Skip to content
Back to the report

Methodological annex

Estimating e-commerce penetration in Argentina

Report, pages 63 to 64

Why an estimate is needed

The share of e-commerce in total retail sales in Argentina has no direct estimate comparable to those published by statistical offices or industry chambers in other countries (such as the ONS in the United Kingdom or the Census Bureau in the United States). To build an approximate value, a three-stage procedure was followed.

Step by step

  1. DataCOU · INDEC

    The retail share of the trade sector

    First, the share of retail sales in the total gross output (GO) of the trade sector was obtained from the Supply and Use Tables (SUT) published by INDEC. For the 2018-2023 period, the SUT for each available year was used. This ratio reflects the structural composition of the sector across its wholesale, retail and vehicle components.

  2. DataARCA

    Formal retail sales

    Second, that ratio was applied to the annual revenue of the trade sector declared to ARCA (formerly AFIP) to obtain an estimate of the level of formal retail sales in each year.

  3. DataCACE

    The ratio to e-commerce

    Third, the ratio was computed between the annual e-commerce revenue reported by the Argentine E-commerce Chamber (CACE) and the estimate of retail sales from the previous step.

  4. DataCuentas Nacionales · INDEC

    2024 and 2025

    For 2024 and 2025, since ARCA sales data are not yet available, the average ratio between GO and sales observed in previous years was used to estimate, based on INDEC's National Accounts, a plausible invoiced value for retail trade. E-commerce penetration for those years was calculated from that estimated denominator. The resulting values should be treated with more caution than those for the 2018-2023 period.

The calculation, in a diagram

E-commerce penetration is annual e-commerce revenue (CACE) divided by estimated retail sales. That estimate is the annual revenue of the trade sector (ARCA) multiplied by the share of retail sales in the trade sector's GO (SUT · INDEC).

The annex describes the calculation, but does not publish its intermediate values.

Main assumptions

The calculation rests on the following assumptions.

  1. First, that the wholesale/retail composition of the SUT is a reasonable approximation of the structure in each reference year.
  2. Second, that the revenue declared to ARCA adequately represents the universe of formal retail trade.
  3. Third, that the revenue reported by CACE captures all, or the vast majority, of e-commerce transactions.
  4. Fourth, that for 2024 and 2025 the historical average of the GO/sales ratio is an acceptable proxy in the absence of tax data.

Limitations and biases

  • Effect on the resultOverstates

    The denominator does not include informal trade

    13.82%Upper bound of penetration in the formal segment, 2023

    The main bias identified is an overestimation of penetration. The denominator (retail sales) is built exclusively on tax records and therefore excludes informal trade: street vending, street fairs, unregistered shops and the under-declared portion of formal shops. Since e-commerce transactions are traceable by nature and are mostly recorded, the numerator has proportionally greater coverage than the denominator, which pushes the ratio up. The resulting value (13.82% for 2023) should be read as an upper bound of penetration in the formal segment and as an overestimate of penetration in total retail trade.

  • Effect on the resultIn the opposite direction

    Informal e-commerce that CACE does not capture

    In the opposite direction, there is a volume of informal e-commerce (sales through social networks, messaging and platforms without invoicing) that CACE does not capture, which would reduce the numerator. However, this effect is unlikely to offset the scale of informality in brick-and-mortar retail in Argentina.

  • Effect on the resultUncertain direction

    The SUT ratio may be outdated

    In addition, the ratio taken from the SUT may not reflect recent changes in the composition of the sector, for example the relative growth of wholesale self-service stores aimed at final consumers, which would introduce an error whose direction depends on how each component specifically evolves.

  • Effect on the resultLess robust

    2024 and 2025, with a historical average

    This limitation is more pronounced for 2024 and 2025, where the lack of tax data makes it necessary to rely on a historical average that might not capture recent structural changes.

Where this calculation is used

These charts from the report show the estimate. Each one carries its methodological note.

  • Chart 2

    E-commerce as a share of total retail sales by country, latest available year (percent)

    Note: the estimate for Argentina tends to overstate the penetration of the digital channel in retail sales. It was built from the weight of the retail channel in total trade and from the sales recorded by AFIP in 2023. In other words, informal trade at physical stalls is left out of the calculation. For more detail, see the methodological note in the Annex.

    Source: UNCTAD, Banco Central de Chile, INDEC, ARCA y CACE, Contrabando: el rol del Estado y de las plataformas de e-commerce, 2026

    See the chart

  • Chart 5

    Share of the digital channel in total Argentine retail trade, 2018-2025

    Note: the data shown for 2024 and 2025 are a less robust estimate than earlier years, since the level of sales recorded by ARCA for those years is not yet available. For that reason, there are discrepancies between the reported level and the overall movement of the e-commerce channel relative to in-store sales.

    Source: CACE, ARCA, INDEC, Contrabando: el rol del Estado y de las plataformas de e-commerce, 2026

    See the chart

The annex, in the report

The methodological annex is on pages 63 to 64 of the report.

Download the report

PDF · 64 pages · 2.8 MB

Text size