# What actually sells out at the Edinburgh Fringe

Source: https://edfringeguide.com/analysis

24,481 ticketed performances across 2,690 shows, 3–16 August 2026. Snapshot pinned at 16 August 2026 (2026-08-16T19:00:00.000Z).

**Every rate below is a floor**: a date counts as sold out only where the outcome was observable, and unobservable dates count as not sold out. Floor 18.14% (4,440/24,481); upper bound 28.67% on the 15,487 dates whose outcome was visible. Traverse Theatre (110 dates), The Stand (289 dates), Laughing Horse / Free Fringe (2,050 dates) are excluded from every ranked table: they sell only through the Fringe's central box office, which reports sellouts but never confirms a miss, so their rates have no upper bound.

## Headlines

- **1 in 5 — dates sell out.** 4,440 of 24,481 ticketed dates in the first two weeks — at least 18.14%, and at most 28.67% once you drop the dates whose outcome nobody could observe.
- **58.8% — of shows never sell out once.** Of 2,690 shows, 1,582 have not sold a single date out. The top 100 shows — 3.7% of the catalogue — produce 39.9% of all sellouts.
- **7× — gap between the best and worst venue.** Monkey Barrel sells out 44.3% of its dates; Just The Tonic 6.1%. The venue spread is wider than the price spread — and once everything is fitted at once, it is price that turns out to be mostly a restatement of which room and which operator.
- **Opposite — directions for comedy and theatre.** Big rooms sell out more — for comedy. For theatre the sign flips, in all four of the operators big enough to test it. Theatre sells out by being small.
- **84% — for an Edinburgh Comedy Award mention.** 51 shows name the Edinburgh Comedy Award or Perrier in their blurb; 84.3% of them sell out at least one date, against a 43.2% base — and it holds after controlling for venue, price and where the act is from.
- **+12.6pp — for a Saturday.** Saturday 15 August hit 32.5% against a clean weekday baseline of 19.85%. It is one of the few effects that survives adjustment for everything else (+9.5pp).

## Where you book matters more than what you pay

_The gap between the best and worst venue operator is seven-fold, and it is not the same fact as the price gap wearing a different hat._

**Share of ticketed dates that sold out, by venue operator**

| | Sold out | Dates |
|---|---:|---:|
| Traverse Theatre (10 shows) | 96.4% | 110 |
| Monkey Barrel (104 shows) | 44.3% | 880 |
| Pleasance (222 shows) | 30.2% | 2,414 |
| Summerhall (72 shows) | 29.6% | 891 |
| The Stand (63 shows) | 26.3% | 289 |
| Underbelly (164 shows) | 24.0% | 1,802 |
| Assembly (191 shows) | 16.3% | 2,344 |
| theSpace (270 shows) | 14.7% | 1,913 |
| Gilded Balloon (171 shows) | 12.3% | 1,766 |
| ZOO (43 shows) | 10.4% | 395 |
| Laughing Horse / Free Fringe (189 shows) | 9.9% | 2,050 |
| Hoots (191 shows) | 7.2% | 1,733 |
| Greenside (123 shows) | 6.5% | 1,010 |
| C venues (76 shows) | 6.2% | 616 |
| Just The Tonic (161 shows) | 6.1% | 1,425 |

Floor rate: a date counts as sold out only where we could see it, so every bar is a lower bound. For most operators the ceiling is only a few points higher, because they sell through their own ticketing as well as the Fringe’s and both outcomes are visible. The three marked ≥ have no ceiling at all — see below.

The Big Four are not a bloc. Pleasance sells out at nearly twice Assembly’s rate and two and a half times Gilded Balloon’s, on comparable ticket prices — a bigger spread than anything the programme itself would suggest. Gilded Balloon is the standout underperformer relative to what it charges.

Monkey Barrel is the outlier at the top, and it is a year-round Edinburgh comedy club running its own rooms through August rather than a festival land-grab. At the bottom, Just The Tonic, Greenside, C venues and Hoots all sit under 8% — roughly a third of the festival-wide rate.

One qualification on the top of the table. Fit everything at once — operator, room size, price, genre, date — and Monkey Barrel’s advantage shrinks from +13.4 points to +5.3 and stops being statistically significant, because part of what looks like a venue effect is the kind of act that venue books. The spread across the table as a whole is real; the exact size of any one operator’s edge is softer than the chart makes it look.

