SHUR IQ / Micro-Drama Intelligence / Issue No. 21 / Week of 2026-07-27
Visualizations Previous Issue Archive
21
Issue No. 21 · W32-2026 · 2026-07-27 to 2026-08-02

HBO Max Shipped a Vertical Feed With Nothing Original In It, and ByteDance Enters This Ranking at Number Six

Every one of the 23 companies this index tracked going into the week scored zero, which has not happened in twenty-one issues. It happened in the busiest week the category has had. Warner Bros. Discovery launched HBO Max Shorts on 2026-07-28, a vertical feed cut by an in-house AI tool from catalogue Warner already owns, with no original vertical drama in it. World 1 unveiled a category of 30-second vertical-native sports on 2026-07-30, funded by Mark Cuban and Abu Dhabi’s Palms Sports with ten Olympic medallists among its founders. ByteDance’s paid microdrama app had surfaced four days before the window opened. None of the three was a company this index scored. The board was pointed in the wrong direction, and the correction is what this issue does: ByteDance enters at #6, immediately above Netflix, and Warner Bros. Discovery enters at #20.

0
Tracked companies that moved, in a week with three platform-class events
#6
Where ByteDance enters, one place above Netflix, on a business it has never announced
1 of 6
Platform giants with a vertical surface that have commissioned original vertical drama to fill it
21 of 23
Companies whose published composite does not reconcile with their own dimension scores
Where this week’s events landed, 2026-07-27 to 2026-08-02
One dated event touched a tracked company, and it was a denial. The three that mattered landed on companies the board did not score. Two of them are now scored.
Dated events, 2026-07-27 to 2026-08-02, plus the two carried in from 2026-07-22 and 2026-07-24 · sources in the W32 corpus

The Window at a Glance

All flatZero gainers, zero decliners, 23 holds across a full seven-day window. Unlike Issue 20’s three-day window, these flat calls carry full evidentiary weight. An all-flat board on seven days of research is a result, and what it says is that the index was measuring the wrong set of companies.
Warner ships emptyHBO Max Shorts launched 2026-07-28 in US iOS test: a vertical scrollable feed of trailers, clips and bonus material, personalised on watch history, with an in-house AI tool doing scene-level selection and human editors cutting the results vertically. No original vertical content is commissioned. Warner says sports will follow.
Netflix says noStoryboard18 reported on 2026-07-28 that Netflix was preparing an India microdrama entry around October 2027 with Yash Raj Films and Red Chillies as likely partners. Netflix denied it within a day. That scores nothing, and it is the first time Netflix has been put publicly to the question on the format’s largest growth market and has answered no.
Sport goes vertical-nativeWorld 1 unveiled Microsports on 2026-07-30: original competitions where a world title is decided in about 30 seconds, built for the feed rather than adapted from an existing sport. Fastest and Strongest are defined, a modular arena is built, and revenue is planned from advertising, sponsorship and licensing.
Board expandsByteDance joins at #6 on monetization 88 and distribution 87. Warner Bros. Discovery joins at #20 on distribution 72 against a content score of 25. The board goes from 23 companies to 25, so ranks below #5 shift by one or two places for reasons that have nothing to do with performance.

What the Board Could See and What It Could Not

The second consecutive issue where the coverage was the story
Scored

Nothing. Twenty-three companies returned no in-window event that priced. The single dated event on a tracked company was Netflix being reported into an India launch on 2026-07-28 and denying it within a day, which is evidence of a decision not to move.

Could not be scored, and now can be

Warner Bros. Discovery shipped a working vertical feed on 2026-07-28 and was not on the board. ByteDance had been selling microdrama subscriptions in the United States since March 2026 and was not on the board. Both are added this issue, at #20 and #6. Issue 20 recommended the ByteDance addition and this issue carries it out.

Still cannot be scored

World 1’s Microsports competes for the same feed attention and the same advertiser budget as everything on this board without being drama at all, and the index has no way to represent it. Paramount+ is testing microdramas on its mobile app with mocked-up titles, no talent and no date, which sits below the entry bar Google cleared with a named slate. Both are recorded as open questions for Issue 22 rather than scored.

