Earlier today I told you on camera that I missed Drake’s ICEMAN first-week number by 62,000 album-equivalent units. I predicted 525,000 on May 4. The actual landed at 463,000. I framed the miss as a release-strategy-level restraint I failed to price — Drake triple-dropped, capped his own per-album peak, and took the top three Billboard 200 spots in one week. The first artist to do that since the chart went weekly in March 1956.
A reader named Dirty South Joe pushed back on Instagram. His question was direct: did I factor in YouTube into the prediction? Because, as he put it, “it’s convenient that they just started not counting those now.”
Joe was half right.
Here’s what happened. On December 16, 2025, Billboard announced a methodology change for the Billboard 200. The ad-supported-to-paid-stream ratio tightened from 1:3 to 1:2.5. In practical terms: under the old rules, 1,250 paid streams equaled one album unit, and 3,750 ad-supported streams equaled one album unit. Under the new rules, that’s 1,000 paid streams or 2,500 ad-supported streams. Both kinds of streams are now worth more per album unit. The new methodology took effect on January 17, 2026, on the chart covering data from January 2 through January 8.
One day before the new methodology took effect — on January 16, 2026 — YouTube stopped sharing data with Billboard. The negotiations broke down over how streams were weighted. YouTube’s stated position, per their public blog post on the dispute, was that “every stream is counted fairly and equally, whether it is subscription-based or ad-supported.” Billboard’s position, as reported across music industry coverage, was that a person hearing a clip in the background of a video is not equivalent to a person seeking out a song on a paid music service. Both sides had reasonable positions. Both sides walked away. ICEMAN is the first major release that has landed entirely under the new rules — no YouTube data flowing to Billboard, and the new 1:2.5 weighting in effect.
That much is uncontested public record. Joe’s question — did I factor that into my May 4 prediction — is a fair one. I didn’t. My model was built on first-week patterns from Drake’s prior releases plus the pre-release stream-momentum and anticipation signals heading into May 15. All of those inputs were calibrated under the old methodology with YouTube data included. The pre-release momentum was real and was telling me the cultural moment was massive — which is why I revised from the April 200,000-range floor up to 525,000. What none of the inputs were priced against was the chart-counting rules-set that would actually measure the result.
The question Joe didn’t ask but is the actually interesting one: in which direction did the methodology change actually move Drake’s number?
Walk through the arithmetic.
Billboard reported ICEMAN’s first-week streaming-equivalent-album units at 449,000, generated from 462.2 million on-demand official streams across the album’s 18 tracks. Pure album sales contributed 13,000 units. Track-equivalent-album units added 1,000 more. Total: 463,000 equivalent album units.
Of that 462.2 million tracked streams, the public record doesn’t disclose the exact paid-versus-ad-supported split. But the math can be reverse-engineered. Under the new methodology, paid streams convert at 1,000 per album unit and ad-supported streams convert at 2,500 per album unit. The reported 449,000 streaming-equivalent-album units from 462.2 million streams implies a blend ratio of approximately 95 percent paid streams and 5 percent ad-supported streams — roughly 440 million paid streams and 22 million ad-supported streams. That’s consistent with Drake’s audience demographics. His listeners are heavily concentrated on Spotify Premium, Apple Music, and Amazon Music — the paid tiers. Spotify free-tier and other ad-supported channels are minority surfaces for his music.
There’s an independent data point that corroborates that split. Apple Music’s Global Editorial Head of Hip-Hop and R&B, Ebro Darden, publicly reported that Drake’s three new projects accounted for 55 percent of all premium plays on Apple Music in the week following release. Apple Music has no free tier — every Apple Music stream is by definition a paid stream. If Drake dominated more than half of Apple Music’s premium consumption in a single week, the platform mix that runs through his streaming numbers is heavily concentrated on the always-paid surface. That’s the kind of audience profile that the new chart methodology disproportionately rewards.
