States where "pop" wins in search
43/ 50
Share of measured states where "pop" phrases dominate.
States where "coke" wins in search
6/ 50
Share of measured states where "coke" phrases dominate.
States where "soda" wins in search
2/ 50
Share of measured states where "soda" phrases dominate.
What the data shows
The Midwest really does drink pop. Iowa, Michigan, Minnesota and Ohio are our top four "pop" strongholds, where more than 60% of all can of X and case of X searches use the word "pop". In Iowa it hits 68%.
The Deep South really does say coke. Alabama, Arkansas, Louisiana, Mississippi, Tennessee and Texas all flip to "coke" — despite the obvious brand overlap with Coca-Cola, which we filter out by using grocery-intent phrases.
But soda almost disappears. Only 2 states — California and Utah — show "soda" as the clear winner. This is the first real surprise: the entire East Coast soda belt that every dialect map calls "soda territory" doesn't show up that way in search.
New York searches like the Midwest. Despite 82% of New Yorkers saying "soda" in linguistic surveys, "can of pop" and "case of pop" out-search their soda equivalents by 2-to-1 statewide. The likely culprit: upstate pop-speakers outnumber Manhattan soda-speakers in commercial queries.
The map
Which word wins your state's search bar
Color opacity scales with confidence (margin between winner and runner-up). Hover any state for the full breakdown.
- 1Iowa68.4%
- 2Minnesota64.1%
- 3Michigan64.9%
- 4Ohio61%
- 5Indiana54.5%
- 1California36.3%
- 2Utah35.7%
- 1Louisiana50%
- 2Alabama50%
- 3Tennessee47.1%
- 4Arkansas42.9%
- 5Mississippi42.9%
Survey says X. Search says Y.
The states that broke the dialect map
Harvard's 2003 Dialect Survey asked Americans what they say. We measured what they type. 22 states split between the two. These are the biggest divergences — the places where commerce has already left linguistics behind.
51.2%
pop
25.6%
soda
23.1%
coke
43.5%
pop
23.5%
soda
32.9%
coke
36.7%
pop
36.7%
soda
26.5%
coke
37.5%
pop
25%
soda
37.5%
coke
37.8%
pop
32.4%
soda
29.7%
coke
36.4%
pop
30.3%
soda
33.3%
coke
38.7%
pop
35.5%
soda
25.8%
coke
41.9%
pop
25.8%
soda
32.3%
coke
36%
pop
36%
soda
28%
coke
Regional concentration
Pop owns the Midwest. Coke owns the South.
Winner: Pop · 3.3K monthly searches
Winner: Pop · 4.3K monthly searches
Winner: Pop · 4.7K monthly searches
Winner: Soda · 3.2K monthly searches
Where each word peaks
The six queries that prove dialect is real
- #1New York
"can of pop"480/mo - #2Michigan
"can of pop"390/mo - #3Ohio
"can of pop"390/mo - #4California
"case of pop"320/mo - #5Florida
"can of pop"260/mo - #6Illinois
"can of pop"260/mo
- #1California
"case of soda"320/mo - #2California
"can of soda"260/mo - #3Pennsylvania
"case of soda"210/mo - #4New York
"can of soda"170/mo - #5Texas
"can of soda"170/mo - #6Texas
"case of soda"170/mo
- #1California
"can of coke"320/mo - #2Texas
"can of coke"260/mo - #3New York
"can of coke"210/mo - #4California
"case of coke"170/mo - #5Florida
"can of coke"170/mo - #6Texas
"case of coke"170/mo
Every state, every number
The full 2026 state-by-state ranking
| State | Region | Search winner | Survey (2003) | Volume | Confidence | Shares |
|---|---|---|---|---|---|---|
| CA California | West | soda | soda✓ | 1,600 | 3.2% | |
| NY New York | Northeast | pop | soda✗ | 1,210 | 25.6% | |
| TX Texas | South | coke | coke✓ | 1,110 | 8.1% | |
| PA Pennsylvania | Northeast | pop | pop✓ | 900 | 13.4% | |
| FL Florida | South | pop | soda✗ | 850 | 10.6% | |
| IL Illinois | Midwest | pop | pop✓ | 840 | 20.2% | |
| OH Ohio | Midwest | pop | pop✓ | 820 | 41.5% | |
| MI Michigan | Midwest | pop | pop✓ | 740 | 46% | |
| NJ New Jersey | Northeast | pop | soda✗ | 490 | 0% | |
| GA Georgia | South | pop | coke✗ | 480 | 0% | |
| IN Indiana | Midwest | pop | pop✓ | 440 | 31.8% | |
| MN Minnesota | Midwest | pop | pop✓ | 390 | 46.2% | |
| NC North Carolina | South | pop | coke✗ | 370 | 5.4% | |
| VA Virginia | South | pop | coke✗ | 330 | 3.1% | |
| WA Washington | West | pop | pop✓ | 320 | 6.2% | |
| MD Maryland | South | pop | soda✗ | 310 | 3.2% | |
| MA Massachusetts | Northeast | pop | soda✗ | 310 | 9.6% | |
| AZ Arizona | West | pop | soda✗ | 250 | 0% | |
| WI Wisconsin | Midwest | pop | pop✓ | 250 | 4% | |
| MO Missouri | Midwest | pop | pop✓ | 230 | 4.3% | |
| CO Colorado | West | pop | pop✓ | 200 | 0% | |
| IA Iowa | Midwest | pop | pop✓ | 190 | 52.6% | |
| LA Louisiana | South | coke | coke✓ | 180 | 22.2% | |
| TN Tennessee | South | coke | coke✓ | 170 | 17.7% | |
| KS Kansas | Midwest | pop | pop✓ | 160 | 31.3% | |
| OR Oregon | West | pop | pop✓ | 160 | 12.5% | |
| KY Kentucky | South | pop | pop✓ | 150 | 13.4% | |
| OK Oklahoma | South | pop | coke✗ | 150 | 13.4% | |
| SC South Carolina | South | pop | coke✗ | 150 | 0% | |
| AL Alabama | South | coke | coke✓ | 140 | 21.4% | |
