Theories of Consumer Trend Adoption and Influence
Fashion trend diffusion has been explained by several classic theories that describe how a trend moves through society. These frameworks consider whether influence flows from the top down, bottom up, or horizontally across social groups:
Top-Down (Trickle-Down) Theory: This is the oldest model of fashion change, originating from sociologists like Thorstein Veblen (1899) and Georg Simmel (1904). The trickle-down theory posits that fashion trends start with the elite or upper classes and gradually trickle down to the masses. In a hierarchical society, people of lower status look to the styles of those at the top and adopt those fashions to emulate prestige. Meanwhile, the upper classes eventually move on to new styles to differentiate themselves again once the original look has become too common (Kenton, 2023). In Veblen’s terms, fashion is a vehicle of conspicuous consumption – the upper strata introduce new styles as a status display, and as soon as the middle class copies them, the elite abandon those styles for new ones. A historical example is the way luxurious styles from haute couture runways would be knocked off by lower-priced brands; by the time a trend reached department store windows, the couture clientele was on to the next thing (Vannini, 2017). Though rooted in old class structures, elements of trickle-down still occur (e.g. high-end designer trends often set the direction for fast-fashion retailers). The famous “hemline index” theory – where skirt lengths are said to rise in good economic times and fall in recessions – is a related idea of elites setting trends that reflect broader social dynamics .
Bottom-Up (Trickle-Up) Theory: Also known as the “bubble-up” theory, this model is essentially the reverse of trickle-down. It suggests that innovative fashion often emerges from street style, youth culture, or lower-income groups, and then moves upward to be adopted by designers and the fashion elite (Vannini, 2017). This concept gained traction in the later 20th century as observers saw trends like denim jeans, T-shirts, punk hairstyles, and hip-hop streetwear originate in grassroots communities and eventually be embraced on high-fashion runways. The trickle-up theory was formally named in the 1970s (and described by field researchers like Ted Polhemus, who talked about “street style” tribes fueling fashion) (Nag, 2025). A classic example is the punk look: the ripped jeans, DIY aesthetics, and subversive graphics of 1970s London punks were initially anti-fashion statements from working-class youth, but within years these elements were appropriated by designers like Vivienne Westwood and eventually showed up in luxury collections (Kiran, 2021). According to this theory, trend initiation is democratic – fashion bubbles up from the ground, driven by creative individuals or subcultures who influence the mainstream. In today’s world of social media, bottom-up influence is more visible than ever: a small TikTok trend can become a global phenomenon picked up by brands. Trickle-up theory highlights that innovation is not exclusive to couture houses; often the streets, music scenes, or minority cultures are fertile ground for new fashion ideas that later gain mass appeal (Fashion L., 2023).
Trickle-Across (Mass Market) Theory: First articulated by fashion scholar Charles King in 1963, the trickle-across theory (also called mass dissemination or horizontal flow) argues that new fashions spread quickly within and across social groups, rather than strictly above or below one another. In this model, “fashion moves horizontally between groups on similar social levels”. Essentially, each segment of the market has its own innovators and trendsetters (for example, celebrities or influencers who appeal to a particular demographic), and once a trend is introduced, it can be adopted simultaneously across various social classes, given little lag time. This became especially plausible after the mid-20th century, when mass media and mass production grew. For instance, during a fashion season, you might see a particular trend (say, animal-print coats or neon sneakers) appear almost concurrently in designer runway lines, mid-range brands, and fast-fashion stores – each targeting their own customers (Crane, 1985). This happens because designers at all price points are observing the same cultural signals and because trends are reported widely in real time. King pointed out that by the 1960s, the old class-based fashion system was giving way to a more democratic fashion landscape where a teenager in a suburb could adopt a trend as fast as a socialite, thanks to the availability of affordable knockoffs and instant media coverage. The trickle-across theory aligns well with the modern era of instant trend knowledge. When haute couture shows present a new look and fast-fashion retailers produce similar styles within weeks, different income groups end up wearing versions of the same trend at roughly the same time (Levchuk, 2018). In other words, fashion leadership exists at multiple levels of the market, and a trend can “trickle across” through peers and networks just as easily as it can filter from higher status to lower status.
