eCommerce Statistics, With the Source Attached to Every Number
Most eCommerce statistics pages are a list of numbers copied from another list of numbers, with no source, no year and no indication of what was measured. Every figure here names its original source, when it was published, and the geography or sample behind it — because a statistic you cannot check is not evidence, it is decoration.
FIGURES
23 figures
SOURCES
9 primary sources
LAST REVIEWED
September 2026
NEXT REVIEW
March 2027
Every number on this page links to the organisation that produced it.
- The standard
What has to be true before a number appears here
Four conditions. A figure that fails any of them is left out, however useful it would have been.
A statistic without its method is an opinion with a decimal point.
The rule we apply is simple: if a journalist, analyst or client could not verify the number from what is on the page, it does not belong on the page. That removes most of what circulates as eCommerce statistics, which is why this list is shorter than the ones that promise a hundred.
- A named original source, linked — not an article citing an article
- A publication or measurement date, stated next to the figure
- The geography and the sample or method it was measured on
- A stated limitation: what the number does not tell you
- A scheduled review date, so stale figures get replaced rather than quietly aging
Where a figure is older than we would like — the mobile share below is from 2024 — we say so rather than presenting it as current.
- Market size
How large online retail actually is
The most quoted numbers in eCommerce, taken from the government statistics office that produces them rather than from a secondary summary.
$340.2bn
in US retail e-commerce sales in a single quarter, seasonally adjusted.
SOURCE
U.S. Census Bureau, Quarterly Retail E-Commerce Sales, 2nd Quarter 2026
Released 18 August 2026 · United States · Q2 2026 · seasonally adjusted
17.1%
of all US retail sales were made online in the second quarter of 2026.
SOURCE
U.S. Census Bureau, Quarterly Retail E-Commerce Sales, 2nd Quarter 2026
Released 18 August 2026 · United States · Q2 2026 · share of total retail
12.2%
year-on-year growth in US e-commerce, against 6.7% for total retail. Online is still taking share.
SOURCE
U.S. Census Bureau, Quarterly Retail E-Commerce Sales, 2nd Quarter 2026
Released 18 August 2026 · United States · Q2 2026 vs Q2 2025 · ±0.9% at 90% confidence
3.8%
quarter-on-quarter growth in US e-commerce sales, adjusted.
SOURCE
U.S. Census Bureau, Quarterly Retail E-Commerce Sales, 2nd Quarter 2026
Released 18 August 2026 · United States · Q2 2026 vs Q1 2026 · ±0.4% at 90% confidence
- By geography
Why the share of retail figure depends on where you ask
The US and Great Britain publish online share figures that differ by eleven percentage points. Almost all of that gap is definitional rather than behavioural.
28.3%
of total retail sales in Great Britain were made online in July 2026 — a very different figure from the US, for definitional reasons explained below.
SOURCE
Office for National Statistics, Retail sales, Great Britain: July 2026
Released 21 August 2026 · Great Britain · July 2026 · share of retail sales by value
29.2%
was the equivalent figure a month earlier. Online share moves month to month, which is why a single month is a weak headline.
SOURCE
Office for National Statistics, Retail sales, Great Britain: July 2026
Released 21 August 2026 · Great Britain · June 2026 · share of retail sales by value
6.5%
year-on-year growth in the value of online spending in Great Britain.
SOURCE
Office for National Statistics, Retail sales, Great Britain: July 2026
Released 21 August 2026 · Great Britain · July 2026 vs July 2025 · value, not volume
- Read carefully
Why two official figures disagree
Before quoting an online share figure, check which of these you are quoting.
Question
U.S. Census Bureau
ONS (Great Britain)
Headline figure
17.1% of retail sales
28.3% of retail sales
Period
Q2 2026
July 2026
Geography
United States
Great Britain, excluding Northern Ireland
What counts as retail
Retail trade as defined by the Census Bureau
Retail sales as defined by the ONS, including a different mix
Seasonal adjustment
Adjusted for season and trading days
Headline proportion published monthly
Measures
Sales value
Sales value
Safe to compare directly?
