College Town Versus Tourist Town Cafe Visibility Scores
An owner in a university town hears that campus traffic makes online visibility easy. An owner in a resort town hears the opposite, that seasonal visitors make it hard to build anything lasting. Neither one has ever actually seen the numbers side by side, because until now nobody had put them side by side.
Short answer: market type is a real setting, but it is a weak predictor of a shop's visibility score. Cities coded as campus towns and cities coded with a high tourist index score modestly better on median visibility than the rest, but the gap between the best and worst median is a handful of points out of forty, and the range within every single group is far wider than the gap between groups. A shop's own site and profile still matter more than which kind of town it sits in.
This is a study built from Hillcane's own nationwide census, more than 1,000 independent specialty coffee shops across six regions, snapshot date September 15, 2026. Every number below traces back to that dataset, joined against a separate city level file that records whether a city has a college campus and how tourist driven its economy is. The method box below explains exactly how that join was built, because the labels in this post belong to cities, not to individual shops.
How the Labels Are Defined and Joined
Two labels matter here, and both are set at the city level, not the shop level. Campus is a simple yes or no flag recorded per city. Tourist index is a coarser rating, low, medium or high, also recorded per city. Neither label describes an individual cafe's customer base directly. It describes the town the cafe happens to sit in, which means every shop in a coded city inherits that city's label whether or not the shop's own clientele actually matches the label.
That distinction is worth sitting with before looking at any numbers. A specialty roaster three blocks from a stadium in a campus town and a drive through kiosk on the edge of the same town both get coded campus yes, even though their actual foot traffic may look nothing alike. The census was not built to measure an individual shop's customer mix. It was built to measure how shops are represented online, and the city labels are the closest available proxy for market type at scale.
The join itself was built from six regional shops files stacked together, matched against six regional city ground truth files using a key built from city name and state, lower cased for consistency. Every one of the 1,109 shops in the combined census matched a city row, across 264 distinct cities. Of those shops, 528 sit in a city coded campus yes and 581 sit in a city coded campus no. On the tourist side, 222 shops sit in a high tourist index city, 421 in a medium tourist index city, and 466 in a low tourist index city.
Those denominators matter because some of the cells this study can build from crossing the two labels are thin. A rough pattern in a cell with a few dozen shops deserves a different level of confidence than a pattern in a cell with several hundred, and this post says the count out loud in every place a comparison is drawn, rather than letting a small cell pass as a firm finding.
It is also worth explaining why the ground truth file records these two labels at all. Both were set during the same city level research effort that produced the census's demand context, and both exist to give a coarse read on the kind of demand a city's cafes are likely competing for, not to score any individual shop's actual clientele. A city level label is a blunt instrument by design, useful for a study like this one precisely because it applies the same rule consistently across every city rather than trying to guess at an individual shop's customer mix, which no file in this census records.
Reading the labels honestly also means accepting their limits up front. A city can be coded tourist index high because of an attraction that draws visitors seasonally, and a shop three miles from that attraction may see almost none of that traffic, while a shop directly next to it sees a great deal. The label describes the city's overall demand profile, not a guarantee about any single address inside it, and every number that follows in this post should be read with that caveat attached.
The Median Scores, Cell by Cell
Across the full census, the overall median visibility score is 16.0 out of 40, recounted September 21, 2026. That is the baseline every cell below should be read against. None of the cells below move far from that baseline, and that is the headline finding as much as any individual number.
Shops in cities coded campus no and tourist index low sit at a median of 15. Shops in cities coded campus no and tourist index high sit at a median of 17. Shops in cities coded campus yes and tourist index low sit at a median of 15. Shops in cities coded campus yes and tourist index high sit at a median of 18. Every one of those medians sits within a couple of points of the 16.0 overall median, and within a couple of points of each other.
Put plainly, the widest gap in this whole comparison, between campus yes plus tourist high at 18 and campus no plus tourist low at 15, is three points out of forty. That is a real, measurable difference and it is worth naming honestly. It is not a difference that explains why one shop is thriving online and another is invisible. A three point median gap between the best coded combination and the worst coded combination says market type nudges the number, it does not decide it.
