Reading import partner concentration before you pick a market
Two markets can have the same import value and completely different entry prospects. The number that tells
them apart is not the total, it is **how concentrated the suppliers are**.
Below: the same HS code, two countries, and what the difference implies. Every figure is from UN Comtrade
(reporter-reported, 2023) and can be re-run in one API call.
The two markets
| | Indonesia | Vietnam |
|---|---|---|
| HS 2106 imports, 2023 | **USD 703,645,643** | **USD 1,060,567,534** |
| Top supplier | China 156,669,628 (**22.27%**) | Singapore 260,939,390 (**24.60%**) |
| Second | Singapore 103,182,165 (14.66%) | China 86,687,282 (8.17%) |
| Third | Malaysia 91,048,707 (12.94%) | Malaysia 54,911,155 (5.18%) |
| Fourth | Korea 72,121,246 (10.25%) | Korea 46,944,084 (4.43%) |
| Fifth | Thailand 43,370,510 (6.16%) | Thailand 39,272,330 (3.70%) |
| Top five combined | **66.28%** | **46.08%** |
| Left to everyone else | **33.72%** | **53.92%** |
Vietnam's market is larger **and** less concentrated. That combination matters more than the headline size:
a long supplier tail usually means buyers already switch suppliers routinely, which is the behaviour a new
entrant needs in order to get a first order at all.
Why the tail matters
A supplier holding a third of the market is defending an account base built over years. A market where no
single supplier holds more than a quarter, and where the top five control less than half, has more room for
a differentiated offer — even a small one.
This is the part of the analysis you cannot get from a market-size headline, and it is also the part most
decks omit, because "USD 1.06bn market" sounds better than "46% concentrated, 54% fragmented".
Three traps in this calculation
1. **Adding the World row to the partner rows.** The response contains an aggregate row (`partnerCode = 0`)
equal to the sum of all single-partner rows. Adding both doubles the total. It is the most common
way this analysis goes wrong.
2. **Reporting shares against the wrong denominator.** After any row is dropped or filtered, every share
must be recomputed against the same total. In our own first attempt a filter shrank the denominator and
pushed China's share from 22.27% to 26.65% — a 20% relative error, invisible unless you re-check.
3. **Treating mapped names as if the API supplied them.** The Comtrade payload carries partner *codes*
without names. Names are a lookup. Any table that shows names should say where they came from.
What this cannot tell you
- **Who the buyers are.** Comtrade reports country-to-country flows, not companies.
- **Prices and margins.** Not in the data.
- **Whether the tail is addressable by you.** A fragmented market of 60 small importers may still be
concentrated in one city or one channel.
If a figure is unavailable — for example the prior year returns HTTP 429 — the honest output is
*not available*. An interpolated growth number is worse than a blank cell, because it invites trust.
Get the underlying numbers
We publish these tables as a spreadsheet: total with its derivation, partner codes and shares, macro
indicators with their years, and a sources sheet with every URL and every gap.
- Free samples: [Indonesia](/samples/Indonesia_HS2106_Trade_Data.xlsx) ·
[Vietnam](/samples/Vietnam_HS2106_Trade_Data.xlsx)
- Order a pack for your own country and HS code: [verified trade data packs](/data-packs.html)
### Examples with the partner tables in full
The concentration figures above come from the same source as these pages; each one lists the top five
partners under their official Comtrade names, with shares:
- [Thailand, HS 2106, 2023](/import-data/tha-hs2106-2023.html)
- [Viet Nam, HS 2106, 2023](/import-data/vnm-hs2106-2023.html)
- [Thailand, HS 8471, 2023](/import-data/tha-hs8471-2023.html)
- [Mexico, HS 8471, 2023](/import-data/mex-hs8471-2023.html)
Need this data as a spreadsheet you can check line by line?
Verified trade data packs — every figure carries its source URL →