Brand Reflection Series · Episode 1 · companion to Sound fingerprint & Strategic topography
A spot's picture either commits to the brand color or commits to a world. Of 47 Czech TV spots from 13 brands, 16 flood with brand color, 11 cede the picture to a world, and 20 sit in between. Few realize they are choosing.
Method
47 publicly available Czech TV spots across 13 brands, sampled at one frame per second. Per-frame palettes via K-means (k = 5) in canonical CIE-LAB. For each spot we compute its dominant chromatic hue, the brand's anchor hue, and the circular distance between them. Strip and radial views aggregate frames across all spots of a brand; the bar chart in section 01 shows every spot separately.
Distance, in degrees on the hue wheel, between each brand's anchor swatch and the dominant chromatic color the spot actually puts on screen. One bar per spot (47 spots from 13 brands), bars filled with the swatch; the dashed reference at 90° marks the perceptual threshold beyond which the two are, effectively, different colors. 16 spots sit below 20°. 11 sit above 100°. 20 sit in between.
Each spot reads as a contact sheet: every frame a vertical column, every column split top to bottom by K-means cluster sizes. Strips are ordered by their distance from brand color, so the page reads brand-led at the top, world-led at the bottom. A saturation sparkline sits below each brand name; the arc glyph summarizes its shape.
Per-spot polar histograms. Angle is hue (0° red, 120° green, 240° blue); radial length is saturation-weighted density. The tick on the rim, in anchor color, marks each brand's logo swatch. In two cases the tick aligns with the spot's color mass. In the rest, the tick is somewhere the picture barely visits.
Hierarchical clustering on pairwise Székely energy distance between every pair of brands' weighted LAB palettes. If the brand-vs-world split were just a sector artifact, brands from the same sector would cluster tightly. They do not. McDonald's two spots sit close to each other; everyone else clusters across sector boundaries, by their color signature.
Saturation tension (the std of per-frame saturation across the spot) splits the corpus by category too. Banking and automotive flatten; performative formats swing. This is a related but distinct axis from the brand-vs-world split; both shape the spot.
Sector means with 95% bootstrap CI on saturation std
Bootstrap resamples brand-level values within each sector. With two to three brands per sector the intervals are wide; the gap between corporate (banking, automotive) and performative sectors is still visible.
01
Across 47 spots from 13 brands: 16 commit to brand color (gap below 20°), 11 cede the picture to a world (gap above 100°), and 20 sit in the middle. The middle is not a third style; it is what happens when no decision is made.
Why it mattersThe extremes are deliberate. The middle is drift. A brief that does not pick a camp lands in the middle by default, and the middle reads as visually weak in every category we measured.
02
Same sector, opposite strategies: Penny Market's five spots compress the gap to 8.6° (brand red on 36% of screen); Albert's five spots stretch it to 174° (Albert blue on 2% of screen). Both Czech retailers. Both consistent internally. The five spots within each brand cluster tight, the two brands sit at opposite ends of the field.
Why it mattersA brand can choose either camp. Albert's warm-bistro world is a legitimate creative move; so is Penny's promo flood. What is not legitimate is choosing accidentally.
03
Banking world-leans (Komerční banka 142°, Fio 85°), automotive world-leans (AAA Auto 159°, Škoda 79°), telecom brand-floods (T-Mobile 12°, Vodafone 15°), retail splits cleanly (Lidl 6°, Penny 9°, Albert 174°). QSR sits brand-led (McDonald's 22°), beverage splits (Kofola 11°, Pilsner 102°).
Why it mattersThe category sets the gravity well; the brief decides whether to follow it. Going against your category's default is bolder almost by construction, but only when the spot can carry the weight.
04
Brand-led brands run anchor share 25 to 56% (Lidl 56%, Penny 36%, Vodafone 31%, T-Mobile 19%, McDonald's 25%). World-led brands sit under 5% (Albert 2%, Pilsner 0.2%, AAA Auto 0.001%, Komerční banka 0.0%). The middle group is between 5 and 25%; almost no brand sits in the 10 to 20% region for long.
Why it mattersBrand color presence is bimodal in practice. Either the swatch is the picture or the swatch is just the logo. Targeting "moderate brand presence" sounds reasonable but usually produces the muddle no one remembers.
Five moves a strategist or creative director can run on next week's pitch.
01
Before greenlight, agree with the client which camp the spot belongs to and what the target gap is. Brand-color: target gap below 20°, anchor share above 30%. World-color: target gap above 100°, anchor share below 10%. The middle is the muddle. Pick a side.
02
Lidl's blue and the supermarket aisles light up together; Penny's red and the price tags blur into one signal; McDonald's yellow and the fries belong on the same shot. Forcing a brand-color flood when the product fights the swatch (a green logo on a beige insurance ad) reads as desperation. If the product cannot carry the color, switch camps.
03
Albert lives in 2% anchor, AAA Auto in 0.001%, Komerční banka in 0%. All three are intentional. The picture is the spot; the logo end-card is the brand. Stop chasing more brand color on screen as a goal. Chase a more deliberately built world instead.
04
Run this pipeline on the client's three competitors. Show them where the category lives on the gap axis. Then name where the next spot will land. Distinctiveness becomes a coordinate. Especially useful where every brand has converged (banking, automotive, premium beverage).
05
The pipeline runs on rough footage in minutes. If the brief committed to brand-color (gap under 20°) and the dailies are scoring 80°, you have caught the drift before it is baked into the edit. Cheap insurance on expensive shoots.