In addition to these three models, other frameworks have been proposed to explain fashion change. Collective selection is one theory (advanced by Herbert Blumer in 1969) which suggests that fashion trends are a collective response to the zeitgeist or spirit of the times (Blumer, 1969). Rather than class imitation or subculture rebellion alone, this view sees fashion as a reflection of the prevailing cultural mood – designers and consumers collectively gravitate toward certain styles because they “feel right” for the moment. For example, after a period of economic austerity or a global crisis, society might collectively embrace more somber, utilitarian styles (or conversely, escapist extravagant styles), and multiple designers will coincidentally showcase similar trends that align with that mood (Sociology, 2024). This theory shifts focus from “who leads and who follows” to the idea that trends arise from a convergence of broader influences (art, politics, technology, social changes) that both designers and the public experience simultaneously (UKEssays, 2025).
Another concept, overlapping with trickle-up, is Polhemus’s idea of “style tribes,” which emphasizes that different groups in society (defined by age, music, lifestyle, etc.) create their own distinctive fashions. These style tribes (e.g. goths, surfers, hip-hop aficionados) incubate trends internally. Sometimes one tribe’s look becomes popular beyond its own boundaries (a bubble-up effect), or sometimes it remains niche. The fashion systems model vs. populist model described by theorists contrasts a centralized approach (akin to top-down) with a polycentric approach where many groups generate trends independently (Flannels, 2024).
In reality, the contemporary fashion landscape sees a mix of these mechanisms at work. A trend might start in street culture (trickle-up), then be adopted by designers (top-down from there to the public), and simultaneously spread through social media to various peer groups (trickle-across). The rapid information flow today blurs the lines – trends can emerge from any level and spread in all directions. High fashion and street fashion constantly trade ideas (Ivy, 2020). For example, consider the sneaker trend: luxury brands draw inspiration from streetwear while sneaker culture itself is influenced by celebrity endorsements (horizontal) and designer collaborations (top-down). Fashion forecasters must therefore track multiple channels of influence. They look at couture runways (for top-down cues), street style and youth markets (for bottom-up cues), and monitor media and influencers (for trickle-across diffusion) (Polhemus, 2012). By combining these perspectives, forecasters build a holistic view of how a trend might propagate and how broadly it might catch on.
The Fashion Trend Life Cycle
Once a fashion trend emerges, it typically follows a life cycle with several stages: introduction, rise, peak, decline, and obsolescence. This five-stage fashion cycle has been a cornerstone concept in fashion marketing for decades (Carbon, 2025). In the Introduction stage, a new style or idea is first introduced – often by an influential designer or a style-setting subculture. At this point the look is new and not widely adopted; it might be a high-fashion runway piece or something spotted on an edgy style leader. If the design gains acceptance, it enters the Rise stage, in which more people start to adopt the trend and it attracts media and consumer attention. This is when early adopters and fashion-forward consumers pick up the style, and it may be featured in magazines or worn by celebrities (Master C. 2024). For example, when a few pop stars or influencers were seen wearing platform heels in early 2021, data showed a 26% rise in popularity of that shoe style among women in Europe by the end of the year. This kind of celebrity endorsement or social media exposure propels the trend upward (LaCkore, 2025).
