How the Nationwide Cafe Visibility Census Works

Thomas Garner

Oct 7

13 min read

A cup of coffee and a notebook on a table

I keep getting some version of the same question once a shop owner hears we scored more than a thousand coffee shops across the country: how did you actually do that, and how do I know the number next to my region is real. Fair questions. This post is the honest, public answer, and it replaces any loose talk about a two hundred shop project being the whole picture.

What started as a Southeast only pass has grown into six regions and more than a thousand shops, and the method needs its own explanation now that the scope has changed this much. I am not going to reprint the full eight dimension walkthrough here, because the Southeast Audit Findings Library already did that work in detail and I would rather point you there than write a thinner copy of it. What follows is the part that is new: how the nationwide count works, what the two different snapshot dates mean, how the waves rolled out region by region, and where the honest edges of this project are.

What Gets Scored

Every shop in this census is scored on eight dimensions, each worth zero to five points, for a possible forty points total. That is the same scoring key the Southeast project used from the start, carried forward unchanged into every region added since. Nothing about the rubric itself has been loosened or tightened as the census grew. A shop scored in the Southeast in the spring is being judged by the identical yardstick as a shop scored in the Pacific Northwest this month.

The rubric looks at what a search engine or an AI model would see if it loaded the shop's own homepage right now: the title tag, the meta description, the H1 heading, the canonical URL, the structured data types present in the page code, the platform the site runs on, the robots file, the sitemap, and where a page could be reached without a captcha wall, what shows up in Maps, organic search, and AI answer evidence. This is a readability score for machines, not a taste score for humans. I will say that a second time later in this post because it is the single most common thing people assume this audit measures and it is not what it measures.

If you want the dimension by dimension breakdown, with worked examples of what a zero looks like next to a five, the Southeast method post covers that in depth and I am not going to clone it here. Read this post for how the national count works. Read that one for how a single dimension gets scored line by line.

One thing worth saying plainly: eight dimensions is a deliberate choice, not an arbitrary one. Fewer dimensions and the rubric starts missing real gaps, a shop could look fine on title and meta alone while its structured data is completely absent. More dimensions and the scoring starts drifting into subjective territory that a different reviewer might weigh differently on a different day. Eight, each worth up to five points, keeps the total at a round forty and keeps every dimension carrying roughly equal weight, so no single factor quietly dominates the score.

Two Dates, Never Blended

This is the part I want to be most careful about, because it is also the part most likely to get flattened into a single misleading number if I am not precise. There are two different snapshots of this data, taken two days apart, and they are not the same thing and should never be averaged together.

The first is an earlier count dated September 13, 2026, covering 1,087 shops across six regions. The second is a recount dated September 15, 2026, pulled directly from the current data files for each region, totaling 1,109 shops. Those two numbers differ because the West regions kept growing between the two dates, not because anyone made an error. I label every number in these posts either SECONDARY, meaning it traces to that September 13 count, or LIVE, meaning it traces to the September 15 recount. Anything that fits neither label is marked deferred and stays out of the published numbers until a dated pass covers it.

The LIVE regional counts as of September 15, 2026 are Southeast 211 shops, Northeast 215, Midwest 175, Mountain West 187, Pacific Northwest 157, and Western Coastal California 164, for a LIVE total of 1,109. The SECONDARY brief from September 13 used Southeast 201, Northeast 215, Midwest 175, Mountain West 186, Pacific Northwest 148, and Western Coastal 162. The right way to describe this project publicly is more than a thousand shops across six regions, then whichever dated table you are quoting from. Never split the difference and never present a blended figure as if it were a third, truer number. It is not. It is two honest snapshots taken two days apart.

Deferred nationwide are West organic search results and AI Overview results, along with a large share of the Wave B Maps pack data, Lighthouse performance scores, Google Business Profile star totals, and Keyword Planner search volume. None of those appear as numbers in these posts. Where there is no dated figure, the post leaves the figure out rather than filling the gap with something that sounds authoritative.

I know this labeling system, LIVE, SECONDARY, and a deferred list kept out of the numbers, is a bit more bookkeeping than most audits bother with. I would rather carry that overhead than let a reader assume every number in these posts was pulled from the same source on the same day, because it was not. A count compiled two days before a recount is still genuinely useful information. It is just a different snapshot, and pretending otherwise would be the kind of small dishonesty that erodes trust in every other number in the project, even the ones that are solid.

How the Waves Rolled Out

This census did not happen all at once. It happened in waves, region by region, and the order matters if you want to understand why some regions have thicker data than others right now.

