew the parse was right.
- Denver. Colorado publishes sales by county back to 2014. Denver's share of the state has fallen in every one of the twelve years since, from 47.7 percent to 23.9 percent. Not most years. Every year.
- Illinois. Sales went from $171.2 million in May 2025 to $132.4 million in June 2025 and never came back, averaging $127.4 million across the thirteen months since. That is a step, not a slide, and it changes what any year-over-year comparison spanning that month actually means.
None of those came from an industry estimate or a press release. They came from a spreadsheet, a Word document and a state dashboard, and each one is linked on the page it appears on so you can check it.
The pages we did not publish
The rule cuts the other way too, and this is the part most agencies leave out of a progress report.
Las Vegas has no guide. Nevada's Cannabis Compliance Board publishes an establishment list with a county column and no city column, so the file cannot tell you how many dispensaries are inside Las Vegas proper. The City of Las Vegas does publish city-level counts, but that dataset stopped updating in 2023. We could have published three-year-old numbers with a footnote. We did not, and the market that most people would expect to see first on a cannabis SEO site is simply absent.
Colorado has no statewide store count. The state's licensed facilities file sits behind a block that refuses automated requests, so the Colorado guide reports sales, taxes and county detail, and says plainly that it is not stating how many dispensaries the state has. Denver's own city data was reachable, so Denver's count is on the page with its date.
Georgia and Tennessee have no hemp guides. Georgia moved hemp licensing behind a login portal and publishes only application-process documents. Tennessee's list is portal-only. Florida did have a usable file, so Florida got a guide.
We keep those refusals in the source files rather than quietly dropping them, because a gap that is written down is a gap someone can close later. Sixteen state source files are recorded in full, including the ones that failed.
Three things we got wrong
We wrote Colorado off early. An initial check hit an old, unmaintained dataset, and the conclusion was that the state's data was stale. It was not. Colorado's actual sales files refresh monthly and were updated the day we finally used them. A state being out of date in one place says nothing about the rest of what it publishes, and that assumption cost us weeks of coverage in a top-tier market.
We nearly published a table of nulls. Illinois publishes its sales through a dashboard rather than a download. The extraction we had used successfully for other states returned empty values for every Illinois figure, because numeric columns in that format are indexed differently from text columns. The failure was loud, which is the only reason it was harmless. A quieter version of the same bug is exactly how wrong numbers reach a live page.
We caught a mislabeled figure with one check. Michigan's monthly report contains a plant-count table showing more than 1.4 million plants in vegetative growth. It sits in the medical section, not the adult-use section, and the adult-use numbers are different. Reading the table without checking which program it belonged to would have put a striking and wrong statistic on the Detroit page.
What we cannot show you today
Keyword positions, organic impressions, referring domains and audit calls booked from the site all come from the same place: a Search Console export that has not been pulled. Until it is, we do not have those figures, and the alternative to publishing them is publishing an estimate dressed as a measurement.
That is the specific shortcut the transparency page exists to refuse, so the four rows on it say "not published this update" instead. They land in the next update with the export date printed beside them, and this eight-week gap stays visible in the history rather than being smoothed over later.
There is a real reason to be honest about this beyond principle. A new domain publishing 50 pages in eight weeks has no business claiming ranking wins yet. Indexation takes time, competitive terms take longer, and any agency showing you a hockey stick eight weeks into a new site is showing you something other than organic search.
What update #2 will contain
The four search figures with their export date. A note on which of the 30 market guides are indexed and which are not, because that is the first real signal and it is measurable. And whatever the numbers say, including if they say very little, which at this stage is the likeliest outcome.
If you want to see the method rather than read about it, the market guides are the artifact. Open Michigan or Denver and check a figure against the state file it cites. That is the same standard we apply to client work, and it is the only reason to trust a number on a marketing site at all.
Frequently asked questions
Why is this update late?
The first monthly update was promised for early August and published on 25 August. The delay is stated on the transparency page rather than backdated, and the gap remains visible in the update history.
Why are there no keyword rankings in this update?
Keyword positions, impressions, referring domains and audit calls come from a Search Console export that has not been pulled. Rather than publish estimates in their place, those four rows are marked not published and will appear in update #2 with the export date beside them.
How many pages has Nearfront published since launch?
The sitemap went from 18 pages on 1 July 2026 to 68 on 25 August 2026. Thirty of the new pages are state and city market guides, and 30 pages publish a sourced dataset with citations.
Where do the numbers on the market guides come from?
Each figure comes from the responsible state agency's own published file, and carries the period the file covers, the date the agency published it, and the date we read it. Where an agency does not publish something, the page says so instead of estimating.
Why is there no Las Vegas market guide?
Nevada's state establishment list carries a county column but no city column, so it cannot show how many dispensaries sit inside Las Vegas proper, and the City of Las Vegas dataset that does carry city counts stopped updating in 2023. Publishing three-year-old counts was not worth the page.