The Giants Are Converging on the Interface

Four vertical feeds, no original commissions behind them

Warner’s arrival makes the pattern hard to miss. Netflix has a Clips feed and no original vertical commission anywhere in the world. Amazon has a partial Clips rollout that has now missed its own summer deadline for a third consecutive issue. Google announced a slate in March 2026 with named talent and has shipped no title roughly five months later. Warner has a feed built from catalogue it already owns.

Disney is the exception, and the gap between Disney and the rest is the whole finding. Locker Diaries launched 2026-02-16 and returned with a Descendants wave on 2026-07-25, which makes Disney the only platform giant running a repeating original vertical programme rather than a single experiment.

The strategic read is that four of the largest media companies in the world have decided the vertical feed is a retention feature for the catalogue they already have, and one has decided it is a content business. Those are different bets with different costs, and only one of them competes with ReelShort and DramaBox for anything. The gap tracking this reopens to critical after four weeks of narrowing, because adding a fourth empty feed widens it.

Visualizations

The window in five views
Where this week’s events landed, 2026-07-27 to 2026-08-02
One dated event touched a tracked company, and it was a denial. The three that mattered landed on companies the board did not score. Two of them are now scored.
Dated events, 2026-07-27 to 2026-08-02, plus the two carried in from 2026-07-22 and 2026-07-24 · sources in the W32 corpus
Twenty-three holds and two new entries
No company changed rank on merit. ByteDance enters at #6 and Warner Bros. Discovery at #20, so every incumbent below #5 reads one or two places lower than in Issue 20.
SBPI composite, 0 to 100 · W32-2026, change from the W31 baseline published 2026-07-26
Six giants with a vertical surface, one with anything original in it
Warner Bros. Discovery became the fourth platform giant to ship a feed and the fourth to ship it empty. Disney is still the only one commissioning original vertical drama.
Platform-giant deployment status at 2026-08-02 · W32 corpus
How the five forecast methods scored on an all-flat week
An all-flat week hands a perfect direction score to any method that predicts stability, so four methods tied at 100% without forecasting anything. mean_reversion scored zero because it has predicted up for every company in every week since W12.
5 methods × 23 forecastable companies · predictions-W32-2026.json, recorded 2026-08-02T06:13
The index’s own arithmetic does not reconcile
For 21 of 23 companies the published composite does not equal the weighted sum of its five dimensions. Twenty issues of independent rounding on the composite and the dimensions accumulated into this.
state/archive/W31-2026.json against the 20/20/25/20/15 weighting · reproduce with reports/W32-2026/check_drift.py

SBPI Stack Ranking

Structural Brand Power Index · W32-2026 · the authority ranking for the micro-drama category: 25 companies scored on dated public evidence. Twenty-three held, none gained, none declined, and two entered.
Rank Company Tier SBPI Score W32 Change What Moved It
Scoring Methodology

The Structural Brand Power Index evaluates 25 companies across five weighted dimensions: Content Strength (20%), Narrative Ownership (20%), Distribution Power (25%), Community Strength (20%), and Monetization Infrastructure (15%). Scores update on verifiable public signals within the 2026-07-27 to 2026-08-02 research window, measured from the W31 baseline published 2026-07-26. Tier thresholds: Dominant 85 to 100, Strong 70 to 84, Emerging 55 to 69, Niche 40 to 54, Limited below 40. A change of 0.40 points or more in one window is material. Two entities entered this issue and carry no change figure, because they have no prior week.

A defect in the arithmetic, disclosed

For 21 of the 23 incumbents the stored composite does not equal the weighted sum of that company’s five dimension scores, with drift from 0.05 to 1.85 points. The cause is that prior issues nudged the composite and the dimensions as separate quantities and twenty issues of rounding accumulated. Recomputing every company from its dimensions would move Amazon from #11 to #13. It is not reconciled here, because rescoring 21 companies and rewriting a published rank would break the week-over-week chain twenty issues were built on. Issue 22 owns the decision.

