E-commerce
22 May 2026
What percent of retail sales are e-commerce? In English: what percentage of retail trade goes through the e-commerce? There is no universal figure that applies to all regions, all years, and all definitions. Published percentages depend on the country, the statistical scope (goods, services, households, vehicles, etc.), and the boundary between online sales and other channels.
This guide helps you read a percentage without getting it wrong: what to check in the methodology note, how to reconcile macro and store-level figures, and why your intra-site rate (share of the web in your revenue) can be very different from the national average.
We do not display here a fixed “official” percentage by year: the series are revised and the definitions change. For precise figures, open the source publication (national institute, Eurostat, federal agency, etc.) and note the edition date. For your management: e-commerce analytics, GA e-commerce tracking, Shopify reports.
Three points before going further. First: “retail” and “e-commerce” do not mean the same thing depending on whether we are speaking colloquially or about a statistical aggregate. Second: a retail percentage share can increase because the numerator (online) goes up, because the denominator (total retail) goes down, or both. Third: comparing two countries without harmonizing the definition creates an illusion of precision.
If you are in a hurry: remember that the useful question for your business is often “what share of my revenue do I want to attribute to digital over three years, and how do I measure it internally?” rather than “what is the worldwide percentage”.
A simple tip: when a news headline cites a percentage, look in two clicks for the chart caption; half the misunderstandings disappear there.
Finally, distinguish the share of e-commerce in retail as measured by public statistics and the share of online in your own sales; the two have no reason to be equal.
Executives sometimes use a macro percentage as a rhetorical argument in a meeting without needing it to decide; keep the proportion between time spent on macro monitoring and time spent on your cohorts aligned with your size.
If you teach or train teams, a useful exercise is to ask them to write down by hand the numerator and denominator before commenting on a chart found online.
Sector case studies (fashion, food, electronics) often show massive gaps under the same national figure; that is normal and expected.
For the link between online retail and brick-and-mortar in the broad sense: how e-commerce works, omnichannel, marketplace and own catalog.
Summary
Why there is no single "world percentage"
You can’t answer « 37% everywhere » without lying by omission. Organizations publish ranges and series; journalists extract an average or a year from them; readers forget the scope.
Why the question keeps coming up
E-commerce is visible, measurable in panels, and useful for illustrating a long-term trend in the digitalization of consumption.
What a share really measures
A share = a numerator divided by a denominator; change either one, and the curve shifts without the « sensational » result being the same.
Reasonable expectations
In meetings, first ask which series you’re citing, not just the isolated percentage.
Students and young brands sometimes confuse e-commerce’s share of GDP and share of retail; these are different denominators.
If your board asks you for « the industry figure », prepare a slide with at most two sources and their definitions side by side rather than ten numbers without context.
Read a methodology note in four minutes
Before reusing a percentage, read four lines of methodology: population (households, businesses...), included goods, how online sales are defined, frequency, and revisions.
E-commerce in the statistical sense
Often: an order placed through an internet channel (website, app...), regardless of delivery; sometimes exclusions (certain services, fuel, vehicles) depending on the country.
Total retail
The denominator may exclude major items; a percentage “high” or “low” only makes sense with that scope.
Current prices versus volumes
A share in value (€, $) is not the same as a share in volume; inflation shifts shares in value: price and margin.
The notes sometimes indicate whether the figures are seasonally adjusted or raw; comparing a raw month with an adjusted average distorts the interpretation.
For international teams, keep a “denominator definition” column in your monitoring table; otherwise you compare countries over ten years without realizing it.
Countries, blocs and cautious comparisons
Each country or bloc publishes its own tables; the harmonizations (Europe, OECD) exist but do not erase all the nuances.
Europe
Eurostat and national institutes publish indicators on internet sales; the detail in the methodological notes takes precedence over the chart highlighted by a media outlet.
United States
Federal and state agencies publish retail and e-commerce series; beware of later revisions.
Hasty comparisons
Two percentages side by side in a country comparison article without a methodological equivalence table: be wary.
Reports from trade associations can be useful for the trend but do not replace official series for an internally «defensible» figure.
If you have to present to investors, prefer «range according to source X, year Y, scope Z» to a single round figure without a reference.
Mainstream media like headlines «record» or «collapse»; the long series is often more flat and more useful for a strategy note.
When a study compares several countries on the same timeline, ask whether the countries have all published their data for the same full calendar year or whether some points are extrapolated.
Channels: the "online" detail is sometimes unclear
The physical store, the website, the marketplace, social commerce and in-store pickup after web ordering sometimes blur the “pure online” picture: social channels, enabling social sales.
Click and collect
The order is digital; part of the revenue can be attributed to the store according to the retailer’s framework: point of sale.
Marketplaces
Revenue for a brand can be strong on platforms; the national “e-commerce” aggregate may classify things differently depending on the organizations: Amazon.
Services
Subscriptions, SaaS, mixed services: depending on the statistics, they sometimes fall outside strict retail even though they are “online”.
Your internal dashboard may be more granular than the national aggregate; that is not an error, it is another level of analysis.
A hybrid retailer that promotes store pickup may see a high web-ordered share while public statistics break it down differently; document your own definition for the board of directors.
Reference year, revisions and peaks
A year in which the e-commerce share « goes up » in retail may reflect a low comparison base, a change in methodology, or a real acceleration; all three coexist in recent history depending on the country.
Revisions
Figures in initial releases may differ from final series; cite the version of the table.
Exceptional events
Periods of transition to online have temporarily distorted curves; the longer view then smooths out part of the effect.
Normalization
After a peak, stabilization can look like a « slowdown » even though the level remains high relative to the distant past.
