The benchmark
We asked whether a room read predicts who plays a room. It doesn't — and the reason it doesn't is the finding, because it tells you precisely what the instrument measures. Both the null result and its mechanism are below.
What was tested
For each venue: does Roomtone score the acts that documentarily played there above plausible acts that didn't? The metric is AUC, computed against the incumbent method on the same candidate pool — ranking by Qloo's popularity percentile, which is how draw is actually decided.
Ground truth is Wikipedia tour date tables: 149 documented plays across 34 verified venues from 57 major tours, independent of Qloo. 149 of those plays were measurable.
Qloo's room affinity measures a venue's local audience, not the touring draw that fills it. That is not a guess — it is what the 149 documented plays establish, and it is visible in the acts Qloo returns. AFAS Live in Amsterdam hosts Harry Styles and Halsey; Qloo's read of that room is Dutch-language pop. Oslo Spektrum hosts Harry Styles and Dua Lipa; its read is Norwegian-language acts. Rogers Arena gets EDM DJs.
Touring decisions are made on draw, and popularity predicts draw almost perfectly (AUC 0.995). Taste fit is a different variable — the one that decides whether the act belongs to the crowd that is already in the room. The original build collapsed those two into one number. This test is why it no longer does.
Chance is 0.5. Fame decides who plays a room, and it does so almost without error.
Below chance, because the room read is answering the other question.
Not one touring act appears in the audience Qloo assigns to the room it played.
What the failure actually is
It would be easy to report a null result and stop. The mechanism matters more. Look at what Qloo returns for a room whose touring history we documented:
AFAS Live in Amsterdam is documented to host Harry Styles, Halsey and Lorde. Qloo's read of that room is Dutch-language pop — Nielson, Gers Pardoel, Bløf, Marco Borsato. Oslo Spektrum is documented to host Harry Styles and Dua Lipa; Qloo's read is Norwegian-language acts. Rogers Arena in Vancouver gets EDM DJs.
Qloo place affinity describes the local audience of a place, not the touring draw that fills it. International acts play those rooms because of how many people will buy a ticket — popularity, which the benchmark scores at 0.995 — not because of taste fit. The signal is real and it is coherent; it answers a different question than the one we first asked of it.
We also tested the obvious follow-up: maybe the read predicts the support bill rather than the headline, since support acts are smaller and more genre-specific. It does not. Across 26 documented support slots — Sabrina Carpenter with Taylor Swift, Griff with Dua Lipa, Zara Larsson with Beyoncé — not one appears in its venue's top 50 either. The touring ecosystem is draw-driven at every level.
But is the read real signal, or noise?
A null result on one question says nothing about the other. The product claims a room has a specific, coherent local audience — so we tested that claim directly across all 22 catalogued rooms. Two things have to hold. If they don't, there is nothing here.
Median audience overlap across 231 room pairs. If the read were generic, most rooms would return the same acts and this would approach 1. It is zero — the read is specific to each room.
The strongest overlap in the whole matrix — and it lands exactly where someone who knows these rooms would put it. Not uniform, not random: structured by scene.
The most-overlapping pairs are shared scenes
Bowery Ballroom and Saint Vitus are both New York indie and punk rooms and share 47% of their audiences. KOKO and Village Underground are both London grime rooms. Barclays and Radio City are both New York rap arenas. Every high overlap is a pair a booker would have named.
And it is hyper-local
Two rooms in the same city share 3.7% of their audience on average; two rooms in different cities share 0.2%. That is a 19× difference. The strongest cross-city overlap anywhere in the matrix is just 0.18 — and that pair is two Los Angeles rooms.
Genre alone does not carry across cities. That is the concrete product consequence: a booking decision researched in one market does not transfer to another, and a tool showing you a genre match is not showing you an audience. It also explains the null above — the same act does not draw the same audience in Amsterdam and Oslo.
What we changed as a result
- The claim is now narrower and true. Every page that once implied a booking recommendation now states that Roomtone measures a room's local audience, and does not forecast draw.
- Draw is a separate, explicit axis. A booking decision has two independent inputs — taste fit and draw — and collapsing them into one number is how a promoter gets misled. Roomtone shows both.
- Scope is now the local-programming decision — which acts to book into a room for its own audience, and which sponsors actually reach the people who are already there. That is where the taste read is the binding constraint rather than a comment on fame.