At the Peak (or maturity) stage, the trend reaches maximum popularity. The style has become mainstream – it’s available at many retailers and worn by a broad audience. When a particular design is at the peak of its popularity, it often gets mass-produced and even oversaturated in the market. For instance, at the peak of the platform heel trend, fast-fashion brands and mass retailers produced inexpensive versions of the designer look, making it ubiquitous (UVU Review, 2024). This saturation often signals that the trend has matured; nearly everyone who wants it has it, and it’s no longer exclusive. Following the peak, the trend enters a Decline. Consumers begin to tire of seeing the style everywhere. As one description put it, when a design is “very common and available throughout the season… people search for something new,” marking the downturn of the cycle. Sales of the trend item slow down. Retailers respond by discounting the remaining stock – during the decline stage, you’ll find the once-hot items on clearance racks as stores make room for new trends (Heuritech, 2024). Finally, in the Obsolescence stage, the trend is considered “out of fashion.” Consumers have fully moved on to newer styles, and the old trend is no longer desirable at any price. The item might feel dated or stale, and wearing it may even carry a stigma of being out-of-style. At this point, retailers have long since stopped reordering the style, and whatever leftover product exists is deeply discounted or eliminated (Textile, 2023).
It’s important to note that different trends travel this cycle at different speeds. A fad (a short, intense craze) might rocket from introduction to obsolescence within a few months – peaking fast and then dying out as quickly as it came. A classic, on the other hand, is a style with enduring appeal that might never truly become obsolete (for example, a Chanel-style tweed jacket or blue jeans go through waves of popularity but never vanish entirely). Many trends follow the bell-shaped curve of consumer adoption, aligned with Everett Rogers’ diffusion model: first adopted by a small group of innovators, then spreading to early adopters, on to the majority, and finally laggards, before disappearing. A visual metaphor often used is the “fashion curve,” which charts the number of adopters over time – starting low at introduction, rising to a peak, then descending (Infomineo, 2024).
In recent years, the classic fashion cycle has accelerated and become less predictable. The “Introduction” and “Rise” phases can happen almost simultaneously now, thanks to social media virality. A trend might go from a niche Instagram post to widespread adoption in a matter of weeks, compressing the cycle. The Textile Focus magazine observes that technology and social networks have rendered the cycle “much shorter and less forecastable than in the past.” (Textile, 2023).
Trends can flame out faster (or conversely, certain trends can linger longer due to continued online interest). Nonetheless, most trends do eventually hit a saturation point and decline. An interesting phenomenon in fashion is the revival of obsolete trends. There’s truth to the adage that “everything comes back in style.” A general rule of thumb in fashion is a roughly 20-year cycle for revival (Fibre, 2021). This means that a trend popular two decades ago might start to look fresh to a new generation too young to remember it the first time. For example, in the 2020s, styles from the early 2000s (“Y2K fashion” like low-rise jeans and baby tees) have made a comeback about 20 years later (Infomineo, 2024). This cyclical resurgence is one reason fashion forecasters also look to the past for clues – archived trends can be mined and reimagined for the present. In summary, while the timeline of trends can speed up or slow down, the trajectory of birth, growth, saturation, decline, and rebirth remains a fundamental pattern in fashion evolution.
Historical Overview: From Early Trend Reporting to AI Forecasting
Early origins (19th to mid-20th century). The practice of fashion forecasting in some form can be traced back centuries. In the court of Louis XIV in France, fashion information began to be communicated systematically – nobility would send updates on Paris styles to provinces, an early attempt to stay ahead of trends (Hansen, 2003).
By the late 19th and early 20th century, with the rise of industrialized fashion and consumer capitalism, forecasting became more organized. The first professional color forecasting service, the Textile Color Card Association of America, was founded in 1915, issuing seasonal color swatches and trend guidance to manufacturers (Insead, 2023).
In 1927, Tobe Coller Davis started the Tobe Report, often cited as the industry’s first dedicated fashion trend forecasting company. Tobe’s weekly reports advised retailers on coming style trends during the 1920s–40s, an era when American buyers were eager for guidance from Paris and broader cultural currents. These early efforts were often tied to trade publications or associations and focused on specific aspects like color or retail trends (Truthplus, 2011).
After World War II, the forecasting field expanded alongside the booming fashion industry. Many modern forecasting agencies emerged from the 1940s through the 1960s. Industry observers note a pattern: several key trend agencies were founded in the wake of economic crises or shifts, when businesses were especially eager to anticipate the future. For example,
The Doneger Group began in 1946 (just after WWII) as a New York buying office that evolved into a trend service. Paris-based Promostyl was established in 1966 by François Vincent-Ricard amid a slump in the French textile sector (Impression, 2020).