The Southeast and Northeast went first, run in parallel as the first real test of whether the eight dimension rubric would hold up outside a single state. Once that pattern held across two separate regions rather than one, the Midwest followed as its own wave. After that came three more Western regions in sequence: Mountain West, then Pacific Northwest, then Western Coastal California. Each wave used the identical scoring key, the identical eight dimensions, and the identical LIVE versus SECONDARY labeling discipline as the wave before it. Nothing about the standard changed as the map got bigger. What changed was simply how many shops had been counted so far.

I mention the order because it explains something that would otherwise look strange in the numbers: why the Pacific Northwest count grew between the two snapshot dates while, say, the Northeast count stayed identical. Regions that were already closed out by September 13 did not move. Regions still being finalized in the West did.

There is also a practical reason to run regions as separate waves instead of trying to audit the entire country in one pass. Each wave becomes a checkpoint. Once Southeast and Northeast were both done and both showed the same basic pattern, thin median scores, a large majority of shops never publishing a blog post, commerce templates standing in for a real homepage, that was the signal the rubric itself was holding up outside a single region, not just describing something particular to the Southeast. Only after that checkpoint did it make sense to keep expanding the map. A single giant pass with no checkpoints in between would have made it much harder to catch a scoring inconsistency before it had already spread across a thousand rows.

The West Reopened Itself

The most significant thing that happened between the two snapshot dates was a final census pass through the Western regions, and it is worth explaining plainly rather than burying it in a footnote. That final pass reopened 101 Western cities that had already been through an earlier round, went back through them a second time, and found 25 additional shops that the first pass had missed.

That is not a small correction. Reopening more than a hundred cities to check the work a second time, on top of the original count, is exactly the kind of thing a careful census does and a rushed one skips. It is also the direct reason Pacific Northwest LIVE sits at 157 shops while the SECONDARY count two days earlier had it at 148. Those shops were not new arrivals. They were always there. The first pass simply had not found all of them yet, and the second pass did a better job.

I want to be honest about what that implies for the rest of the country too. If a second, more careful pass through 101 already audited Western cities turned up 25 more shops, it is reasonable to assume other regions, including the Southeast and Northeast that went first and earliest, likely have a small number of shops still missing from their own counts. I am not going to pretend the census is a closed, final, perfectly complete number. It is a careful, dated snapshot that gets more accurate with each additional pass, not a claim that every coffee shop in America has now been found and scored.

Why go back through 101 cities at all instead of just moving forward to new territory. Because a census that only ever moves forward and never checks its own earlier work tends to accumulate quiet blind spots, a shop that opened after the first pass came through, a shop that was missed the first time because its listing sat under an unusual category, a shop whose website changed platforms between the two passes and looked different the second time it was checked. Reopening those 101 cities was slower than simply adding new cities to the map, but it produced a genuinely more accurate Western count, and that tradeoff was worth making.

What This Audit Is Not

I would rather spell this out directly than let someone assume something the data was never built to say. This audit does not measure how good the coffee tastes. Nothing in the eight dimensions touches the cup at all. A shop with a mediocre pour and a technically clean, well structured website can score well here, and a shop pulling the best espresso in its whole metro area can score poorly if its homepage is a bag store with no room description, no structured data, and a dead blog tab. That gap is the entire reason this project exists, not an oversight in how it was built.

This is also not a weeks to rank experiment. Nothing in this census tracks how long it takes a page to climb a search result after a change is made, because that was never the question being asked. The question being asked is simpler and more immediate: right now, today, what does a search engine or an AI model see when it looks at this shop's own homepage. That is a snapshot, not a forecast, and I am not going to dress it up as one.

And this is not a claim that every specialty coffee shop in America has been found. A thousand plus shops across six regions is a real, substantial, carefully labeled sample. It is not a national registry. New shops open, existing shops close, websites get redesigned, and platforms change hands constantly. Every number in these posts carries a date next to it for exactly that reason. Treat each figure as accurate as of the date attached to it, not as a permanent fact about the American coffee industry.

Reading the National Numbers Versus the Southeast Library

One more thing worth being precise about. The Southeast Audit Findings Library, sixteen posts deep, still speaks to its own original vintage: 201 shops across 53 cities, captured as of a September 9, 2026 snapshot. That library has not been rewritten to match the newer, larger LIVE count, and it should not be read as if it had been. The current LIVE Southeast count in this nationwide series is 211 shops, ten more than the library's original number, for the same reason the West grew between its own two dates: later passes keep finding shops earlier passes missed.

So here is the simple rule I use, and the rule I would encourage anyone citing this project to use as well. For Southeast specific method depth, the dimension by dimension walkthroughs, the score band breakdowns, the Carolinas schema corridor, and anything else that lives at the 201 shop, 53 city vintage, go to the Southeast library directly. For the national picture, the six region table, what the Midwest and the three Western regions added to the original two region hedge, and the mega metro contrast list, these posts are the source. The two are companions, not competitors, and neither one should be quietly rewritten to sound like the other.

