A Trend Is a Category, Not a Compliment

"That's so trendy" is usually meant as praise. Sometimes as an insult. Either way, it's an opinion about taste. In trend research, calling something a trend should be neither. It's a classification, and like any classification it needs criteria you can say out loud and defend when someone disagrees.

The awkward part is that the people who write about trends don't agree on what a trend is. That sounds like a flaw in the field. I think it's the most useful thing to learn first.

Four definitions that don't line up

Here's how four well known sources define the word:

Henrik Vejlgaard (Anatomy of a Trend, 2008) treats a trend as a process of change in style and taste, pushed from the top down by identifiable groups of trendsetters and trend creators.
Martin Raymond (The Trend Forecaster's Handbook, 2019) calls it "the direction in which something tends to move and which has a consequential impact on the culture, society or business sector through which it moves." He gives you direction and impact, but says nothing about where it comes from or how long it lasts.
Henry Mason, David Mattin, Maxwell Luthy and Delia Dumitrescu (Trend-Driven Innovation, 2015) describe a new behavior, attitude or expectation that reflects deeper human needs. The product is just where the trend shows up. Nobody starts it.
Albert Postma and Bálint Papp ("Of trends and trend pyramids", 2021) go further: a trend is "a pattern we read in a series of events." It isn't a thing out in the world at all. It's an interpretation made by an observer.

So Vejlgaard looks at styles, Mason and colleagues look at behavior, and Postma and Papp look at the person doing the looking. They can study the exact same object and reach different conclusions, and nobody is wrong. They're just not talking about the same thing.

Try it on one example

Take buying secondhand clothes as a normal way to shop.

Under Vejlgaard, you ask: is there an identifiable group that adopted it first and spread it outward? If yes, it's a trend.

Under Mason, you ask: did people's expectations of shopping change, or only where they shop? If only the channel changed, it isn't a trend yet.

Under Postma and Papp, you ask: are we reading a real pattern, or are we reading our own interest in it existing?

Same object, three different verdicts. That's the lesson. Whenever you write "this is a trend," you should also write which definition you're using. It sounds like a formality. It's actually the difference between an argument and a claim.

Classify in two steps, not four boxes

Most people sort things into a ladder: fad, micro, macro, mega. The problem is that this ladder assumes a fad is just a very small trend. Postma and Papp are explicit that it isn't: hypes, fads and one-hit wonders "have a very short lifecycle and are not considered trends." They're a different category.

So the classification works better in two steps.

Step 1: is it a trend at all?

Hype: lots of noise, no phenomenon underneath. It rises because people talk about it and falls when they stop.
Fad: real adoption, but it rises and falls for the same reason, which is novelty. When it stops being new, it vanishes without changing anyone's behavior. The fidget spinner in 2017 is the textbook case.
Trend: what people do, value or expect has changed, and that change survives after the novelty wears off.

Step 2: only then, what size?

Postma and Papp use duration:

Microtrend, up to 5 years. New events, innovative products and services, emerging practices. Most of what you'll spot in daily life lives here.
Mesotrend, 5 to 10 years. Changes that cut across a whole domain: an industry, a discipline, a market. (You'll also see this middle level called "macro" in other sources.)
Megatrend, 10 years or more. Global forces with breadth and deep effects. John Naisbitt popularized the term with Megatrends (1982), and the European Commission's Competence Centre on Foresight keeps a public Megatrends Hub written in plain language.

One honest warning about those numbers: they're a convention in the foresight literature, not a measurement. Other authors cut the ranges differently. When you classify, cite the range you're using.

A note on the Hype Cycle

You'll run into Gartner's Hype Cycle constantly. It says new technologies pass through five phases: the innovation trigger, the peak of inflated expectations, the trough of disillusionment, the slope of enlightenment and the plateau of productivity.

It's handy vocabulary. Saying "this is at the peak" starts fewer arguments than saying "this is overrated." But keep in mind who publishes it: Gartner sells research subscriptions, and placing technologies on the curve is part of its product. The common criticism is that it isn't a verifiable predictive model, since you only know which phase something was in after the fact. Use it to describe, not to forecast.

A movement is signed. A trend isn't.

This is the distinction people get wrong most often, and it's the most useful line in the whole topic.

A movement is a group of people working toward a shared goal, and saying so publicly. It has a program, spokespeople and a start date. The Bauhaus was founded in Weimar in 1919, published a manifesto and program signed by Walter Gropius, and had a school, teachers and named students. It ended when the school closed in 1933.

