Crowd Control Digital

CHART
CONTROL

THE HOT 100,
 REVERSE-ENGINEERED.

LIVE
WEEK OF APR 3 – 9, 2026PUBLISHED APR 15
THIS WEEK · #1

Choosin' Texas

Ella Langley · Columbia

PEAK
#1 · W25
299,600
SONG EQUIVALENT POINTS
POINT COMPOSITION
100%
PREMIUM
198.1K66.1%
AIRPLAY
69.8K23.3%
AD-SUPP
21.6K7.2%
SALES
8.9K3%
01 · THRESHOLDS

POINTS TO HOLD POSITION

Minimum Song Equivalent points at each chart rank, this week.

#1
299.6
SEQ Pts
0.0% WoW
#10
146.7
SEQ Pts
1.2% WoW
#25
88.9
SEQ Pts
0.5% WoW
#50
55.6
SEQ Pts
2.1% WoW
#75
36.3
SEQ Pts
1.8% WoW
#100
31.1
SEQ Pts
0.3% WoW
02 · CALCULATOR

Will Your Song Chart?

Input a consumption profile. Predicted Hot 100 position updates in real time.

18.0M
Top-10 avg: 18–28M premium streams/week
5.0M
Spotify Free & ad-supported platforms (YouTube excluded)
2.0M
Pandora, iHeart programmed, SiriusXM streaming
50.0M
Impressions from Hot 100 Panel radio stations
25,000
Paid downloads & physical singles (CD / vinyl)
TOTAL POINTS
270,346
PREDICTED POSITION
#3
±3 position margin
#1
299,600 pts
-29K
#10
146,717 pts
+124K
#25
88,931 pts
+181K
#50
55,619 pts
+215K
#75
36,298 pts
+234K
#100
31,079 pts
+239K
Already past #50.
03 · METHODOLOGY

The Formula

Luminate's Song Equivalent, reverse-engineered from 13 weeks of chart data and 42 consumption drill-ins. Re-fit every week as new data lands.

MODEL v13Last calibrated Apr 18, 2026In-sample MAPE 5.35%Confidence 87%
PREMIUM ON-DEMAND
÷ 106
Calibrated
Subscription on-demand streams — Spotify Premium, Apple Music, Amazon, Tidal-15.2% vs 2024 prior (÷125)
AD-SUPPORTED ON-DEMAND
÷ 364.8
Calibrated
Ad-supported on-demand streams (Spotify Free, etc.). YouTube excluded since Jan 16, 2026-2.7% vs 2024 prior (÷375)
PROGRAMMED STREAMS
÷ 700
Needs Data
Non-interactive radio-style streams (Pandora, iHeart programmed, SiriusXM streaming)+40.0% vs 2024 prior (÷500)
AIRPLAY AUDIENCE
÷ 846.7
Calibrated
Radio airplay audience impressions from Hot 100 Panel stations (Mediabase via Luminate)+5.8% vs 2024 prior (÷800)
DIGITAL + PHYSICAL SALES
÷ 1
Locked
Paid downloads and physical singles (CD, vinyl). One unit = one point.

The model learns every week. New drill-ins trigger a refit; divisors drift toward their true values as sample size grows. Sales is locked at ÷1. Programmed will move off the 2024 prior once we capture per-song Service Type breakdowns starting next chart week.

04 · VALIDATION

THIS WEEK'S AUDIT

Predicted vs actual Song Equivalent across six threshold positions.

Choosin' Texas

Ella Langley

#1
ACTUAL299,600
PREDICTED298,359
ERROR0.41%

Folded

Kehlani

#10
ACTUAL146,717
PREDICTED145,703
ERROR0.69%

Body

Don Toliver

#25
ACTUAL88,931
PREDICTED78,840
ERROR11.35%

Porch Light

Noah Kahan

#50
ACTUAL55,619
PREDICTED46,213
ERROR16.91%

DAÑO

Peso Pluma, Tito Double P

#75
ACTUAL36,298
PREDICTED35,573
ERROR2.00%

Sweet Boy

Malcolm Todd

#100
ACTUAL31,079
PREDICTED25,328
ERROR18.50%
05 · BACKTEST

Does It Generalize?

In-sample MAPE is always optimistic. The honest test: fit the model on onlyprior weeks, then predict the next one without peeking. Here's how the model performs week-by-week in that strict walk-forward setup.

IN-SAMPLE MAPE
5.35%
Fit on all 42 drill-ins, tested on the same. Optimistic by nature.
WALK-FORWARD MAPE
5.85%
Fit on prior weeks only, predict the next. 36 out-of-sample predictions. The real test.
GENERALIZATION GAP
+0.50pp
Held-out matches in-sample. Model is well-regularized, not overfit.
ACCURACY OVER TIME

Weekly Held-Out MAPE

lower = more accurate
Walk-forwardis the strict test — each point uses divisors fit only from weeks before it, simulating how the model would have performed in production. First two weeks are omitted (need ≥2 prior weeks to fit). Leave-one-out uses every other week to predict the held-out one — less strict but shows per-week fit quality.
OUT-OF-SAMPLE PREDICTIONS

Predicted vs Actual Points

dashed line = perfect prediction
Each point is an out-of-sample prediction (model never saw this week during training). Points on the diagonal = perfect prediction. Tight clustering along the line means the model generalizes across chart positions and weeks.
06 · HISTORICAL

Threshold Drift

How the points required at each rank shift week to week. Spikes mark big release weeks; valleys mark softer chart weeks worth targeting.