The medium tourist index cells sit closer to the middle still, consistent with what you would expect from a coarser, three way label rather than a sharp cutoff. None of these medium cells produced a result far enough from 16.0 to call out on its own, and that consistency across the medium band is itself part of the evidence that this is a soft pattern, not a hard rule.
It is worth naming what a three point median gap looks like in practice. Out of a forty point scale, three points might separate a shop that has a working, if thin, blog and a shop with no blog activity at all, or a shop with complete schema markup and one with markup that is only partially present. That is the scale of difference the strongest market type combination buys a city's shops on average, and it is smaller than the difference a single fixable item on a shop's own site can produce for that individual shop.
Why the Overlap Matters More Than the Median
A median tells you the middle of a distribution. It does not tell you how wide that distribution is, and the width is where the real story sits. Inside every single cell in this study, whether campus yes or campus no, tourist high or tourist low, there are shops that score well above their cell's median and shops that score well below it. The spread inside each group is large enough that a shop's own individual score cannot be predicted from its city's label with any real confidence.
This is the practical takeaway for an owner reading this from either side of the comparison. If you run a shop in a campus town with a high tourist index, the data here says your city sits in the highest scoring cell measured, median 18. It does not say your specific shop scores 18. Plenty of shops in that same cell score well below that median, and plenty of shops in the lowest cell, campus no and tourist low at median 15, score well above it. Whatever city a shop sits in, that shop's own choices about its site, its profile and its content still explain more of its individual score than its city's label does.
This is also why the post avoids any causal language about why campus or tourist status might raise or lower a score. It is entirely plausible that campus towns produce more built in search demand, or that tourist towns produce more incoming links and mentions from travel content, but none of that is measured here. What is measured is a correlation between two city level labels and a shop level score, and the correlation is real but modest. Reasoning about a possible mechanism belongs in a different post with different evidence behind it, not stapled onto this one as if it were proven.
This study also does not repeat or replace the live State of Specialty Coffee Visibility research, which reports the national picture without cutting by market type. This post is a narrower cut of the same underlying census, built specifically to answer the campus versus tourist question and nothing broader than that.
One more limit worth spelling out, because the data assets this study draws from are explicit about it: this comparison uses the census's own /40 visibility scoring scale, and it must never be combined or averaged with the separate Southeast audit's score bands from the earlier, smaller vintage study. Those two scales measure different things on different ranges, and mixing them would produce a number that looks precise but means nothing. Every figure in this post stays inside the census /40 scale, consistently, from the overall median down to each individual cell.
What This Means for an Owner Deciding Where to Focus
The temptation on reading a study like this is to treat a city's label as an excuse or a guarantee. An owner in a low tourist, no campus town might read this and conclude their market is simply harder, full stop. An owner in a high tourist, campus town might read this and assume their market is simply easier, full stop. Both conclusions overstate what a three point median gap supports.
A more useful reading is that market type sets a mild tailwind or headwind, not a ceiling or a floor. The blog and schema patterns recorded across the same census, blog activity absent in the large majority of shops nationwide and schema markup incomplete or missing in the majority as well, show up across every one of these cells, campus or not, tourist heavy or not. The baseline problems this project has documented elsewhere in the census are not solved by sitting in a favorable city, and they are not a lost cause in an unfavorable one.
If anything, the overlap inside every cell is the more actionable finding here. A shop in the lowest scoring cell that fixes the same basics, a working blog, complete schema, an accurate and complete profile, has every reason to expect to land well above its own cell's median, because plenty of shops in that same cell already do. Market type describes the neighborhood a shop is playing in. It does not decide the score.
The honest summary an owner in either kind of town can take from this study is that neither the campus flag nor the tourist index should be treated as a verdict. Both labels are real, both were measured carefully, and both produce a small, consistent nudge on the median. Neither one comes close to explaining why one specific shop scores well and its neighbor two doors down does not, and that gap between neighbors, inside the same city and the same label, is where the actual opportunity for an individual owner sits.
