01 · The curve bends fast.
Across 1,420 ads scored through the RoastIQ pipeline against benchmark pool v.2026-05, the relationship between first-distinctive-cue latency and Brand Impact is monotonic, steep and remarkably stable. Ads whose first distinctive brand cue lands inside 1.5 seconds carry a median Brand Impact of 71 and land Scale 62% of the time. The same cohort, re-cut so the cue lands past 6.0 seconds, collapses to a median of 31 and 71% Rebuild.
The slope is not linear. It bends sharply between 1.5s and 3.0s, the region where the short-form scroll-stop median sits. Past 3 seconds the marginal cost of every additional half-second of latency falls because the audience has already left; the cut is now writing brand assets onto an empty room. The implication for planners is that the actionable lever is concentrated in the first three seconds. Latency past that point is mostly damage control.
| Latency band | Sample (n) | Median BI | Median Build Brand | Scale rate | Rebuild rate |
|---|---|---|---|---|---|
| 0.0–1.5s | 412 | 71 | 68 | 62% | 8% |
| 1.5–3.0s | 461 | 58 | 54 | 31% | 22% |
| 3.0–6.0s | 336 | 44 | 41 | 14% | 54% |
| >6.0s | 211 | 31 | 28 | 4% | 71% |
"The 1.5-second rule is not a creative preference. It is a structural property of every short-form vertical surface where the median scroll-stop is 1.9 seconds. Late brand is late audience."
02 · Industry latency, ranked.
The eight-industry ordering is the most stable cross-industry pattern in our benchmark. FMCG brands lead at a median 0.8 seconds, driven by mascots, sonic logos and ownable colour systems that work as the cue rather than around it. Finance trails at 5.3 seconds, late enough that on a 1.9-second scroll-stop surface the brand handshake is invisible to the median viewer.
FMCG's lead is not magic. It is the residue of a decades-long discipline around distinctive assets, colour systems (Coca-Cola red, Cadbury purple), packaging silhouettes (Liquid Death can, Toblerone wedge), mascots (Magalu's Lu, M&M characters), that compress the brand handshake into a single frame. Auto and finance are the inverse: creative traditions built around establishing shots, narrator voice-over, and end-card badge reveals that worked on TV and broke on short-form vertical.
Telco sits in the middle for a structural reason: ownable colour systems (T-Mobile magenta, Orange's orange, Vodafone red) carry the cue early even when the rest of the creative is uninspired. The industry's median is propped up by paint, not by creative discipline. Remove the colour system and telco's latency collapses into the auto-finance band.
03 · What a distinctive asset is.
The Ehrenberg-Bass Institute's definition, refined over twenty years of category research, is the working definition here: a distinctive asset is a non-name, non-logo cue that is uniquely owned by the brand and triggers brand recognition without the wordmark present. Distinctive assets are the brand's hooks into the long-term memory structures that drive future purchase. They are not optional, decorative or stylistic, they are the asset under management.
Three properties separate distinctive assets from generic creative codes. Uniqueness: only your brand uses this cue in this category. Fame: a meaningful share of the category audience associates the cue with your brand. Consistency: the cue appears across cuts, formats, and years. A new colour palette every campaign is not a distinctive asset; it is a refresh.
04 · The five most-owned cues in our pool.
Five non-logo distinctive assets dominate the top decile of Brand Impact across our 1,420-ad pool. Each is uniquely owned (no category competitor uses the same cue), famous (recognised without wordmark by the brand's category audience), and consistent (appears in the brand's cuts across multiple years).
- Magalu's "Lu" mascot, Brazilian retailer Magazine Luiza's CGI brand character. Cue median latency 0.4s. Carries the brand handshake in a single frame; the Lu silhouette is the brand. Brand Impact median 78 in the cohort.
- Cadbury purple, Pantone 2685C. Cue median latency 0.9s. The colour is the brand handshake before any product or wordmark appears. Brand Impact median 72.
- Liquid Death can shape, the gothic-tallboy silhouette. Cue median latency 0.7s. Packaging-as-brand-asset; the silhouette reads as Liquid Death from across the room. BI median 75.