> **What the ≥ on three of those bars means**
>
> Traverse, The Stand and Laughing Horse sell every date through the Fringe’s central box office, and it retires a past date rather than freezing it: the availability gauge snaps to zero whether or not the show sold. Checked against performances still 70%+ available three hours before curtain, that end-of-run reading was wrong 20 times out of 34, against 0–1.4% for venues running their own ticketing. Their sellouts are real — those come from live readings taken before curtain — but their misses can never be confirmed, so each rate is a floor with no ceiling. The Stand’s ≥ 26.3% is 76 confirmed sellouts across 289 dates and could truthfully be anything up to 100%, which is why it should not be read as "below Underbelly". Traverse is the one case the floor settles by itself: even counting every unresolved date as a miss, it is the highest rate at the festival.


## Room size runs backwards for theatre

_The festival-wide finding is that bigger rooms sell out more. Split it by genre and the sign flips — inside every operator big enough to test._

**Sellout rate by room capacity — all genres pooled**

| | Sold out | Dates |
|---|---:|---:|
| Under 50 seats | 15.6% | 6,381 |
| 50–99 | 14.2% | 9,253 |
| 100–199 | 21.6% | 5,475 |
| 200–399 | 34.0% | 1,511 |
| 400+ | 27.9% | 1,030 |

Capacity is the room’s, from the venue’s own seating data, and is known for 96.6% of dates.

That is the opposite of the intuition that a small room is easy to fill. A 40-seat room is not scarce, it is invisible: the scarcity only bites once enough people know the show exists. But pooled like this the finding is mostly comedy, because comedy is the larger half of the sample.

**The same bands, split by genre**

| Seats | Comedy | n | Theatre | n |
|---|---:|---:|---:|---:|
| Under 50 | 8.5% | 1,760 | 15.1% | 2,324 |
| 50–99 | 17.6% | 3,281 | 12.3% | 2,991 |
| 100–199 | 30.3% | 1,638 | 13.4% | 1,146 |
| 200–399 | 36.0% | 594 | 15.7% | 198 |
| 400+ | 41.3% | 208 | 5.9% | 51 |

Excludes dates sold only through the Fringe’s central box office, which reports sellouts but not misses and is 25.6% of comedy against 12.2% of theatre — pooling the two without that exclusion understates comedy. The 400+ theatre cell is 51 dates; treat it as a direction, not a number.

Comedy rises almost five-fold across the range. Theatre is flat and then falls, and in the smallest rooms theatre beats comedy outright. The obvious objection is that big rooms are big-name venues — so hold the operator fixed and split at 100 seats:

**Same operator, same festival, opposite sign**

| Operator | Comedy under 100 | Comedy 100+ | Theatre under 100 | Theatre 100+ |
|---|---:|---:|---:|---:|
| Pleasance | 36.0% | 41.4% | 25.6% | 15.2% |
| Assembly | 18.1% | 33.7% | 10.4% | 1.0% |
| Underbelly | 17.5% | 23.0% | 36.0% | 19.2% |
| Gilded Balloon | 11.0% | 19.2% | 5.0% | 3.0% |

Minimum 40 dates per cell. Comedy rises in four of four operators; theatre falls in four of four.

The plainest reading: a comedy audience is largely buying “a comedy at 8pm”, so a bigger room draws a bigger crowd. A theatre audience is buying that play, so demand is bounded per show and a bigger room mostly buys emptier seats. Theatre sells out by being small.


## What a credential in the blurb is worth

_Specific award names carry real signal. Generic “award-winning” carries almost none. Both survive controlling for venue, price and region — or fail to._

**Shows whose blurb makes each claim, and how often they sell out at least one date**

| Claim in the blurb | Shows | Ever sold out | Still there after controls? |
|---|---:|---:|---|
| Edinburgh Comedy Award / Perrier | 51 | 84.3% | Yes — strongly |
| Melbourne Comedy Festival | 35 | 80.0% | Yes — strongly |
| So You Think You’re Funny / new-act awards | 32 | 68.8% | Yes |
| Any overseas festival credit | 111 | 63.1% | Yes |
| “Sold-out run” claim | 44 | 61.4% | Yes |
| Adelaide Fringe | 67 | 58.2% | Partly |
| Returning / “by popular demand” | 200 | 56.0% | Weakly |
| TV credit | 169 | 55.6% | Weakly |
| Any award claim at all | 527 | 55.2% | Weakly |
| Fringe First | 26 | 42.3% | No — not significant |

Base rate: of 2,239 shows with at least three past dates, 43.2% sold out at least one. “After controls” is a Mantel–Haenszel odds ratio stratified separately by venue operator, by region and by price band — the three obvious ways a credentialled act is different anyway.

The pattern is that naming a specific award beats claiming one. “Edinburgh Comedy Award nominee” nearly doubles the base rate and keeps most of that after you account for the better venue and higher price such an act commands. “Award-winning”, pooled across 527 shows, barely moves.

