China city opportunity index
One row per Chinese city or prefecture-level unit held in our own city table. It exists to answer a narrow question honestly: which cities do we actually have a profile for, and which ones are we still missing?
391 city records · 358 carry an industry profile · 40,499 place records
Across 34 province-level divisions. Snapshot 2026-09-29; every figure below is recomputed from the published file.
The numbers
| Value | What it counts | Source / definition |
|---|---|---|
| 391 | city rows | One row per city or prefecture-level unit in our own city table. Source: this dataset's records array. |
| 358 | cities with an industry profile | Rows where data_status is a scored profile rather than missing. That is 91.6% of all rows. |
| 34 | province-level divisions covered | Distinct values in the province field across the 391 rows. |
| 40,499 | place records | Places held against these cities, summed from the poi_rows field. Median 75 per city, mean 103.6. |
| 388 | cities with place records | Cities with at least one place on file. Three cities have none. |
| 389 | rows with an administrative code | Rows carrying an region_adcode. The remaining two do not. |
What one record is
One record is one city or prefecture-level unit. It carries the city name, its province, its administrative level, a short industry list, a data status, the number of places on file, and an administrative code where we have one. A record can exist while its industry profile does not; those rows are kept and labelled rather than dropped.
Records
- The gap is entirely county-level. All 33 rows without an industry profile are county-level units: 21 counties, 11 county-level cities and 1 forestry district. Not one prefecture-level city is missing — all 312 are present. So the missing 8.4% is a level-of-government artefact, not a regional blind spot.
- Place records are heavily skewed. Beijing holds 975 place records, 13.0× the median of 75. Three cities sit at zero. Any average read off this field will be pulled upward by a handful of large cities.
- Two rows still lack an administrative code. 389 of 391 rows carry one. That is a small integrity gap we publish rather than round away.
- Why we publish it. Coverage claims are easy to make and hard to check. This page states exactly how many cities we hold, how many carry a profile, and which kind are missing — so the number can be checked against the file rather than taken on trust.
What this cannot tell you
- No economic indicators. Population, GDP and growth are being re-collected and are not published in this snapshot. This page says nothing about how large or how fast-growing any city is.
- A profile is not a ranking. An industry list marks what a city is recorded as doing, not how good the opportunity is. The scoring that sits on top of this data is published separately and can be argued with.
- The counts describe our table, not China. 391 rows is what we hold, not a claim about how many cities exist.
Data & methods
- Source. Our own city database, with place counts joined from our place records. Not a third-party dataset, and not collected by a partner on our behalf.
- Snapshot. 2026-09-29 19:49 (Asia/Shanghai). This is a snapshot, not a live figure.
- Fields. city, province, level, industries, data_status, last_updated, poi_rows, region_adcode — the published file carries the same definition used above.
- Licence. CC BY-NC 4.0 — attribution: InfoGap (infogap.store), link to this page.
Where to go next
- Download: china-city-index.json
- Related data page: City × industry matrix
- Method page: How to check our numbers
- Hub: https://infogap.store/data/
- Machine-readable: registry.json