No
No
Both are correct. They answer slightly different questions, which is why any article presenting one as the global eCommerce share is wrong before it starts.
- How people shop
Device, conversion and the arrival of AI referrals
Three different measurement approaches: platform data, aggregated trading data and analytics data. Each is labelled so you can weigh it.
66%
of global online orders were placed on a mobile device, from 78% of traffic. Mobile carries more browsing than buying.
SOURCE
Salesforce, Shopping Index, Q1 2024
Published 23 April 2024, covering Q1 2024 · Global · Q1 2024 · 1.5bn shoppers across 67 countries
64%
was the equivalent US mobile order share, from 75% of traffic.
SOURCE
Salesforce, Shopping Index, Q1 2024
Published 23 April 2024, covering Q1 2024 · United States · Q1 2024 · Salesforce platform data
2.26%
average conversion rate across UK and Irish stores on one platform, up from 1.94% a year earlier.
SOURCE
IRP Commerce, eCommerce Market Data, July 2026
July 2026 · UK & Ireland · July 2026 · one platform’s merchant base
393%
year-on-year growth in AI-sourced traffic to US retail sites in the first quarter of 2026.
SOURCE
Adobe Analytics, Quarterly AI Traffic Report, Adobe Digital Insights
Published April 2026, covering Q1 2026 · United States · Q1 2026 · 1tn+ visits analysed
42%
higher conversion from AI-referred visitors than from other traffic — a reversal from a year earlier, when they converted at half the rate.
SOURCE
Adobe Analytics, Quarterly AI Traffic Report, Adobe Digital Insights
Published April 2026, covering Q1 2026 · United States · March 2026 · Adobe Analytics customers
37%
higher revenue per visit from AI-referred sessions than from other sources.
SOURCE
Adobe Analytics, Quarterly AI Traffic Report, Adobe Digital Insights
Published April 2026, covering Q1 2026 · United States · Q1 2026 · Adobe Analytics customers
- Where revenue leaks
Abandonment and returns
The two largest sources of revenue that looks like it arrived and did not.
70.22%
average documented cart abandonment rate, averaged across 50 separate studies.
SOURCE
Baymard Institute, Cart Abandonment Rate Statistics
Aggregate of studies 2006–2025, retrieved 2026 · Mixed geographies · studies 2006–2025 · rolling average
40%
of shoppers who abandoned cited extra costs — shipping, tax and fees — as the reason. It is the single largest stated cause.
SOURCE
Baymard Institute, Cart Abandonment Rate Statistics
Aggregate of studies 2006–2025, retrieved 2026 · United States · excludes browsing-only abandonment
18%
abandoned because the site required them to create an account.
SOURCE
Baymard Institute, Cart Abandonment Rate Statistics
Aggregate of studies 2006–2025, retrieved 2026 · United States · excludes browsing-only abandonment
19.3%
of online purchases were returned, against 15.8% across all retail.
SOURCE
National Retail Federation, Consumers Expected to Return Nearly $850 Billion in Merchandise in 2025
Published 15 October 2025 · United States · 2025 · survey of 2,006 consumers and 358 merchants
$849.9bn
in merchandise expected to be returned across US retail in a single year.
SOURCE
National Retail Federation, Consumers Expected to Return Nearly $850 Billion in Merchandise in 2025
Published 15 October 2025 · United States · 2025 forecast · NRF with Happy Returns
20%
abandoned because delivery was too slow — the second most cited reason after cost.
SOURCE
Baymard Institute, Cart Abandonment Rate Statistics
Aggregate of studies 2006–2025, retrieved 2026 · United States · excludes browsing-only abandonment
Cart abandonment is covered in more depth, with what to do about each stated reason, on the cart abandonment statistics page.
- Platform scale
What the largest platforms report
Company-reported figures from regulatory filings and earnings releases, which are the only numbers in this category worth quoting.
$115.6bn
in merchant sales processed through Shopify in one quarter, up 32% year on year.
