Which Brands Appear Together in the Same Consumer's Purchase History Across Categories?

What are cross-category brand clusters?

Cross-category brand clusters are groups of brands that repeatedly appear together in the same individual's purchase history while never competing in the same category. A running-shoe brand, a dairy brand, and a technical outdoor-gear brand share no shelf space — yet the same consumer buys all three at rates five to six times the panel average. Because these brands never compete, traditional category-level market research never surfaces the connection. Person-level behavioral data does.

Predict measures this with an affinity index: the share of one brand's buyers who also buy another brand, divided by the share of the full panel that buys it, times 100. An index of 100 means parity with the general population; the pairings below run from 530 to 615 — consumers are five to six times more likely than average to hold both brands in their purchase history.

Which brand pairings show the strongest cross-category affinity?

The strongest pairing in the US panel is On Running × Arla Foods at index 615. The table below shows the sharpest non-competing pairings across fashion, food and beverage, health and wellness, and home.

What consumer identities do these brand constellations reveal?

Three coherent identity clusters emerge from the co-occurrence data, each spanning fashion, food, fitness, and home without any internal competition.

The endurance athlete: On Running × Arla Foods × Rab × POC × Bai

The sharpest cluster in the panel. On Running buyers over-index at 615 on Arla Foods (dairy with a protein, clean-eating signal), 609 on Rab technical outdoor gear, 601 on POC performance cycling and ski equipment, and 593 on Bai functional beverages. This is a single, coherent endurance-sport identity — protein, performance drinks, high-spec gear — expressed through brands that never meet in a category.

The quiet-luxury curator: Gucci / Dior / Lululemon × Jo Malone × Estée Lauder

Gucci, Dior, and Lululemon audiences all converge on the same two premium beauty brands: Jo Malone (536–568) and Estée Lauder (545–572). Gucci and Dior buyers extend the pattern into artisanal beverages — Califia Farms (530–532), Topo Chico (538), Cutwater Spirits (536), Vitaminwater (532). These consumers spend premiumly across beauty and better-for-you food rather than defaulting to standard CPG.

The accessible-wellness shopper: H&M × Pret A Manger × Zoe × Billie × FLOS

H&M's affinities skew mass-premium: Pret A Manger at index 574 reaching nearly 4 in 10 of its audience, Flow water at 553, wellness brands Zoe (593) and Billie (531), and — notably — FLOS designer lighting at 558 in the home category. The profile is design-aware and wellness-oriented without being luxury: fast fashion paired with deliberate, better-for-you choices everywhere else.

How can brands act on cross-category purchase data?

Frequently asked questions

What is a cross-category brand cluster?

A cross-category brand cluster is a set of brands from different categories — such as fashion, food, fitness, and home — that the same consumers buy at rates far above the population average, even though the brands never compete with each other. It is detected through person-level purchase co-occurrence, not category market share.

Which brand pairing has the highest cross-category affinity?

In Predict's US panel, On Running × Arla Foods is the strongest pairing at index 615 — On Running buyers are over six times more likely than the average panelist to also buy Arla Foods. On Running × Rab (609) and On Running × POC (601) follow closely.

How is the cross-category affinity index calculated?

Index = (% of brand A's audience who also buy brand B) ÷ (% of the full panel who buy brand B) × 100. An index of 100 means parity with the general population; 615 means brand A's buyers are 6.15× more likely than average to buy brand B. Only pairings with meaningful audience scale are reported.

Why don't traditional market research methods find these clusters?

Traditional research analyzes one category at a time, so brands that never share a category never appear in the same dataset. Cross-category clusters only become visible in longitudinal, person-level behavioral panels that observe the same individual's purchases across fashion, food, fitness, and home simultaneously.

What consumer identities do the clusters reveal?

Three identities emerge from the US panel: the endurance athlete (On Running, Arla Foods, Rab, POC, Bai), the quiet-luxury curator (Gucci, Dior, Lululemon, Jo Malone, Estée Lauder, Califia Farms), and the accessible-wellness shopper (H&M, Pret A Manger, Zoe, Billie, FLOS). No single brand currently owns any of these identities across all categories.

Methodology

Findings are drawn from Predict, Measure Protocol's behavioral intelligence engine, using brand-level pairwise audience overlap across the US behavioral panel. Affinity indices compare the share of a brand's audience that also buys an affinity brand against the full-panel baseline (index 100 = parity). Only non-competing, cross-category pairings with meaningful audience scale are included.