Entrepreneurs sometimes use these cycles to justify a digital investment plan; align the narrative with your own order history, not just with a magazine cover.
For the commercial calendar, compare the same windows (sales, holidays) year over year before inferring a macro trend.
Your store is not the national average
Your digital share of revenue can be twenty points above or below a national average without your strategy being "better" or "worse"; market positioning matters.
Internal indicators
Web revenue, web orders, average basket by channel, mobile share: average basket, mobile.
Conversion
A favorable macro trend does not increase your conversion; conversion, industry benchmarks.
Tag consistency
If your "web" sales exclude marketplaces in one report and include them in another, your own share will jump artificially.
Pure players readily compare their growth to a flat national average; brick-and-mortar retailers do the opposite; both have perspective bias.
Set an internal annual review: channel share, margin after returns by channel, acquisition cost by channel; the macro percentage does not replace this table.
Decision: macro as context, micro as a lever
Using a macro percentage to decide without your own numbers leads to misaligned budgets: inventory, advertising, hiring.
Scenarios
Low, base, high for your internal web share over twelve to twenty-four months, regardless of the national average.
Acquisition
If digital grows everywhere, ad competition can also rise: marketing costs, traffic and conversion.
Inventory
Anticipate spikes if your line follows national online seasonality: inventory.
A common mistake: extrapolating “everyone buys online” to justify a doubling of SKUs without testing demand by item.
For pitch decks, a slide “macro + our trajectory” with two distinct curves avoids the objection “you’re confusing the market and your store”.
A complementary indicator often overlooked: the share of households that have already ordered online at least once during the year, when a survey publishes it; it is not the same indicator as the retail share in value terms, but it helps calibrate the user experience.
Data teams can cross-check press claims with a pull API from Eurostat or an institute when a stable endpoint exists; invest half a day once to automate the quarterly update if the topic is critical for you.
Private reports versus public series
Private actors sometimes publish estimates of market size. They can be useful for a trend, not for an accounting entry.
Commercial scope
B2B, B2C, logistics services, retail media advertising: a report can broaden the definition of « digital retail » without saying so explicitly.
Samples
Merchant panels, consumer surveys: useful but subject to response bias.
Cross-checking
When a private study and a public series diverge, note why (scope, year, constant prices or not) before drawing conclusions.
Marketing teams like big « eye-catching » numbers; finance prefers reportable series; a compromise is to cite two sources transparently.
If you integrate third-party data into a data warehouse, document the license and extraction date to avoid phantom reports.
Communicate the numbers without contradicting yourself
When you present a change in e-commerce share, include three elements: definition, source, period. Without that, a curve may “speak” but does not allow action.
Visualization
Same scale, same year baseline if comparing across countries; otherwise the chart can make a false idea look convincing.
Uncertainty
Mention intervals or revisions when the source provides them; caution strengthens internal credibility.
Action
Link the statistic to a decision: SEO budget, marketplace, pilot store; roadmap, funnel.
Committees that spend twenty minutes on a macro percentage and then zero minutes on the internal average basket size invert the order of operational priorities.
Archive screenshots of source tables with URL and date; institutional links sometimes change structure.
If you present a percentage in a CSR or annual report, specify whether the communication targets shareholders or internal teams; the acceptable level of detail differs; unions or works councils may question a macro trend used to justify a logistics reorganization plan.
For joint ventures or co-brands, agree in advance on a common definition of the “digital” channel in shared tables; two Excel joins with different definitions produce incompatible percentages at month-end.
Qstomy: customer experience without brittle statistics
Understanding e-commerce’s share of retail helps size up customer questions: delivery, availability, promised lead times. An assistant can keep a reassuring tone about the experience without citing shaky statistics.
Qstomy for Shopify: AI chatbot, automated customer support, customer experience.
Links: demo, offers, assisted selling, support, analytics. Your macros should avoid unsourced numerical claims about market size.
Even without AI, a stock phrase like « we track the lead times announced on the product page » is better than a made-up percentage about buying habits.
Measure tickets related to delivery expectations after major promotional periods; it’s often more informative than a macro figure on e-commerce.
Go further: LTV, margins, SEO
To extend the thinking around measurement and performance without locking yourself into a single market share indicator:
Retention
Digital’s share in macro retail says nothing about your LTV: customer retention, CAC and LTV.
Profitability
Models and margins: e-commerce models, profitability of an e-commerce business.
SEO
Capturing informational demand for your categories: e-commerce SEO, organic traffic.
Building a “macro” watch alongside a “direct competitors” watch helps avoid overreacting to a national figure when your niche is decoupled.
Early-stage founders benefit from spending more time on their customer cohort than on slide three of a global report.
Summary, FAQ, and Further Reading
In brief
No single percentage sums up « online commerce in retail » without a country and without a definition.
The denominator and the numerator change everything; read the note.
Your internal metrics determine budgets, not the national average alone.
Prefer dated official sources to recycled round numbers.
FAQ
What is the exact percentage of e-commerce in global retail?
There is no single reliable answer without a scope; consult organizations that publish aggregates with methodology.
Why do two articles give two different percentages for the same year?
Definitions, revisions, seasonal adjustment, or translation error in the chart.
Should I align my strategy with the national average?
No: your positioning and your competitive advantage come first; macro serves as context.
Is the mobile share the same as the e-commerce share?
No; mobile is often a subset of the online channel.
Do marketplaces inflate national statistics?
They can influence the aggregates depending on how the agencies classify the flows; read the notes.
How should I present this to an investor who demands numbers?
Provide a sourced range + your measured trajectory + explicit assumptions.
For further reading
Enzo
22 May 2026