SOURCE
Shopify Inc., Second Quarter 2026 Results, filed with the SEC
Released 5 August 2026 · Global · Q2 2026 · GMV, not Shopify revenue
34%
year-on-year revenue growth for Shopify itself in the same quarter, to $3.58bn.
SOURCE
Shopify Inc., Second Quarter 2026 Results, filed with the SEC
Released 5 August 2026 · Global · Q2 2026 · company revenue
$200.6bn
in Amazon net sales for the quarter, up 20% year on year across all its businesses.
SOURCE
Amazon.com, Inc., Amazon Q2 2026 earnings report
Released 30 July 2026 · Global · Q2 2026 · includes AWS and non-retail
26%
year-on-year growth in Amazon advertising, the fastest-growing part of the marketplace economics sellers pay into.
SOURCE
Amazon.com, Inc., Amazon Q2 2026 earnings report
Released 30 July 2026 · Global · Q2 2026 · advertising services
- The sources
Every source behind this page
What each organisation measured, how, and what it does not cover. If you cite a figure from this page, cite the original rather than us.
1
U.S. Census Bureau
Published: Released 18 August 2026
Method: Official government estimate from the Monthly Retail Trade Survey, seasonally adjusted for holiday and trading-day differences. Confidence intervals stated at 90%.
What it does not cover: United States only. Measures retail e-commerce, so it excludes most services and business-to-business trade. Not adjusted for price changes.
2
Office for National Statistics
Published: Released 21 August 2026
Method: Monthly Business Survey of retailers in Great Britain, reporting internet sales as a proportion of total retail sales by value.
What it does not cover: Great Britain only, and it excludes Northern Ireland. Measures retail sales value rather than order volume, and a different definition from the US figure, which is why the two are not comparable.
3
Baymard Institute
Published: Aggregate of studies 2006–2025, retrieved 2026
Method: Rolling average of 50 separate published studies, alongside their own programme of 25 rounds of usability testing across 4,400+ participant sessions and 12 quantitative studies totalling 20,240 participants.
What it does not cover: An average across studies with different definitions of a cart and different sectors. Treat it as an order of magnitude, not as a target for your own store.
4
Adobe Analytics
Published: Published April 2026, covering Q1 2026
Method: Analysis of more than one trillion visits to US retail sites and more than 100 million SKUs, plus a survey of 5,000+ US respondents in March 2026.
What it does not cover: US retail sites using Adobe Analytics, so it skews toward larger retailers. AI-referred traffic is still a small share of total traffic despite the growth rate.
5
Salesforce
Published: Published 23 April 2024, covering Q1 2024
Method: Activity of more than 1.5 billion unique global shoppers across 67 countries, drawn from Salesforce Commerce, Marketing and Service Cloud data.
What it does not cover: The oldest figure on this page, and due for replacement at the next review. Covers merchants on Salesforce platforms rather than the whole market.
6
IRP Commerce
Published: July 2026
Method: Aggregated operational trading data recorded within the IRP platform, using consistent attribution across the merchant network rather than survey responses.
What it does not cover: UK and Ireland, and one platform’s merchant base. Conversion rate varies enormously by sector, price point and traffic mix, so this is context rather than a benchmark.
7
National Retail Federation and Happy Returns
Published: Published 15 October 2025
Method: Two surveys: 2,006 US consumers who had returned an online purchase in the previous twelve months, and 358 eCommerce professionals at US merchants above $500 million in revenue.
What it does not cover: United States. Return rates vary so widely by category that the headline figure tells you little about apparel or furniture specifically.
8
Shopify Inc.
Published: Released 5 August 2026
Method: Audited company reporting filed with the US Securities and Exchange Commission.
What it does not cover: One platform. GMV is merchant sales processed through Shopify, not Shopify’s own revenue, and the two are frequently confused in secondary reporting.
9
Amazon.com, Inc.
Published: Released 30 July 2026
Method: Company quarterly reporting.
What it does not cover: Net sales include AWS and other non-retail businesses, so this is not a marketplace-only figure. Amazon did not disclose a third-party unit share in this release.