- T-Mobile magenta, Pantone Rhodamine Red. Cue median latency 1.1s. The colour system carries the cue early even when the cut otherwise underperforms. BI median 69.
- Coca-Cola contour bottle, the 1915 silhouette. Cue median latency 0.6s. The shape is the brand independently of the colour or the wordmark. BI median 77.
"Every cue in the top decile is non-logo. The brands winning Brand Impact at the frame-one timescale are not winning with their wordmark, they are winning with mascots, colour systems, packaging silhouettes, and sonic logos."
05 · Why the logo is not enough.
The logo is the brand. The distinctive asset is what tells the viewer the brand is there before the logo arrives. Confusing the two is the most common brand-team failure mode in our pool. Cuts that rely solely on logo-as-cue underperform cuts with non-logo distinctive assets by a median of 18 Brand Impact points, even when both are placed at identical latency.
The reason is mechanical. On a 1.9-second scroll-stop surface, the logo competes for attention with every other on-screen element in the first second. A wordmark in the corner reads as decoration. A mascot mid-action, a recognisable colour wash, a packaging silhouette mid-reveal, these read as content. Content earns the second second of attention; decoration does not.
The second reason is memorial. Distinctive assets are over-coded into long-term memory because they pre-date any single ad. The viewer has seen Lu hundreds of times before this cut; the cut is a memory refresh, not a memory build. The logo is doing memory-build work in every cut, which is expensive when you have 1.5 seconds to spend.
06 · Three worked examples.
The Magalu Lu mascot lands mid-action at 0.4 seconds, well inside the 1.5-second window. The mascot itself does three jobs simultaneously: brand handshake, emotional opener, and category identifier (retail / e-commerce). Brand Impact 81, Build Brand 76. The cut also scores top-quintile on Beat the Skip (84) because the mascot's silhouette is its own pattern-break inside a feed of human-led UGC.
An illustrative re-edit cycle: the same Cadbury footage cut twice. V1 opens on the gifting moment, the purple wash and bar reveal arrives at 5.0s, brand-late. Brand Impact 51, Sharpen. V2 cuts the gifting setup in half, brings the purple wash forward to 3.0s, and the bar reveal to 3.4s. Brand Impact 68, Scale. The Build Brand lift is +17 on the same shot list. No re-production, only a brief shift on where the brand handshake lives.
A 60-second cinema cut down-edited to 30 seconds for short-form vertical. The drone establishing shot survives the cut down; the badge reveal lands at 6.0 seconds, past the 75th-percentile scroll-stop on every surface in our cohort. Brand Impact 39, Build Brand 34. The cut has no non-logo distinctive asset to fall back on, the silhouette of the vehicle reads as "luxury sedan," not as the brand. The category-typical trap, in canonical form.
07 · Implications by stakeholder.
// For the CMO
Audit your category's median cue latency before approving any brief. If you are in auto, finance or luxury fashion, your category baseline is already late by default. Set a hard 2.0-second cue ceiling on every short-form vertical cut. Promote the latency metric to a brand-health KPI alongside SOV and prompted recall.
// For the brand team
Inventory your distinctive assets, rank them by frame-one viability, and brief at least one into every cut. If your brand book has only the logo as a "brand asset," your brand book is a logo book. Build out a non-logo asset system: colour, mascot, packaging silhouette, sonic, recurring talent.
// For the agency
Stop pitching "brand-reveal" structures on short-form cuts. The reveal is for cinema. On vertical short-form, the brand belongs in the first 1.5 seconds, and it belongs as the distinctive asset, not as the logo slate. Brief the edit around the cue, not around the story.
// For the media buyer
Your cuts with late cue latency are silent bleeders. They look fine on completion rate (the viewer who completes has already given the cut the audience it deserves) and devastating on Brand Impact. Pull spend from any cut whose distinctive cue lands past 3.0s on a short-form vertical buy.
// Three things to change tomorrow
08 · Method.
What we do not claim. Brand Impact correlates with public engagement and click-intent signals. It does not measure in-market sales, ROAS, attributed conversion or brand recall. Cadbury and other brand examples are illustrative re-edit cases drawn from our benchmark pool with brand-identifying detail held back for licence reasons. The 8-industry ordering is stable across every benchmark_pool_version since 2025-Q3.