This distinction is not pedantry. It is the same sourcing discipline that runs through every fact in this post: label what you are citing, date it, and never let two different counts blur into one number that neither snapshot supports.

Why I Am Publishing the Method at All

I could have just published the findings and left the method as an internal document. I did not, on purpose. If a shop owner is going to read that their region's median score sits in the teens out of forty, or that a supermajority of shops in their region have never published a single blog post, they deserve to know exactly how that number was produced, not just be handed a conclusion and asked to trust it.

That is also, honestly, how I would want to be treated if someone were scoring my own business. Show the rubric. Show the dates. Show which measurements are in the numbers and which are still deferred. A method a business owner cannot inspect is not a method worth trusting, and I would rather have this project be slower and more transparent than fast and vague.

If you run a coffee shop and you are curious where your own site would land against this same eight dimension rubric, that is a conversation worth having directly rather than guessing from a regional median. Reach out through hillcane.co/contact or call (256) 384-2449, and we can walk through what a real, dated look at your own homepage would show.

Frequently Asked Questions

How many coffee shops has Hillcane scored?

As of September 15, 2026, the nationwide census covers 1,109 shops across six regions: Southeast, Northeast, Midwest, Mountain West, Pacific Northwest, and Western Coastal California. An earlier count dated September 13, 2026 covered 1,087 shops. Both numbers are accurate for their own date. The honest public description is more than a thousand shops across six regions, with the dated table specified.

What are the eight dimensions in the scoring rubric?

The rubric scores what a search engine or AI model would see on a shop's own homepage: title tag, meta description, H1 heading, canonical URL, structured data types, platform, robots file and sitemap, plus available Maps, organic, and AI answer evidence. Each dimension is worth zero to five points for a possible forty points total. The Southeast method post covers each dimension with worked examples in more depth than this post repeats.

Why do some regions have different shop counts on different dates?

Because the census was built in waves, and a final pass through the Western regions reopened 101 already audited cities and found 25 additional shops that the earlier pass had missed. That is why Pacific Northwest LIVE sits at 157 shops while the September 13 count had it at 148. Regions that were already closed out before the second pass, like the Northeast, did not change between the two dates.

What order did the regions get audited in?

Southeast and Northeast went first, run together as the initial test of the rubric outside one state. Midwest followed as its own wave. Then three Western regions ran in sequence: Mountain West, Pacific Northwest, and Western Coastal California. Every wave used the identical eight dimension scoring key, with no changes to the standard as the map grew larger.

Does this audit measure how good the coffee tastes?

No. Nothing in the eight dimensions touches the cup. This is a readability score for what a machine can see on a homepage, not a taste score. A shop with excellent coffee and a thin, structured data free website can score low here, and that gap between reputation and what a search engine can parse is the entire reason this project exists.

Is this a claim that every coffee shop in America has been found?

No. A thousand plus shops across six regions is a real, carefully labeled sample, not a national registry. The West final census pass found 25 shops a first pass had missed across 101 cities, which suggests other regions likely have a small number of shops still uncounted too. Every figure in these posts carries a date for that reason.

What do LIVE and SECONDARY mean in these posts?

LIVE means the number traces to the current data recount dated September 15, 2026, 1,109 shops. SECONDARY means it traces to the earlier count dated September 13, 2026, 1,087 shops. Measurements that fit neither, such as West organic search results, AI Overview results, Lighthouse scores and Keyword Planner volume, are deferred and do not appear as numbers in these posts.

Should I still read the Southeast Audit Findings Library?

Yes, for Southeast specific depth. That sixteen post library still speaks to its own original vintage of 201 shops across 53 cities, dated September 9, 2026, and covers dimension by dimension detail, score bands, and regional corridors these nationwide posts do not repeat. These posts are the source for the national six region picture; the library is the source for Southeast method depth.

Does a higher score mean a shop will rank higher in search results next week?

No. This is not a weeks to rank experiment and does not track ranking movement over time. It is a snapshot of what a search engine or AI model can currently see on a shop's own homepage, scored against a fixed rubric on a specific date. Ranking depends on many factors this census does not attempt to model or predict.

How can a coffee shop owner find out their own score?

The most direct way is to ask. Reach out through hillcane.co/contact or call (256) 384-2449, and we can walk through what a dated look at your own homepage against the same eight dimension, forty point rubric used across all 1,109 LIVE shops would show, which of the eight dimensions sit at zero today, and what specifically would move that score.

Work with Hillcane

Eight dimensions, forty points, more than a thousand coffee shops, and two different dated snapshots that should never be blended into one number. Here is exactly how the nationwide cafe visibility census actually works.

Curious where your own homepage would land against the same eight dimension rubric used across all 1,109 LIVE shops. Reach Hillcane at hillcane.co/contact or call (256) 384-2449.