StubHub moved $3.1bn of tickets in the three months to June and booked $573.1m of that as its own revenue. That is roughly 18.5 cents of every dollar passing through the marketplace, and it is the clearest publicly audited number any organiser will ever see on what percentage-based ticket pricing actually costs the people buying and selling.
The resale giant published its second quarter results on 12 August. Revenue was up 33% year on year, gross merchandise sales up 34%, and yet the line that mattered to the market was a rounding error.
Record quarter, unremarkable bottom line · credit: Maxim Hopman / Unsplash
The numbers, without the adjectives
| Metric | Q2 2026 | Q2 2025 |
|---|---|---|
| Gross merchandise sales | $3.1bn | $2.3bn |
| Revenue | $573.1m | $430.3m |
| Adjusted EBITDA | $105.7m | $54.3m |
| Net income | $14.6m | Loss of $53.8m |
| Free cash flow | $309.7m | $9.7m |
Adjusted EBITDA nearly doubled. Debt came down by $200m across the year to date, and net leverage improved from 4.5 times to 3.0 times trailing earnings. The company credited a record World Cup, which is fair: a tournament across three countries is about as good as demand gets.
Why a record quarter produced almost no bottom line
Net income was $14.6m, a genuine swing from a $53.8m loss a year earlier. But net income attributable to common stockholders came in at a loss of roughly $40,000, and the shares fell after hours despite the headline growth. A business with $1.7bn of cash on hand and $1.2bn owed to sellers is, at the level of the ordinary shareholder, running to stand still.
That gap between spectacular gross figures and a flat net result is the most useful thing in the filing for anyone who does not own the stock. Volume is not margin. Marketplaces at this scale spend enormous sums acquiring the demand that generates the volume, and the take rate has to cover all of it.
What an 18.5% take rate means for your own pricing model
Resale is not registration, and StubHub's cut is not comparable to what a B2B ticketing platform charges. But the arithmetic of percentages is identical wherever you find them, and this is what a mature one looks like when it is applied to real ticket prices.
| Ticket face value | Retained at an 18.5% take |
|---|---|
| £20 | £3.70 |
| £50 | £9.25 |
| £150 | £27.75 |
| £400 | £74.00 |
The delivery cost of those four transactions is effectively the same. One database write, one email, one QR code, one scan at the door. The percentage is not pricing the work. It is pricing your ticket.
A percentage does not care whether your ticket costs twenty pounds or four hundred. That is the entire appeal of it for the vendor, and the entire problem with it for the organiser.
This is why the question changes shape depending on what you run. A £15 community event barely notices. A two-day conference at £395 a delegate is handing over the price of a speaker's travel for every twenty registrations. If you have never sat down and multiplied your own numbers, the flat fee versus percentage comparison is a twenty minute exercise that occasionally rewrites a budget.
Tired of Fee Announcements? Go Flat
eventcloud charges one subscription with no per-ticket fees, so platform news stops being budget news.
The part the earnings coverage left out: take rates move
Almost every write-up of these results treated the take rate as a fixed property of the business. It is not. It is a dial, and across the industry it has been turning in one direction.
Humanitix raised its US rate this year from 2.1% plus $0.99 per ticket to 5% plus $1.29. Cvent moved its per-registrant charges up and shifted the increase to a July cycle. Ontario spent months discovering that "face value" is not a stable concept once fees are layered on top of it, and a German court ruled against FIFA over how ticketing costs were being presented to buyers. The direction of travel across the market has been upward and the disclosure has been getting less clear, not more.
For an organiser this has a practical consequence that has nothing to do with StubHub. Whatever percentage you agreed when you signed is a snapshot, not a promise. The relevant question at renewal is not what the rate is today, but what the contract permits it to become, and how much notice you get. Anyone who has changed platform mid-cycle knows the answer to "what does switching cost" is mostly measured in data, not money, which is why the comparison work is easier to do before you need it than after.
What to actually take from a resale marketplace's quarterly report
Three things. Demand for live events in 2026 is strong enough to produce a 34% jump in a single quarter, which should embolden anyone sitting on a pricing decision. Scale does not automatically produce profit, so a vendor's size tells you nothing about how sustainable its pricing is. And percentage-based models are extremely comfortable for the party collecting them, which is precisely why they are so widely offered.
None of that makes a percentage wrong. It makes it a choice, and one worth making deliberately rather than inheriting from whichever platform you happened to sign up to first. At eventcloud we price flat because we would rather the cost of selling a ticket stay the same when the ticket gets more expensive, but the useful move for any organiser this week is simply to run the multiplication on your own event and see what falls out.