The Apple-Drake concentration isn’t an accident. When Apple Music launched on June 30, 2015, Spotify was already at 22 million paid subscribers plus 68 million on the free tier. Apple Music started with zero. To compete, Apple needed a marquee artist to anchor premium-listener loyalty in hip-hop — the audience demographic Spotify was leaving the most exposed. Apple recruited Larry Jackson as Global Creative Director in June 2014 from Interscope, and Jackson built the artist-relationship architecture that brought Drake to Apple Music’s launch event at WWDC 2015. The Drake-Apple multi-year deal, reported at the time around 19 million dollars, was the cornerstone artist-loyalty trade — OVO Sound Radio as an Apple Music exclusive, exclusive listening windows on Drake’s releases, the Apple Music premiere of Views in 2016. That 2015 trade built the premium-listener foundation in hip-hop that Spotify never fully replicated. Ten years later, in May 2026, that same foundation translates to 55 percent of Apple Music premium plays in one week being Drake. And that same foundation is exactly what the new Billboard methodology, which weights paid streams 25 percent more generously than the old rules, structurally rewards. The strategic bet Apple made in 2015 — when they were 22 million paid subscribers behind Spotify — is now the structural reason Drake’s first-week number reverse-engineers to 95 percent paid. That’s a ten-year strategy paying off at the chart-math layer in 2026.
Apply that same paid-ad split to the old methodology and the math shifts in a specific direction.
Old methodology, same 462.2 million streams: 440 million paid streams divided by 1,250 paid-per-unit equals 352,000 streaming-equivalent-album units from paid streams. 22 million ad-supported streams divided by 3,750 equals about 5,900 units from ad-supported streams. Total under old methodology, tracked streams only: roughly 358,000 streaming-equivalent-album units.
Compare to the actual 449,000 streaming-equivalent-album units under the new methodology. The methodology change alone — same streams, different weighting — added about 91,000 album-equivalent units to Drake’s number. The new 1:1,000 paid rate is 25 percent more generous per paid stream than the old 1:1,250 rate. The chart-math implies Drake’s measured stream mix is roughly 95 percent paid. Result: Drake was the type of artist the new methodology disproportionately rewards.
That’s only one side of the ledger. The other side is YouTube.
YouTube was excluded under the new methodology. Under the old methodology, YouTube views of audio uploads and Content-ID-matched user-generated uploads of officially licensed music did count toward the Billboard 200 — typically at the ad-supported rate, occasionally at the paid rate for Vevo views from YouTube Music Premium subscribers.
How much of Drake’s YouTube footprint would ICEMAN’s first week have captured under the old rules? Chartmetric data shows Drake’s YouTube channel gained 172.4 million views between May 15 and May 21, the tracking week. Baseline weekly run-rate before the release was approximately 70 million views. The release-attributable incremental views are roughly 100 million. Of those, ICEMAN represents 18 of the 43 tracks released across the three albums — about 42 percent of the catalog by track count. A simple proportional allocation puts ICEMAN’s share at roughly 40 to 60 million YouTube views during the first week.
That’s the sensitivity. And the sensitivity confirms the direction of the methodology effect.
Under the conservative case — 60 million ICEMAN YouTube views, all counted at the old ad-supported rate — old methodology total comes to approximately 388,000 album-equivalent units. Methodology change net effect: a positive 75,000-unit boost for Drake under the new rules.
Under a more aggressive case — 150 million ICEMAN YouTube views, all counted at the old ad-supported rate — old methodology total reaches approximately 412,000 album-equivalent units. Still a positive 51,000-unit methodology boost for Drake.
The only sensitivity corner case in which old methodology produces a higher number than 463,000 requires 200 million ICEMAN YouTube views with at least 50 percent of them counted at the paid rate. That’s a stack of two implausibilities. YouTube’s audience is overwhelmingly ad-supported. Vevo and YouTube Music Premium are minority slices of total YouTube music consumption. The realistic range puts the methodology change in net positive territory for Drake under every defensible assumption.
The mechanism here isn’t unique to Drake. Any artist with a heavily paid-streaming audience benefits from the new methodology. Any artist with a heavily YouTube-driven audience — Latin reggaeton acts with massive YouTube view counts but lighter Spotify Premium penetration, for instance — gets hurt by the same change. An artist with Bad Bunny’s audience-channel mix — heavy YouTube share, broader ad-supported penetration — would face a measurably tougher first-week number under the new rules than the old. The reverse pattern. Drake’s audience type — Anglo hip-hop, premium-tier-skewed, lower ad-supported share — is exactly the audience profile the new methodology rewards.