W32 Movers

No score changed. Two companies entered the board.
NEW · ByteDance / LimeShorts · enters at 67.55 · #6

Content 45, narrative 38, distribution 87, community 80, monetization 88. ByteDance runs three microdrama products at once. LimeShorts is a paid standalone app in US test since March 2026, charging $19.99 a week or $199.99 a year for the full library, with coin packs from $4.99 to $69.99 for individual episodes and a paywall after the first free episodes. PineDrama is free and standalone. A microdrama feed runs inside TikTok itself. LimeShorts replaced the earlier Mytopia app and licenses its catalogue from third parties.

Monetization at 88 and distribution at 87 carry the entry. The paywall, the coin ladder and the subscription tiers are live and priced rather than announced, and they sit on the largest short-video audience in the West. Content at 45 and narrative at 38 are held low on purpose: the catalogue is licensed with no originals, and ByteDance has never made a public claim on this category. LimeShorts surfaced through reporting rather than announcement, and a company that has not said it is in this business does not own the story of it. The entry lands one place above Netflix at 67.30, which is the useful comparison.

NEW · Warner Bros. Discovery / HBO Max · enters at 46.85 · #20

Content 25, narrative 30, distribution 72, community 48, monetization 55. HBO Max launched Shorts on 2026-07-28: a vertical scrollable feed in the bottom navigation carrying trailers, clips and bonus material from the existing library, personalised on watch history, in test with select US iOS users. Warner says it will add sports and expand devices and markets. An in-house AI tool performs scene-level analysis and human editors verticalize the results.

Content at 25 is the lowest score on the board above the niche tier and it is the point of the entry: Shorts carries no original vertical drama and re-cuts things Warner already owns. Netflix sits at 27 and Amazon at 31 for the same reason. Narrative at 30 reflects that Warner framed this as content discovery rather than as a microdrama product. Distribution at 72 sits below Disney’s 91 and Amazon’s 88.1 because the feed is one platform, one operating system and a subset of one market. Monetization at 55 credits a working subscription business with no microdrama mechanism at all: no coins, no per-episode gate, no paid tier for the format.

Ranks Moved and None of It Was Performance

Board expansion, not competition
Read the ranks carefully this issue

ByteDance enters at #6 and Warner at #20, so former ranks 6 to 18 shift down one place and former ranks 19 to 23 shift down two. Netflix reads #7 instead of #6. Amazon reads #12 instead of #11. Mansa reads #25 instead of #23. None of that is a company doing better or worse. The build script asserts that the order of the 23 incumbents is unchanged from W31 and fails if it is not.

Flat Lines

Twenty-three companies held. Six of those silences carry information.
Dates that have now expired, not just slipped
  • Amazon, 59.80. Summer full-US Clips availability unmet at 2026-08-02, still described as select US users, press page unchanged since 2026-05-08. Third consecutive issue of slippage against the same date, with the season two-thirds gone.
  • Mansa, 24.10. The May-to-July slate window closed 2026-07-31, inside this reporting period, with no confirmed July premiere and no metric beyond an aggregate 6.5M first-30-day impressions figure from April. This date did not slip. It expired.

Both hold flat because missing a self-imposed deadline is not scoreable under current rules. This is the third consecutive issue recommending a decay rule for publicly missed self-set dates.

Shipped surfaces with no numbers

Netflix, 67.30. The Clips feed is live and still carries no usage figure and no original vertical commission, and the company denied an India entry on 2026-07-28. Google / 100 Zeros, 59.45, has a named slate from March 2026 with no dated title roughly five months on. Disney, 77.70, disclosed nothing in the seven days after a franchise premiere, which is a real absence: no platform sitting on good numbers stays quiet about them.

Pressure from outside the board

VERZA TV, 33.15. World 1’s Microsports landed directly in VERZA’s niche on 2026-07-30, backed well beyond anything VERZA has raised. Held flat rather than marked down, because World 1 builds original sports competitions rather than sports microdrama, and a competitor’s launch is not an event on the incumbent. Lifetime / A+E, 57.90, watched a direct legacy peer ship a vertical surface and did not respond.