> **Read this as a marker, not a lever**
>
> These are counts of what the blurb says, not verified nominations, and the numbers behind the strongest rows are small — 51 shows and 35 shows. An act that can honestly write “Edinburgh Comedy Award nominee” is a different act from one that cannot, in ways no control here captures. Adding the phrase to a blurb does not add the audience.


## Price looks like the strongest driver, and mostly is not

_The raw relationship is clean, monotone and about four-fold. Almost all of it disappears once you account for which room and which operator the show is in._

**Sellout rate by ticket price**

| | Sold out | Dates |
|---|---:|---:|
| £1–6 | 8.5% | 3,122 |
| £6–10 | 10.0% | 5,067 |
| £10–14 | 20.0% | 8,374 |
| £14–18 | 21.8% | 5,956 |
| £18+ | 37.3% | 1,445 |

Cheapest listed price per date; bands include their upper figure, so a £10 ticket is in “£6–10”. That is worth stating because £10 alone is 2,445 dates — put it in the band above instead and the £6–10 rate reads 8.4% rather than 10.0%. Free-but-ticketed dates are excluded: 53% of them are one show, and they behave nothing like the rest of the programme.

> **The adjustment that kills it**
>
> Fit everything at once — price, operator, room size, genre, date, time, region — and the £20+ effect shrinks from +23.9 points to +6.0, with a confidence interval crossing zero. Shows that expect demand both charge more and get booked into bigger rooms by better-selling operators; the model credits the room and the operator. Room size, Saturday, and where the act is from all survive that adjustment. Price largely does not.

For an audience this still works as a signal, just not as a cause: an £18 ticket is a reliable tell that the operator expects to fill the room. For a producer it is not a lever — putting the price up does not buy the demand that usually accompanies it.


## Genre

_Music and children’s shows lead. Circus trails badly. Comedy beats theatre, and the gap is wider at show level than at date level._

**Sellout rate by genre**

| | Sold out | Dates |
|---|---:|---:|
| Children’s shows (118 shows) | 21.6% | 1,083 |
| Music (217 shows) | 21.3% | 1,106 |
| Cabaret (115 shows) | 17.9% | 918 |
| Comedy (1,060 shows) | 17.4% | 10,129 |
| Opera (127 shows) | 17.2% | 991 |
| Spoken word (69 shows) | 16.2% | 216 |
| Theatre (775 shows) | 15.4% | 7,757 |
| Circus (84 shows) | 10.0% | 722 |

Exhibitions and events are excluded: walk-in and timed-entry formats barely use the booking channel, so their rate measures the format, not demand.

At show level rather than date level the comedy/theatre gap is much starker: 49.6% of comedy shows sell out at least one date, against 31.7% of theatre shows. Theatre is also the only major genre that is more expensive than comedy while selling out less — £13.16 against £10.39.

Format tags cut across genre and carry signal genre alone does not. Dates tagged “Interactive” sell out at 30.7% across 67 shows; “Classical” at 2.1% across 45.


## When shows sell out

_Prime time is a myth, Saturday is real, and the festival gets harder to get into as it runs._

**Sellout rate by start time**

| | Sold out | Dates |
|---|---:|---:|
| 11am | 22.4% | 1,494 |
| 3pm | 21.1% | 2,170 |
| 7pm | 19.8% | 2,134 |
| 8pm | 19.4% | 1,732 |
| 6pm | 18.5% | 1,989 |
| Noon | 15.3% | 2,059 |
| 9pm | 16.9% | 1,725 |
| 10pm | 13.4% | 1,115 |
| 11pm | 14.9% | 606 |

UK start time. Hours with fewer than 150 dates are omitted.

The 8pm–10pm slot everyone competes for sits at or below the festival average. The peaks are 11am and 3pm, and the weak points are the lunchtime trough and anything after 10pm. The whole range is modest — smaller than the venue effect by some distance.

**The weekend effect, measured on a clean week**

| Window | Dates | Sold out |
|---|---:|---:|
| Monday–Friday, 10–14 August | 11,459 | 19.85% |
| Saturday 15 August | 2,463 | 32.48% |

Sunday 16 August is not shown: the snapshot cuts off at 7pm that day, so its evening slots had not resolved. Earlier weekend dates sit inside a monitoring outage (6–10 August) and are not usable either.

Across the run the rate climbs steadily — 10.5% on 5 August to 21.4% on 14 August — which is the same show catalogue getting harder to book into as word of mouth accumulates and the tourist population peaks. Booking late is a worse strategy in week three than in week one.

Selling out is not usually a cliff. Of every ticketed date, 35.2% dropped below half availability at some point, 21.2% got under 20% left, and 18.1% went to zero — so most of what runs low does eventually go. Eventual sellouts are separable from the rest about three weeks out, when they are already at a median 63% availability against 87% for everything else.