That’s the structural fact. The methodology change wasn’t a Drake-suppression move. It was a streaming-economics rebalance that disadvantaged some kinds of artists and advantaged others. Drake was on the advantaged side.
I ran the back-calculation across five distinct sensitivity dimensions: paid-versus-ad-supported stream split from 80/20 through 99/1; YouTube-attributable-to-ICEMAN views from 20 million through 250 million; YouTube-views-at-paid-rate fraction from 0 through 100 percent; ICEMAN’s track-share allocation from 30 to 60 percent of release-week YouTube incremental; and a Vevo-subscription-tier inclusion test at multiple thresholds. Across all five sensitivity dimensions, only one corner case produces an old methodology number higher than the actual 463,000. That corner case requires extreme assumptions in two directions simultaneously and isn’t defensible on the public record.
The methodology change helped Drake. The math is robust across reasonable assumptions.
Here’s where the EHIQ AI model failed, and where we will fix it for next time.
When I made the May 4 prediction at 525,000 album-equivalent units, I was anchored on first-week patterns from Drake’s prior releases — measured under the old methodology with YouTube data included. I didn’t update my model when Billboard announced the methodology change in December. I didn’t update when YouTube pulled its data in January. I didn’t price the structural shift.
Two distinct factors moved Drake’s number, and I failed to model both.
The methodology change moved Drake’s number up by roughly 75,000 album-equivalent units. That made my 525,000 prediction less wrong than it would have been under the old rules. Under old methodology, the actual would have landed near 388,000, and my 525,000 prediction would have been 137,000 high — not 62,000 high.
The triple-drop moved Drake’s per-album number down. ICEMAN dropped May 15 alongside HABIBTI and MAID OF HONOUR. Fans split their first-week listening across 43 tracks on three records, not 18 tracks on one. The per-album peak got capped. Drake traded the bigger ICEMAN number for historic catalog dominance — first artist to take the top three Billboard 200 spots in one week.
The net of the two factors: my 62,000 miss is the residual after the methodology change helped me and the triple-drop hurt me. Under a no-methodology-change, no-triple-drop counterfactual, my 525,000 prediction for a single-album Drake release under the old rules would have been roughly in line with the actual outcome — within a few percent. The model wasn’t broken at the cultural-moment level. The model was broken at the structural-rules level.
Joe’s instinct was the right instinct in the wrong direction. He was looking for a methodology answer, and there is one — it just runs the opposite way to his framing. There was no suppression. There was a structural rebalance that helped a paid-streaming-heavy artist and hurt a YouTube-heavy one. Drake won the type of streamer Billboard now rewards. That’s a real fact about how the chart works in 2026.
The deeper point is this: methodology is a price level. When the chart-counting rules change, every artist’s first-week number resets to a new baseline. Models built on the old baseline produce predictions calibrated to the old rules. The rules changed in December and January. I didn’t update the model. Drake’s strategy then traded the per-album peak for historic catalog dominance — a separate decision I also didn’t price.
Methodology is a price level. Drake’s strategy is the trade. Mine was the trade I didn’t price.
The model gets updated.
---
Position disclosure: EMJ Capital holds no position in Universal Music Group (UMG.AS) or Vivendi (VIV.PA). No position in Drake-related catalog rights as of publish date.
---
When the methodology changes, your prediction needs to change. When the strategy changes, your prediction needs to change. When the audience-platform mix changes, your prediction needs to change. The EHIQ AI model gets all three updated by Monday.
If you want predictions that update at the methodology layer, not just the cultural-moment layer, EHIQ Premium is $249/month at eventhorizoniq.com/pricing. You get the daily AM brief, the live dashboard, the receipts ledger of every tracked call, and the analytical scripts behind pieces like this one. The free Substack stays free. The work behind it is what $249/month buys.
Day 279 — same method, every day. Show up. Put in the work. It compounds. Make that investment in yourself.
WHO CARES ABOUT DRAKE THIS IS THE MOST BORING ARTICLE EVER