PROGRESS
13 / 13
WEEKS COLLECTED
42 drill-ins logged · 87% model confidence
Next collection: Tuesday after Billboard publish · dataset re-calibrates divisors automatically
07 · TIKTOK FINGERPRINT

Creates Per Point

TikTok creates don't correlate with Hot 100 position directly. Pearson −0.002 across our 11-song sample.

But the ratio of lifetime creates to chart points is a genre fingerprint. Low ratio = US radio & streaming engine. High ratio = global TikTok hit the Hot 100 undercounts.

US-NATIVE
4
Ratio < 2
Country / Rock / Rap Heartland
MIXED
2
Ratio 2–10
Indie Pop / Crossover
GLOBAL-VIRAL
5
Ratio > 10
Huge on TikTok, Undercounted
Genre Benchmarks

Median creates-per-point by normalized genre. Each song's ratio is most meaningfully compared to its cohort here.

COUNTRY
0.61
median
N=2 · 0.10–1.11
ROCK
2.49
median
N=2 · 0.37–4.62
POP
12.7
median
N=4 · 2.80–39.37
LATIN
22.4
median
N=1 · 22.45 only
HIP-HOP
33.4
median
N=2 · 1.52–65.28
#001
Choosin' Texas
Ella Langley
Country·31.4K/299.6K
0.10
US-NATIVE
#051
Freakin' Out
Dexter and the Moonrocks
Alt. Rock·19.7K/53.7K
0.37
US-NATIVE
#008
Be Her
Ella Langley
Country·176.2K/158.4K
1.11
US-NATIVE
#025
Body
Don Toliver
Rap·135.1K/88.9K
1.52
US-NATIVE
#009
STATESIDE
PinkPantheress
Pop·412.2K/147.2K
2.80
MIXED
#016
BABYDOLL
Dominic Fike
Alt. Rock·587.8K/127.4K
4.62
MIXED
#002
Man I Need
Olivia Dean
Pop·2.5M/224.2K
11.21
GLOBAL-VIRAL
#005
SWIM
BTS
Pop·2.5M/173.9K
14.23
GLOBAL-VIRAL
#023
DtMF
Bad Bunny
Latin - Rhythm·2.1M/95.0K
22.45
GLOBAL-VIRAL
#006
So Easy (To Fall In Love)
Olivia Dean
Pop·6.8M/172.0K
39.37
GLOBAL-VIRAL
#049
Raindance
Tems, Dave
R&B/Hip-Hop·3.8M/58.9K
65.28
GLOBAL-VIRAL
OPERATIONAL READ

Normalize TikTok expectations against genre, not an absolute number.

A country song at 500K creates is a breakout signal. A pop song at 500K creates is underperforming its cohort. The Hot 100 undercounts global TikTok virality: Raindance has 65× more creates per chart point than Choosin' Texas because its audience streams outside the US.

Source: Cobrand lifetime Total Creates · N = 11songs · weekly snapshots begin Wednesday

08 · THE CHART

Billboard Hot 100 · Week of Apr 3 – 9, 2026

All 100 charting songs, ranked by Song Equivalent units. Top 10 featured; remainder expandable.

TOP 10
01
Choosin' Texas
Ella Langley
299,600
PTS
02
Man I Need
Olivia Dean
224,225
PTS
03
I Just Might
Bruno Mars
209,224
PTS
04
Ordinary
Alex Warren
186,820
PTS
05
SWIM
BTS
173,905
PTS
06
So Easy (To Fall In Love)
Olivia Dean
171,963
PTS
07
Golden
AUDREY NUNA, EJAE, Kpop Demon Hunters Cast, REI AMI, HUNTR/X
160,274
PTS
08
Be Her
Ella Langley
158,363
PTS
09
STATESIDE
PinkPantheress, Zara Larsson
147,197
PTS
10
Folded
Kehlani
146,717
PTS
#11 – #100
15 / 44
011
The Fate of Ophelia
143,133
012
WHERE IS MY HUSBAND!
139,469
013
SLEEPLESS IN A HOTEL ROOM
130,289
014
Risk It All
129,175
015
Opalite
128,286
016
BABYDOLL
127,356
017
Dracula
122,182
018
iloveitiloveitiloveit
117,196
019
E85
114,478
020
YUKON
107,937
021
BE BY YOU
106,908
022
HOMEWRECKER
99,435
023
DtMF
94,978
024
AMERICAN GIRLS
92,210
025
Body
88,931