Ordinary silence

ReelShort, DramaBox, JioHotstar, iQiYi, Holywater, DramaReels, CandyJar, GoodShort, ShortMax, NetShort, GammaTime, COL Group, Viu, RTP, KLIP and Both Worlds recorded no dated event. Two items in this group are worth continuing to watch: DramaBox’s $100M raise reaches week 19 of silence, and COL Group’s monetization-95 anomaly enters week 16 with its revenue half still unanswered.

Structural Gaps

The active ledger · W32-2026. Two new gaps enter, one on an adjacent format the board cannot represent and one on the index’s own arithmetic. The platform-giant content gap reopens to critical.
Reopened · narrowing to critical · Platform-Giant Vertical Surface ↔ Original Microdrama Content

Warner Bros. Discovery became the fourth platform giant to ship a vertical feed on 2026-07-28 and the fourth to ship it with nothing original behind it. The count is now Netflix, Amazon, Google and Warner with feeds and no original commissions, against Disney alone with a repeating original programme. Adding a fourth empty feed widens this gap rather than closing it, so four weeks of narrowing reverse.

NEW · critical · Published Composite ↔ Weighted Dimension Sum

For 21 of 23 incumbents the stored composite does not equal the weighted sum of the five dimension scores under the index’s own 20/20/25/20/15 weighting. Drift runs from 0.05 to 1.85 points, largest on COL Group at -1.85, GammaTime at -1.65, Amazon at -1.58 and Netflix at -1.15. Only NetShort and DramaReels reconcile exactly. Recomputing would move Amazon from #11 to #13. Not reconciled this issue; Issue 22 decides between restating the affected weeks and adopting the composite as canonical.

NEW · medium · Microdrama Board Scope ↔ Vertical-Native Non-Drama Formats

World 1’s Microsports competes for the same feed attention and the same advertiser budget as everything on this board without being drama at all. The board scores vertical drama and carries one sports name in VERZA TV. Issue 22 needs a scope decision before an entity decision: either the index tracks vertical-native entertainment and admits Microsports, or it tracks microdrama and states that adjacent formats are out of scope. It currently does neither explicitly.

Narrowing · Tracked Board Coverage ↔ Largest Distribution Entrant

Acted on. ByteDance joins at #6 and Warner Bros. Discovery at #20, closing the two largest holes. It does not close because the window opened a third in World 1, and because Paramount+ remains outside on the strength of a town-hall slide with mocked-up titles and no talent attached.

Escalating · US Supply Land Grab ↔ Organic Demand Reality

Seven new supply-side entries in eleven days: four US entrants on 2026-07-22, ByteDance’s paid app on 2026-07-24, HBO Max Shorts on 2026-07-28 and World 1’s Microsports on 2026-07-30. No demand measurement was published in the window against any of it. The last reading, 820 total branded queries on 2026-07-23, is now ten days old.

Open · US Standalone Content Bet ↔ Incumbent Interface Bet

Warner landed squarely on the interface side on 2026-07-28: a vertical feed built from existing catalogue with an in-house AI tool doing scene selection, and no commissioned vertical content. ByteDance remains the only entity running both bets at once. The split is now four interface-only giants against one content-and-interface giant and one pure content bet.

Open · Platform-Giant Deploy Velocity ↔ Holdout Risk

Unchanged in direction and sharper in evidence. Warner shipped a working feed inside the window while Amazon’s summer full-US Clips date passed a third consecutive issue unmet and Google’s March slate reached five months with no dated title. The spread between giants that ship and giants that announce widened again.

Open · AI Production Tools ↔ Content Output

This window’s AI datapoint was about repackaging rather than production. HBO Max Shorts uses an in-house AI tool for scene-level metadata analysis so human editors can verticalize existing footage, which is AI applied to catalogue mining rather than to origination. Q2’s China reversal, priced in Issue 20, stands unchallenged.

