Color Psychology in UX: What M&M's Data Reveals

Color kills conversions quietly. A palette that feels instinctively right in a brand presentation can — and routinely does — suppress click-through rates, erode trust signals, and fragment recognition across platforms without anyone noticing until the quarterly numbers arrive.

M&M’s recent color controversies, dissected in detail by Rita Kind-Envy on UX Collective, offer a surprisingly rigorous case study in how chromatic decisions interact with consumer psychology at scale. The brand’s historical color manipulations — removing red in 1976 over unfounded Red Dye No. 2 fears, reintroducing it in 1987, controversially retiring tan in 1995, adding purple in 2022 after a public vote — were rarely accidental. Each shift was a calculated signal, and each generated measurable public response. That’s not confectionery marketing trivia. That’s a longitudinal dataset on color perception most UX teams would never have the budget to commission.

Why Color Is a UX Variable, Not a Brand Preference

The neuroscience behind color response is fairly settled: humans process chromatic information before conscious cognition engages. In a UX context, this means your primary CTA color, your error state palette, and your product photography backgrounds are all making decisions before your copy gets a chance to argue its case.

Kind-Envy’s analysis highlights how M&M’s color choices consistently exploited this — warm tones like red and yellow activate urgency and appetite cues, while cooler additions (the green character’s associations, the blue introduction in 1995) were used to signal modernity and approachability. The commercial logic mapped directly to shelf psychology. For digital teams building product pages on Shopee or Lazada, the same mechanism applies: warm accent colors around promotional pricing tend to outperform cool ones in A/B tests, not because warm colors are universally superior, but because they trigger familiar urgency associations baked in by years of retail conditioning.

The implementation trap most teams fall into is treating color as resolved once a brand guide exists. It isn’t. Color behaves differently across surfaces — an orange CTA button on white desktop looks authoritative; that same orange on a dark-mode mobile interface at 375px can read as aggressive or low-contrast depending on the ambient rendering environment.

The Craft Argument: When Constraint Produces Clarity

There’s an adjacent lesson in Hyejin Song’s archival project on Jiro Ono’s sushi tools, covered by It’s Nice That. Song’s documentation of Jiro’s bench — the specific utensils, their worn surfaces, their deliberate limitation — frames craft mastery as the product of constrained toolsets, not expansive ones. Jiro doesn’t use forty knives. He uses a few, exceptionally well.

This is a useful corrective for design systems work. The temptation when building a color token library is inclusivity — accommodate every possible use case with a palette broad enough to never say no. The result is usually a 200-token system that no one uses consistently, and brand coherence that degrades at precisely the touchpoints where it matters most: third-party platform storefronts, partner co-branded content, campaign microsites built under deadline pressure.

Constraint is a feature. M&M’s operated for years with a tightly controlled palette of six to seven colors, each with defined character and consumer association. When they added purple in 2022, it was news precisely because the palette was disciplined enough that one addition registered culturally. If your brand uses 34 accent colors, adding a 35th won’t move anyone.

---

Southeast Asia Implementation: Mobile-First Color Realities

For regional teams, color psychology in UX has a layer of complexity that Western design literature underweights: screen rendering variance across a fragmented Android device ecosystem, cultural color semantics that diverge meaningfully from European defaults, and the practical reality that a substantial portion of your audience is experiencing your brand on a mid-range phone in variable lighting conditions.

Red, for instance, carries fortune and celebration associations across much of Southeast Asia — a positive valence that aligns usefully with urgency mechanics in e-commerce. But that same red deployed in error states or warning messages creates a semantic collision with users who’ve been trained to associate it with good outcomes. Lazada and Shopee have both navigated this by using red primarily for promotional mechanics while reserving orange or amber for cautionary UI states — a deliberate decoupling of cultural association from functional signal.

Multi-language interfaces add another constraint: when your button label shifts from four characters in Thai to twelve in Bahasa Indonesia, the color contrast ratios and visual weight of that CTA change even if the hex code doesn’t. Build color decisions with text-length variance in mind, not just the English version of your UI.

From Intuition to Instrumentation

The M&M’s case study is useful not because candy is design-adjacent, but because it demonstrates that color decisions made on cultural intuition can be validated — or invalidated — by behavioral data. The brand’s internal research on which characters drove purchase intent by demographic wasn’t guesswork; it was structured measurement of a variable most companies treat as settled at the brand guidelines stage.

The same instrumentation is available to digital teams today. Hotjar and VWO both support palette-level A/B testing at a cost accessible to mid-market brands. Google Optimize’s deprecation pushed many Southeast Asian teams toward Optimizely or home-built flag systems, but the testing infrastructure exists. The gap isn’t tooling — it’s the organizational habit of treating color as a hypothesis rather than a decision.

Run your next color change as an experiment before you push it to production. Define the metric — conversion rate on the primary CTA, scroll depth on the landing page, add-to-cart rate on the product card — and treat the result as data, not verdict. Color psychology gives you the prior; behavioral testing gives you the posterior. Use both.

Key Takeaways

• Treat color as a measurable UX variable with defined success metrics, not a brand preference resolved at the style guide stage.
• In Southeast Asian markets, decouple cultural color associations from functional UI signals — red for promotion, amber for caution — to avoid semantic collisions.
• Constrained design systems produce stronger brand recognition; resist palette expansion until the existing tokens are consistently applied across all channels.

The deeper question for regional digital teams isn’t which colors convert best in the abstract — it’s whether your current design system is instrumented well enough to know when a color is working and when it’s quietly losing you money. Most aren’t. That’s the gap worth closing.

---

At grzzly, we work with digital marketing teams across Southeast Asia on design systems and conversion-focused UX that hold up across the region’s fragmented device and platform landscape — from Shopee storefronts to LINE-integrated brand experiences. If your palette decisions are still living in a PDF brand guide rather than a tested, token-based system, that’s a conversation worth having. Let’s talk