## Where the act is from

_Australian shows sell out at nearly twice the rate of American ones, and it is one of the effects that survives adjustment._

**Sellout rate by the company’s home country**

| | Sold out | Dates |
|---|---:|---:|
| Australia | 28.4% | 1,431 |
| Scotland | 22.7% | 3,882 |
| England | 16.1% | 10,479 |
| USA | 13.0% | 3,797 |
| Elsewhere | 12.1% | 3,333 |

Australia holds up inside every venue operator large enough to test, and survives the multivariate model at +10.1 points; Scotland at +7.2. The Australian circuit sends acts to Edinburgh with a Melbourne or Adelaide run already behind them, which is the same signal the credential table picks up from a different direction.


## Sellouts are concentrated in very few shows

_Selling out is a property of a small set of shows, not something that visits the programme evenly._

- **58.8% of shows** — 1,582 of 2,690 have not sold out a single date.
- **3.7% of shows** — produce 39.9% of all sellouts — the top 100 titles.
- **5.6% of shows** — sold out every date that has resolved. A full-run sellout is rarer than a hit.
- **40.9% of dates** — finished under a quarter full; 13.1% finished completely full. The middle is thin — the distribution is barbelled, not bell-shaped.

Across the first two weeks that works out at roughly 1.24 million seats sold against 1.31 million empty ones, and about £18.5m of tickets at the cheapest listed price — a floor on gross, not an estimate of it. Just under half of the seats the festival put on sale were used.


## Appendix

### How this was measured

_One pinned snapshot, one definition of “sold out”, and one population used for every number above._

The data is Ed Fringe Guide’s own availability monitor, which polls the Fringe’s ticketing feed through the day and records every change. This page reads a snapshot pinned at 16 August 2026, covering 3–16 August 2026 — 24,481 ticketed performances across 2,690 shows. It is pinned because the monitor writes continuously: an unpinned query moves the population between two questions asked ten minutes apart, and the totals stop reconciling.

A date counts as sold out if the ticketing ledger recorded it crossing to zero, or a trusted final reading showed zero availability having previously shown some. That is narrower than “no seats exist”: it means the central box office allocation is gone. Venues frequently hold back their own stock, so a sold-out date here can still be gettable at the door.

- **Floor rate — used everywhere on this page** — 4,440 of 24,481 dates, 18.14%. Dates whose outcome could not be observed count as not sold out, so every figure is a lower bound.
- **Observed rate — quoted once, as the other bound** — 4,440 of 15,487 dates, 28.67%. Restricted to dates whose outcome was visible at all. It contains every sellout by construction, so it is an upper bound.
- **Why the two are not averaged** — The truth is between them, but not at a fixed point: coverage varies by venue, so the gap is wider for some operators than others. Mixing the two inside one table would rank reporting behaviour rather than selling.


### What these numbers cannot tell you

_The honest list, including the two mistakes this analysis made and had to correct._

- **None of it is causal** — Every table above is an association. Where a multivariate fit (held-out AUC 0.742) disagrees with the raw table — price, most obviously — the section says so inline. A producer cannot pull these as levers.
- **Three venues have a floor and no ceiling** — Traverse, The Stand and Laughing Horse sell only through the Fringe’s central box office, which retires a past date by zeroing its availability gauge whether or not it sold — wrong 20 times out of 34 when checked against shows still 70%+ available three hours before curtain. Their sellouts are real and are counted; their misses cannot be confirmed. They are on the page as bounds (“≥”), and are kept out of the sorted chart because a floor and a bracketed measurement do not belong in one ranking.
- **Reviews and press are missing** — There is no star-rating or press-coverage data in this dataset, and a four-star review in week one is very plausibly the largest single driver of a week-two sellout. Its absence is the biggest known gap.
- **Capacity is the room’s, not the show’s allocation** — Several shows share a room across a day, and the seating figure does not vary by performance. Fill percentages are therefore estimates of the central allocation, not house occupancy.
- **6–10 August is a hole** — An upstream outage stopped the monitor for four days. Those dates are excluded from every weekday and trajectory claim rather than averaged in.
- **Two earlier passes got this wrong** — The status the feed sets when a central allocation is exhausted was twice stripped out as a supposed artefact — it carries 61% of all sellouts, and removing it inverted the country findings and understated fill by six-fold. It is included here, and the correction is why some figures on this page differ from an earlier draft of the same analysis.
- **This is two weeks of a three-and-a-half week festival** — The cutoff is 16 August 2026 and the festival runs to 31 August. Rates rise through the run, so the final numbers will be higher than these.