| CT Connecticut | Northeast | pop | soda✗ | 140 | 14.3% | |
| UT Utah | West | soda | soda✓ | 140 | 0% | |
| NV Nevada | West | pop | soda✗ | 130 | 0% | |
| NE Nebraska | Midwest | pop | pop✓ | 110 | 27.2% | |
| ND North Dakota | Midwest | pop | pop✓ | 80 | 25% | |
| SD South Dakota | Midwest | pop | pop✓ | 80 | 25% | |
| WV West Virginia | South | pop | pop✓ | 80 | 25% | |
| AR Arkansas | South | coke | coke✓ | 70 | 14.3% | |
| MS Mississippi | South | coke | coke✓ | 70 | 14.3% | |
| AK Alaska | West | pop | soda✗ | 60 | 0% | |
| DE Delaware | South | pop | soda✗ | 60 | 0% | |
| DC District of Columbia | South | pop | soda✗ | 60 | 0% | |
| HI Hawaii | West | pop | soda✗ | 60 | 0% | |
| ID Idaho | West | pop | pop✓ | 60 | 0% | |
| ME Maine | Northeast | pop | soda✗ | 60 | 0% | |
| MT Montana | West | pop | pop✓ | 60 | 0% | |
| NH New Hampshire | Northeast | pop | soda✗ | 60 | 0% | |
| NM New Mexico | West | pop | soda✗ | 60 | 0% | |
| RI Rhode Island | Northeast | pop | soda✗ | 60 | 0% | |
| VT Vermont | Northeast | pop | soda✗ | 60 | 0% | |
| WY Wyoming | West | pop | pop✓ | 60 | 0% | |
| WV West Virginia | South | — | pop✗ | 0 | 0% |
Methodology
How the dialect signal was isolated
The problem with searching "pop": it matches K-pop, pop music, "pop it" toys, and soft drinks — all at once. The same noise hits "coke" (Coca-Cola brand + slang) and "soda" (baking soda + the drink). Generic searches are useless for dialect research.
Our fix: query Google Ads for can of [word] and case of [word]. These phrases almost exclusively fire when someone is buying soft drinks and reaching for the word they use in conversation. Pure dialect, no brand noise.
The ground truth: Bert Vaux's 2003 Harvard Dialect Survey asked 30,000+ respondents which word they use. We treat that as linguistic ground truth and compare it against 2026 search volume.
What this measures: not what Americans say to each other, but what they type when buying. Those two things agree in 56.9% of measured states — which means that in the remaining 43.1%, commercial behavior has already started to diverge from regional dialect. That's the story most dialect maps miss.
Data source
Google Ads
Historical Metrics · state-level geo targets
Keywords measured
9 phrases × 51 states
= 459 data points
Total monthly volume
15,510
Dialect-sensitive queries only
Questions I get about this map
Real questions from linguists, journalists, and ad buyers who've looked at the data.
How is a search-volume map different from a survey of soda dialects?+
The classic 2003 Harvard dialect survey asked people which word they say. Search volume measures which word they actually type when they want one. The two diverge in interesting places — entire states have switched dialects between 2003 and 2026, and the search map catches it where memory does not. We query Google Ads live for all 50 states; 43 states pick "pop", 6 pick "coke", 2 pick "soda".
Why does "coke" still win in the South when most people there don't say "Coke" anymore?+
Two reasons. First, a lot of search intent is for the brand Coca-Cola itself (where to buy it, calorie content, recipes), so it bundles with the dialect signal. Second, the dialect is sticky in older cohorts and in rural areas with slower demographic turnover. Survey memory lags search behavior.
How were the per-state winners calculated?+
For each state we pulled Google Ads monthly search volume on three queries: "pop", "soda", "coke" (lowercase, no brand qualifier). Highest-volume term wins. Ties are broken on the next-most-relevant variant (e.g., "soda pop", "fountain coke"). Methodology section at the bottom shows the full filter.
Are the volumes a real population signal or biased by tourists?+
Google Ads volume is location-targeted by the IP/profile of the searcher, so tourist queries do show up in the visited state. We normalize per state population to reduce that effect; the absolute volumes are still useful for ad-buyer planning where tourist intent counts.
How current is the search data?+
Pulled in the month before publication via Google Ads historical metrics. The 12-month rolling window catches seasonal swings (summer pulls soda up, holiday pulls coke up). The map shows the rolling annual winner per state, not the spot-month winner.
Can I cite this in my article or research?+
Yes — the dataset is released under CC-BY 4.0. Attribution to "Sextaris — Pop vs Soda vs Coke search map (2026)" with a link to this article is enough. The downloadable CSV is on the roadmap; reach out via the contact page if you need the raw numbers now.
Want to run a search map like this — for your keyword?
Sextaris queries Google Ads for 234 countries, 51 states, and 19,656 US cities in a single run. One credit, one minute — the same infrastructure that built this article.
Share this research