Link to the original source in your own work. We would rather be useful than credited.
- Our process
How this page is maintained
Statistics pages rot faster than any other kind of content. This is the process that keeps this one honest.
Where figures come from
- Government statistics offices first, where the measure exists
- Company filings and earnings releases for platform figures
- Research organisations that publish their method
- Never an article citing an article citing a figure
Review schedule
- Full review twice a year, in March and September
- Quarterly figures checked when the next quarter publishes
- Any figure older than two years is flagged on the page itself
- Superseded numbers are replaced, not stacked underneath
What we will not publish
- Figures with no traceable original source
- Projections presented as measurements
- Averages across categories that have nothing in common
- Our own client data dressed up as an industry benchmark
Corrections
- Tell us if a figure here is wrong or out of date
- We correct on the page rather than quietly deleting
- Source links are checked at each review
- If a source withdraws a figure, we remove it
Found a number here that has moved on? Tell us and we will check it against the original.
- The rest of the set
Statistics by subject
Deeper cuts, each built to the same citation standard. Pages in production are named but not linked.
Published
The abandonment rate, every stated reason with its share, and what each one actually costs to fix.
- Rate across 50 studies
- Reasons ranked, with sources
- What to do about each one
In production
Shopify Statistics
Merchant GMV, revenue and growth from company filings, with the distinction between GMV and revenue made explicit.
In production — not yet published
In production
Amazon Seller Statistics
Marketplace scale, advertising growth and seller economics, from Amazon reporting rather than from aggregators.
In production — not yet published
In production
Mobile Commerce Statistics
Device share of traffic and orders, and why the gap between the two is the interesting part.
In production — not yet published
In production
Conversion Rate Statistics
What published conversion benchmarks can and cannot tell you about your own store.
In production — not yet published
In production
Fashion eCommerce Statistics
Category-specific figures, where returns rates make the headline numbers misleading.
In production — not yet published
Need a sourced figure we have not published? Ask us and we will find it or tell you it does not exist.
- Questions
About these statistics
What people ask before quoting a figure from this page.
Can I cite these statistics in my own work?
Yes, and please cite the original source rather than us. Every figure links to the organisation that produced it, and that link is what makes the claim checkable for your own readers. If it helps to mention where you found it, a link back is welcome but not required.
Why are there only a couple of dozen figures and not a hundred?
Because each one has to survive the sourcing standard. Pages promising a hundred statistics are almost always recycling secondary summaries, and a large share of those numbers trace back to nothing at all when you follow them. We would rather publish fewer figures you can stand behind.
Why do the US and UK online share figures differ so much?
Definition, mostly. The two statistics offices define retail differently, cover different geographies and apply different adjustments. The comparison table above sets the two side by side. Quoting either as the global eCommerce share is a mistake.
How often is this page updated?
Fully reviewed twice a year, in March and September, with quarterly figures checked as each new quarter is published. The last review and the next are stated at the top of the page so you can judge for yourself whether it is current.
One of the figures looks out of date. What happens?
The mobile share figure is from 2024 and is labelled as the oldest on the page. Where a better source exists we replace it at review; where it does not, we keep the figure and keep the date visible rather than implying it is current.
Do you publish your own client data as benchmarks?
No. Our engagement data appears in case studies, clearly labelled as our own results, and it is never presented as an industry benchmark. Those are different claims and mixing them is how most agency statistics pages lose credibility.
What about global eCommerce market size?
We have deliberately not published one. The widely quoted global totals come from commercial forecasting houses whose methodology sits behind a paywall, which fails the standard on this page. If you need one for a board paper, buy the report rather than citing a blog.
Can you pull these figures for our category?
Sometimes. Category-level data is thinner and more often proprietary. A free eCommerce audit compares your own numbers against what we see across our client base, which for most decisions is more useful than a published average anyway.
Benchmarks are context. Your own numbers are the decision.
A free eCommerce audit puts your actual conversion rate, abandonment, returns and channel mix against what we see across the accounts we run, and tells you which gap is worth closing first.