Research library
Measurement and experiments research
Source data and original analysis for ecommerce teams. Start with the featured study, inspect the definitions and calculations, and follow each finding to its source.
13research pages
263source citations
26research exhibits
November 2026library updated
Featured study
Holiday Ecommerce Sales Statistics and Seasonal Adjustment
Compare eleven U.S. holiday quarters, raw and seasonally adjusted growth, ecommerce share and annual sales using the November 2026 Census estimates.
159 citations · 6 charts · Updated November 2026 Read the study- U.S. unadjusted ecommerce sales reached $335.7 billion in Q3 2026.
- The seasonally adjusted Q3 2026 number was $347.3 billion.
- Unadjusted U.S. ecommerce rose 20.1% from Q3 to Q4 2025.
- Adjusted U.S. ecommerce rose 1.7% across the same quarters.
Research library
Explore the studies and their source data.
- Branded paid search incrementality statistics from the eBay experimentDoes paid branded search add visits, or change which search link gets credit? Published research, original calculations and limits for DTC merchants.8 citations · 1 chartsAbout 99.5% of branded search traffic was retained.
- A/B test duration statistics for one-week and two-week resultsDoes a second week merely narrow uncertainty, or can the measured effect itself change? Published research, original calculations and limits for DTC merchants.7 citations · 1 chartsA second week widened standard errors in 17% of cases.
- Ecommerce order total reconciliation statisticsAn exact-decimal Olist analysis of payment and item totals, tolerance bands, delivery status and the records needed before another payment request.16 citations · 1 chartsExact equality held for 98,089 of 98,665 matched Olist orders.
- Ecommerce A/B Testing Statistics on Effect Size and Where Experiments HappenA 2,732-test ecommerce study shows a skewed distribution of conversion movements and why the direction of each effect matters.24 citations · 1 chartsMean movement was seven times the median.
- Email holdout statistics by prior purchase channelA 64,000-customer randomized retail replay compares email purchase lift among past phone, web and multichannel buyers.4 citations · 2 charts
- Revenue lift statistics for ecommerce experiments with high-spend ordersA randomized retail dataset shows how uncapped email revenue lift compares with two alternative customer-level spend caps.4 citations · 2 charts
- Checkout holdout statistics for repeat order counts and customer varianceA retail ledger replay shows how treating purchase events as evenly dispersed can make a repeat-order test look more precise than it is.4 citations · 2 charts
- Incremental advertising return statistics and what a holdout can tell youIn 25 randomized ad experiments, the median ROI interval exceeded 100 percentage points. Learn how to read an uncertain checkout recovery result.11 citations · 2 charts
- Sample ratio mismatch statistics for checkout A/B testsCompare published experiment integrity findings and reproduce a 50/50 sample-ratio check at several sizes before trusting a checkout result.9 citations · 2 charts
- A/B test peeking statistics for checkout experimentsReplay 100,000 no-effect simulations and see how repeated checks change the chance of a false positive in a checkout experiment.7 citations · 2 charts
- Email holdout statistics for visits, purchases and revenueA reanalysis of 64,000 randomized retail customers shows how the same email test looks at three stages of the funnel.6 citations · 2 charts
- Experiment variance statistics for pre-test and post-test retail spendA reproducible split of 64,000 randomized retail records tests whether past spend makes an email revenue experiment more precise.4 citations · 2 charts
Method
Follow the evidence and its limits
- 01
Identify the population
Every study states who or what was measured, and when.
- 02
Inspect the calculation
Downloads preserve source values, denominators and derived results.
- 03
Keep the scope
Historical data and controlled comparisons retain the conditions that determine where the findings apply.
Study methods distinguish original calculations, reported experimental results and practical interpretation.