Related Reading
More from Hillcane on this topic, plus the pages on the site that sit next to the work.
Frequently Asked Questions
Which market type scores best on visibility?
Cities coded campus yes with a high tourist index show the highest median in this study, 18 out of 40, against an overall census median of 16.0. That is a real but modest gap of two points above the baseline, not a large or decisive one. The overlap inside every cell is wide enough that a shop's own site and profile explain far more of its individual score than its city's market type.
Are these labels set per shop or per city?
Per city. Campus is a yes or no flag and tourist index is low, medium or high, both recorded at the city level in the ground truth data and inherited by every shop located there. A specialty roaster and a quick service kiosk in the same coded city both carry the same label even though their actual customer bases may differ significantly.
How many shops and cities does this study cover?
The combined census includes 1,109 shops across 264 distinct cities, all matched successfully to a city ground truth row, snapshot date September 15, 2026. By campus status, 528 shops sit in campus yes cities and 581 in campus no cities. By tourist index, 222 shops sit in high tourist cities, 421 in medium, and 466 in low, all counted September 21, 2026.
Does living in a campus town or tourist town cause higher visibility scores?
This study does not claim that, and the data cannot support a causal claim either way. It reports a correlation between a city level label and a shop level median score. The gap between the highest and lowest cell is three points out of forty, small enough that other explanations, including the shop's own choices, plausibly explain more of the difference than the city label does.
What is the overall median visibility score across the census?
The overall median is 16.0 out of 40, recounted September 21, 2026 across all 1,109 shops in 264 cities in the six region census. Every cell in the campus versus tourist comparison sits within a few points of it: 15 for both tourist low groups, 17 for campus no with tourist high, and 18 for campus yes with tourist high. That tight clustering is itself evidence that market type is a modest factor, not a dominant one.
Should an owner in a low tourist, no campus town assume their market is simply harder?
No. The median for that cell, 15, sits only one point below the overall census median of 16.0, and the spread of individual shop scores inside that cell is wide. Plenty of shops in that exact cell score well above their cell's median by working on the same basics, a maintained blog, complete schema markup and an accurate profile, that matter in every cell measured.
Is this the same study as the State of Specialty Coffee Visibility report?
No. The State of Specialty Coffee Visibility report covers the national picture from the same underlying census without cutting by market type. This post is a narrower cut that joins the 1,109 shops to city level labels, a campus yes or no flag and a low, medium, or high tourist index, to compare those groups. Read it alongside that report, and never mix its figures on the census scale of 40 with the separate Southeast audit's score bands.
Do blog and schema gaps show up across every market type?
Yes. The pattern of thin blog activity and incomplete schema markup recorded across the full census shows up inside every cell in this comparison, campus or not, tourist heavy or not. Favorable market type does not appear to insulate a city's shops from these baseline gaps, and unfavorable market type does not appear to guarantee them either.
What should a thin cell in this comparison be read as?
A thin cell, one with a small shop count, is described rather than treated as a firm conclusion. This study states the count wherever a comparison is drawn, because a pattern among a few dozen shops deserves less confidence than one among several hundred. The base groups are sizable: 528 campus yes and 581 campus no shops, and 222 high, 421 medium, and 466 low tourist index shops. Any cell too small was set aside.
How can Hillcane help a shop that sits in a lower scoring market type?
Hillcane can help build the fixes that matter more to an individual shop's score than its city's label: a working blog, complete schema markup, and an accurate, complete profile. In this data the widest median gap between market types is three points out of forty, 18 versus 15, and plenty of shops in the lowest cell already score above their median. Reach out at hillcane.co/contact or call (256) 384-2449 to see where your shop stands.
Work with Hillcane
A campus town shop and a resort town shop both assume their market decides their visibility score. The census says the gap between market types is three points out of forty, and the overlap inside each type is wider than the gap between them.
If you want to see where your shop actually stands against shops in similar and different market types, reach Hillcane at hillcane.co/contact or call (256) 384-2449.
Comments