Implications

What this window changes for the instrument and for the category
For the categoryFour of the largest media companies in the world have now decided that the vertical feed is a retention feature for catalogue they already own. Disney alone is treating it as a content business. Only Disney is competing with ReelShort and DramaBox for anything, and the other three feeds are a different product aimed at churn. Anyone sizing the competitive threat to the pure-plays should count one giant, not four.
For the boardTwo additions closed the coverage hole this index has been reporting on itself for two issues. The lesson is about latency rather than judgement: ByteDance had been selling subscriptions in the United States for four months and Warner shipped inside a week, and both were found by reading trade press rather than by any signal in the system. A watchlist of unscored entrants, refreshed weekly, would have caught both earlier.
For the scoring rulesAmazon has missed the same self-set date for three consecutive issues and Mansa’s delivery window expired inside this one. The index recorded both as stability. A decay rule for publicly missed self-set dates is now the third-issue recommendation and should be decided rather than repeated.
For the instrumentThe composite does not reconcile with the dimensions for 21 of 23 companies, and the discrepancy is large enough to move a published rank. Every ranking this index has produced since roughly Issue 10 carries it. Disclosing it is the minimum; Issue 22 has to choose between restating and redefining, and the choice determines whether the dimension scores mean anything or are commentary attached to a number computed elsewhere.
For the forecast layerAn all-flat week gave four of five methods a perfect direction score without any of them forecasting anything, which is the clearest demonstration yet that direction accuracy on this board is close to meaningless. Two defects surfaced with it, both long-standing, both fixable, and neither previously quantified.
Data quality flags open at close of Issue 21

Four flags stand. The composite reconciliation defect is described above. The nightly semantic layer produced no fresh priors for a second consecutive issue: all seven digests covering this window are byte-identical apart from their timestamps and their predictive signals still describe the W31 window, so the research phase ran without direction again. ReelShort’s parent is still recorded as Jiuzhou Culture while public sources attribute ReelShort to COL Group and Crazy Maple Studio, which would make #1 and #18 related entities scored independently, and this is carried unresolved from Issue 20. World 1’s Microsports is dated 2026-07-30 from secondary coverage because the Hollywood Reporter original is paywalled and returned HTTP 402.

Predictions

Five methods scored against W32 actuals, and two defects the all-flat week exposed
How the five forecast methods scored on an all-flat week
An all-flat week hands a perfect direction score to any method that predicts stability, so four methods tied at 100% without forecasting anything. mean_reversion scored zero because it has predicted up for every company in every week since W12.
5 methods × 23 forecastable companies · predictions-W32-2026.json, recorded 2026-08-02T06:13
Reading the scoreboard honestly

Four methods scored 100 percent on direction and none of them forecast anything. Every company was flat, so any method that predicts stability was always going to be right, and persistence and kg_default predict stability unconditionally. The direction column cannot separate a good forecaster from a constant this week. The error column can, and it does: 0.000 for the two constants, 0.065 for naive_momentum, and above 1.2 for the two methods that predicted movement.

Defect one · mean_reversion has said “up” for every company since W12

Across 21 consecutive prediction files covering W12-2026 to W32-2026, mean_reversion has emitted direction “up” for every company, every time, and has never once predicted down or stable. Its magnitudes vary, so something is computing, but the direction label is a constant. Its accuracy in any week is just the share of companies that rose that week, which is why it scored zero here. Either fix the sign logic or retire the method.

Defect two · kg_optimized’s label has been decoupled from its magnitude since W22

Issue 20 flagged this as a single observation. It is not. The defect appeared at W16, worsened, and has been total for eleven consecutive issues: at W14 the method labelled 14 companies up and 7 stable with no disagreements, and from W25 onward every single prediction is labelled stable while carrying a magnitude. In W32 those magnitudes ran from +0.55 on ReelShort to +1.87 on Mansa, several at 0.95 confidence. The result is that kg_optimized scores 100 percent on direction and 1.217 on error in the same week. Until the label derives from the magnitude, its direction accuracy should be excluded from every scorecard and the W16 to W32 figures already in the evaluation log treated as void.

What no method predicted

Nothing in the forecast layer models board expansion, and board expansion is what changed the index this week. Every method forecasts a delta for a company already in the state file, so there is no mechanism to say that an untracked company is about to matter. That is the specific failure being corrected by hand for the second issue running. ByteDance was identified in Issue 20 prose and added in Issue 21 prose, and no model saw it or could have.