Peclers Paris was founded in 1970 by Dominique Peclers with a mission to “democratize style” by making trend insights widely available. In 1980, renowned forecaster Lidewij Edelkoort launched Trend Union (Edelkoort is often called a futurist for her uncannily accurate lifestyle and fashion predictions). Nelly Rodi followed in 1985, bringing a new method that blended creativity with marketing and consumer research to explain how trends start and how consumers perceive them (Impression, 2020).
By the late 20th century, trend forecasting had matured into a distinct profession, with agencies in fashion capitals producing polished trend books, color forecasts, and consulting services for global clients.
The digital transition (1990s–2000s). For decades, a small elite circle of forecasting companies (perhaps a dozen major players worldwide) essentially dictated trends from the top, in a relatively closed loop. This began to change dramatically in the 1990s with the advent of the internet and real-time global communication (Promostyl, n.d), (Profile Magazine, 2012) WGSN (Worth Global Style Network), founded in 1998 by brothers Julian and Marc Worth, was a game-changer: it moved forecasting onto a constantly updated online platform. Instead of waiting for semi-annual trend books, subscribers could access daily reports, runway analyses, and street style photos from around the world (The Impression, 2025), (Colin Turek Blog, 2015).
An insider in 2007 noted that “WGSN changed the landscape with online daily trends, and books are seen as being of secondary importance now.” (China Daily. 2007).
Traditional agencies like Carlin and Peclers had to add web-based updates to keep up. The 2000s thus saw a hybrid period – companies still used physical trend books and presentations, but increasingly supplemented them with digital content and faster trend turnaround. The forecasting timeline itself began to speed up, as fast-fashion brands like Zara proved that reacting to trends in near real-time (designing and delivering products in a few weeks) could outperform long-range predictions (King, n.d.). Some critics argued that predicting styles 18–24 months out had become an “outmoded system when fast fashion demands products in store within six weeks”. Nevertheless, most brands continued to plan at least 2–4 seasons ahead, relying on forecasters for early signals while building in more flexibility to adjust closer to the season (McKinsey & Company, 2023), (Brand Experts, 2022).
AI-enhanced forecasting (2010s–2020s). In the last decade, fashion forecasting has been increasingly transformed by big data and artificial intelligence. Traditional forecasters used qualitative methods – essentially human “curation” of trends – but now agencies are augmenting that with machine learning to analyze massive datasets. As one researcher noted, the industry has split between “traditional and data-led forecasting agencies,” and the use of AI has grown due to the explosion of available digital data (Cassidy, 2023). Since the 2010s, trend companies and fashion brands have started mining everything from social media images and influencer posts to Google search patterns and retail sales data for trend insights. These AI tools can detect patterns across millions of data points far beyond the scope of manual analysis. For example, Heuritech, a Paris-based forecasting tech firm founded in 2013, uses proprietary image recognition to scan about 3 million social media images per day for fashion signals. It categorizes consumer posts into groups like “edgy early adopters,” “trendy mainstream consumers,” etc., to quantify how a style spreads from niche to mass. This allows Heuritech to spot a burgeoning micro-trend (say a certain sneaker style or print) when it’s still confined to edgy influencers, and project if it will grow. Luxury brands like Dior and Louis Vuitton have partnered with Heuritech to inform their design and marketing – for instance, using its data to decide on new handbag designs or to gauge customer response to a revival of a vintage design (Heuritech, 2024).
Established agencies like WGSN have also invested in data analytics. WGSN’s service today touts a blend of “expert trend forecasting combined with data science” to refine its predictions (Caldana, 2023). The overall goal of these AI-driven systems is to make trend forecasts faster and more accurate. As The Guardian Reported, forecasters now feed algorithms with years of runway imagery, Instagram posts, online shopping data and more, which “can help spot emerging trends more accurately and, crucially, more quickly.” (Biehlmann, 2023).
