"1.7 trillion yuan" and "low to mid single digits" describe the same fortnight of Chinese e-commerce, and both numbers are defensible. Syntun's estimate for the full Double 11 period is 1.7 trillion yuan of gross merchandise value, up 14.2 percent, with e-commerce alone at 1.62 trillion yuan, up 12.3 percent. Citi, working from platform disclosures and its own checks, expects industry GMV to grow only low to mid single digits. As of November 10, 2026, the distance between the two estimates is not a disagreement about demand. It is a disagreement about what gets counted.
Two estimates, two definitions
Syntun's figure covers the whole festival window and includes every platform it monitors, which is why the total is larger and the growth rate higher than anything a single company reports. Citi's number is a growth estimate for the e-commerce industry and deliberately leaves out the instant retail increment that sits outside the shelf commerce GMV convention. Neither is wrong. They answer different questions: how much was sold in total, and how fast the core online retail business grew.
What makes this year harder to compare is that platforms changed both their mechanics and their reporting. Tmall and JD.com have not published a headline total for several years, and JD.com removed pre-sale entirely this time, opening with spot sales on 12 October. That single change moves volume between reporting periods as well as between channels, so a year on year comparison needs the window stated before the number means anything.
What the platforms did disclose
Numbers that survive a change of definition
The disclosures that hold up are the narrow ones. Tmall said 124 brands passed 100 million yuan in the first four hours of sales and that 27,681 brands beat their first day sales from a year earlier. JD.com reported double digit growth in transacting users on its opening day and, in press summaries of its own release, buyer numbers up 40 percent and orders up nearly 60 percent, with no GMV figure attached. Douyin said 67,000 brands doubled sales, and Kuaishou put 3 billion yuan into red packets and subsidies while reporting search GMV up more than 33 percent.
Those are counts of things rather than values of things, which is exactly why they stay comparable across years. A brand that plans on totals is planning on a number nobody publishes. A brand that plans on category growth and its own share of transacting users is working with data it can check, even if the check is uncomfortable.
| Metric | Published by | Comparable across years | Use |
|---|---|---|---|
| Total GMV | Syntun and other third parties | Yes, with a fixed method | Market size |
| Brands above a threshold | Tmall | Yes | Category depth |
| Transacting users | JD.com | Partly | Acquisition |
| Order count | JD.com | Partly | Basket behaviour |
| Search GMV | Kuaishou | No | Channel mix |
Consumption structure adds a second reason the totals diverge. Official data for the first three quarters showed goods retail up 1.1 percent while services retail grew 5 percent, and economists argue the next leg of consumption growth sits in services. A festival built on goods will therefore grow more slowly than a number that includes delivery, dining and travel attach rates, and that is the increment instant retail captures.
How to read the gap without a total
Three signals worth more than a headline number
Start with category growth where a platform publishes a rate rather than a value, because a rate survives a definition change. Then look at the funding model behind the discount, since a merchant funded item discount and a platform funded coupon produce the same consumer price and very different margins. Finally, track the instant retail increment separately: transaction value in that channel rose 112.3 percent year on year during the 618 period, and folding it into a shelf commerce total is what produces the 4 percent and 10 percent versions of the same campaign.
Demand is not the variable in dispute. The dispute is about where demand is fulfilled, and that determines which number a brand should manage against in January, when the campaign is over and the only thing left is the allocation decision for the next one.
FAQ
Which number should a brand put in a board deck?
The one whose method is stated. A rate from a named marketplace, with a defined window, survives scrutiny better than a total with neither, and it can be checked against your own product listing data.
Why do third party totals run higher than platform disclosures?
Because they cover more channels, including instant retail and some content commerce, and because platforms pick the metric that flatters them.
Does the difference change any decision?
It changes allocation. If growth is coming from store level fulfilment rather than parcel, the budget follows availability and trade area coverage, not media.
