Why Taste Fails as Product Design Methodology in the AI Era
Objective Principles and User Research Must Anchor Product Decisions Against Unfalsifiable Opinions
1The Taste Moat Fallacy
As artificial intelligence automates competent interface generation, the tech industry increasingly claims that taste is the primary remaining differentiator. However, relying on subjective preference rather than verifiable evidence turns product reviews into unfalsifiable opinion contests. Real product judgment requires factual grounding combined with the agency to challenge prevailing room consensus.
2Subjectivity Masquerading as Fact
Taste reflects personal exposure and admiration rather than objective quality, making disagreements inherently unresolvable between designers. A 2001 wine study proved perception alters easily when white wine dyed red tricked 54 tasters into using red wine descriptors. Similarly, personal design preferences satisfy only the individual, whereas commercial products must serve thousands.
3Aesthetic Usability Trap
In 1995, Hitachi researchers Masaaki Kurosu and Kaori Kashimura demonstrated that perceived usability correlates heavily with interface beauty rather than actual functionality. Generative design tools amplify this bias by producing inherently polished layouts. When superficial polish becomes cheap, reviewing software based on taste provides zero verifiable failure conditions for teams.
4Status Reinforces HiPPO Culture
Treating taste as an evaluation standard functions primarily as a status marker that inherently favors the highest-paid person's opinion. Ron Kohavi highlighted how HiPPO dynamics undermine objective user needs across organizations. A culture built on taste simply rebrands senior executive authority with improved typography, silencing junior contributors lacking formal organizational power.
5Foundational Laws Provide Grounding
User experience relies on long-standing scientific principles rather than arbitrary aesthetics. Paul Fitts demonstrated target acquisition mechanics in 1954, while William Edmund Hick quantified decision time relative to option volume. Jakob Nielsen and Rolf Molich formalized ten usability heuristics in 1990, converting personal opinions into auditable, testable interface criteria.
6Understanding Contextual Boundaries
Established design laws require careful contextual application rather than blind enforcement. Hick's Law assumes equally probable choices, showing diminished relevance in specific scenarios studied in 1985. Similarly, Fitts's original cockpit stylus measurements do not perfectly map to mobile thumb stretches. Uncritical application turns rigorous empirical laws into mere cited opinions.
7Continuous Research Validates Principles
Principles establish baseline probabilities, but live testing reveals actual user behavior. Microsoft data from Ronny Kohavi showed only a third of controlled experiments moved target metrics positively. As continuous discovery advocate Teresa Torres emphasizes, weekly customer contact by core product builders bridges the gap between general laws and specific user cohorts.
8Aesthetic Friction and Consequences
While attractive design creates positive emotional forgiveness, excessive aesthetic focus can mask critical functional failures. Jony Ive's design leadership produced celebrated hardware but also yielded Apple's flawed butterfly keyboard, resulting in a $50 million class-action settlement in 2022. Flawed mechanics eventually outlast visual charm regardless of executive taste.
9Audience Demographics Alter Preferences
Designers cannot substitute their personal aesthetic sensibilities for user evaluations. A Nielsen Norman Group study evaluated five screenshots across age brackets: young adults favored flat design, whereas older users rated it lower. Evaluating beauty requires structured measurement tools like Microsoft's Desirability Toolkit alongside task completion and error metrics.
10Feedback Loops Build Judgment
Daniel Kahneman and Gary Klein established that intuitive expertise requires regular environments and immediate feedback loops. Junior designers rarely experience closed loops between UI decisions and long-term retention outcomes. Teaching foundational design heuristics gives early-career staff objective standing to challenge menu lengths and missing recovery flows without relying solely on seniority.
11Misinterpreting Steve Jobs
Steve Jobs is frequently cited to justify ignoring user research, yet Apple systematically tracked market data and behavioral trends. When Jobs commissioned Paul Rand in 1986 for the $100,000 NeXT logo, Rand presented one solution backed by rigorous deductive reasoning. Modern founders often succumb to survivorship bias by copying executive certainty without analytical rigor.
12Balancing Metrics and Intuition
Overcorrecting toward pure measurement creates paralysis, as documented when Doug Bowman left Google in 2009 after testing 41 blue shades. Extreme data reliance dodges strategic judgment entirely. Teams require an objective baseline where empirical UX laws, weekly customer evidence, and clear architectural reasoning overrule ungrounded opinions across the hierarchy.