- The room reads still stand as evidence of the signal: The Stone returns a jazz canon, Sony Hall returns classical repertoire, KOKO returns UK grime. Those are local scenes, which is exactly what the mechanism says they should be.
Every venue, with its mechanism visible
| Venue | Qloo's read of the room | Doc. | AUC | Pop. |
|---|---|---|---|---|
Rogers Arena Vancouver | Jai Wolf, Louis the Child, San Holo | 14 | 0.00 | 1.00 |
Ziggo Dome Amsterdam | Gers Pardoel, Guus Meeuwis, Nielson | 13 | 0.00 | 1.00 |
Spark Arena Auckland | Cub Sport, Meg Mac, Ocean Alley | 11 | 0.00 | 1.00 |
Palau Sant Jordi Barcelona | Melendi, Sidecars, Dani Martín | 9 | 0.00 | 1.00 |
Royal Arena Copenhagen Municipality | Rasmus Seebach, Burhan G, Niklas | 7 | 0.00 | 1.00 |
Scotiabank Arena Toronto | Ella Mai, PARTYNEXTDOOR, Jhené Aiko | 7 | 0.00 | 0.99 |
AFAS Live Amsterdam | Nielson, Gers Pardoel, Bløf | 5 | 0.00 | 1.00 |
TD Garden Boston | Meno Tody, Bonde da Stronda, Black Days | 5 | 0.00 | 1.00 |
Accor Arena Paris | NikitA, Алексей Воробьёв, Светлана Лобода | 5 | 0.00 | 1.00 |
Lanxess Arena Cologne | Fettes Brot, Sido, Clueso | 5 | 0.00 | 0.97 |
MEO Arena Lisbon | Aurea, Светлана Лобода, Anjos | 5 | 0.00 | 1.00 |
Principality Stadium Cardiff | Tom Grennan, Circa Waves, Sam Fender | 4 | 0.00 | 1.00 |
PGE Narodowy Warsaw | Kamil Bednarek, Mrozu, Paktofonika | 4 | 0.00 | 1.00 |
Rogers Centre Toronto | Jon Bellion, Bryce Vine, Quinn XCII | 4 | 0.00 | 1.00 |
Stade de France Saint-Denis | Макс Барских, SEREBRO, Светлана Лобода | 4 | 0.00 | 1.00 |
National Stadium Bukit Jalil Kuala Lumpur | Sick Individuals, Thomas Gold, Nicky Romero | 4 | 0.00 | 1.00 |
Oslo Spektrum Oslo | Emilie Nicolas, Julie Bergan, Elsa & Emilie | 4 | 0.00 | 1.00 |
AO Arena Manchester | Sneakbo, Stormzy, Wretch 32 | 4 | 0.00 | 1.00 |
Brisbane Entertainment Centre. Brisbane | Meg Mac, Amy Shark, Thelma Plum | 4 | 0.00 | 1.00 |
AO Arena Manchester | Sneakbo, Stormzy, Wretch 32 | 4 | 0.00 | 1.00 |
Wembley Stadium London | SEREBRO, Макс Барских, Сергей Лазарев | 3 | 0.00 | 1.00 |
Johan Cruijff Arena Amsterdam | Nielson, Gers Pardoel, Bløf | 3 | 0.00 | 1.00 |
Etihad Stadium Manchester | The Courteeners, Reverend and the Makers, Ocean Colour Scene | 3 | 0.00 | 1.00 |
BC Place Vancouver | French Montana, Ty Dolla $ign, PARTYNEXTDOOR | 3 | 0.00 | 1.00 |
Movistar Arena Santiago | Los Prisioneros, Glup!, Los Bunkers | 3 | 0.00 | 1.00 |
Palacio de los Deportes Mexico City | DLD, Siddhartha, Ximena Sariñana | 3 | 0.00 | 1.00 |
SM Mall of Asia Arena Pasay | EXO-K, SUPER JUNIOR, NU'EST | 3 | 0.00 | 1.00 |
Climate Pledge Arena Seattle | Dustin Kensrue, Rocky Votolato, Good Old War | 3 | 0.00 | 0.93 |
Grant Park Chicago | Megan Thee Stallion, Saweetie, Doja Cat | 3 | 0.00 | 0.97 |