A trend is unplanned and anonymous. Nobody signed the Y2K revival. Nobody wrote a manifesto for the "Bauhaus-core" boards you can find on Pinterest. Their goal is to look good, and they fade when the algorithm moves on.

Put simply: a movement is planned, authored and argumentative (it makes a claim about how the world should be). A trend is unplanned, authorless and descriptive (it's a pattern someone reads afterward).

Movements produce trends. Swiss style is a movement. The trend it left behind, which is still very much alive, is a preference for visual clarity and simplicity. Art Nouveau is a movement; the trend it left is nature as ornament. So if you ever write "my trend is the Bauhaus," you've misclassified it. The Bauhaus is a movement that produced trends people still adopt and abandon.

The line I'd keep from all of this: a trend doesn't convince you to want something new. It resonates with something you already needed but hadn't seen, that didn't exist yet, wasn't available, or didn't have a name.

The four kinds of drivers

A trend doesn't float on its own. Something pushes it. Drivers usually fall into four groups:

Technological. Something became possible. The useful question isn't "does the technology exist?" but "since when, and for whom, is it within reach?"
Cultural. What people value, or what an action means, has changed. This is the hardest one to date and the one that explains the most.
Economic. Buying power, relative prices or business models changed. This is often the real driver behind trends that get told as cultural stories.
Regulatory and environmental. A law, a deadline, a shortage. These come with a date built in, which makes them the easiest to cite well.

Most analyses name the cultural driver and stop there. The good ones dig until they find the economic driver underneath. Secondhand clothing is a nice example: there's a real cultural shift in what buying used says about you, but there's also a resale industry with a business model, investors and platforms that profit from it. Resale companies like ThredUp publish annual reports projecting the market's growth, which is useful data, as long as you write down next to it that the company lives off that same market.

What a documented trend includes

If you want to write up a trend so someone else can check your reasoning, it needs six parts:

Name. Short and specific. If you need a paragraph to name it, you don't understand it yet.
Definition. One sentence that says what changed, in whom, and compared to what earlier state.
Drivers. At least two of the four types, each written with a subject, a verb and a date.
Manifestations. Three or more, from industries that don't copy each other. If all three come from fashion, you haven't proved much.
Adoption. Who adopted it first and who is joining now. Everett Rogers' adopter categories (innovators, early adopters, early majority, late majority, laggards) are handy here.
Argued classification. Both steps: first whether it's a trend at all, then its level, with the duration range you're citing.

Here's an exercise worth doing with whatever you currently suspect is "a trend." Run it through both steps, say which definition you're using and why, then write the strongest counterargument against your own classification. Finish with one sentence: what evidence would make you change your mind? If you can't answer that last one, you have a preference, not a classification.

Four definitions that don't line up

Here's how four well known sources define the word:

Henrik Vejlgaard (Anatomy of a Trend, 2008) treats a trend as a process of change in style and taste, pushed from the top down by identifiable groups of trendsetters and trend creators.
Martin Raymond (The Trend Forecaster's Handbook, 2019) calls it "the direction in which something tends to move and which has a consequential impact on the culture, society or business sector through which it moves." He gives you direction and impact, but says nothing about where it comes from or how long it lasts.
Henry Mason, David Mattin, Maxwell Luthy and Delia Dumitrescu (Trend-Driven Innovation, 2015) describe a new behavior, attitude or expectation that reflects deeper human needs. The product is just where the trend shows up. Nobody starts it.
Albert Postma and Bálint Papp ("Of trends and trend pyramids", 2021) go further: a trend is "a pattern we read in a series of events." It isn't a thing out in the world at all. It's an interpretation made by an observer.

Try it on one example

Take buying secondhand clothes as a normal way to shop.

Under Vejlgaard, you ask: is there an identifiable group that adopted it first and spread it outward? If yes, it's a trend.

Under Mason, you ask: did people's expectations of shopping change, or only where they shop? If only the channel changed, it isn't a trend yet.

Under Postma and Papp, you ask: are we reading a real pattern, or are we reading our own interest in it existing?

Classify in two steps, not four boxes

So the classification works better in two steps.

Step 1: is it a trend at all?

Hype: lots of noise, no phenomenon underneath. It rises because people talk about it and falls when they stop.
Fad: real adoption, but it rises and falls for the same reason, which is novelty. When it stops being new, it vanishes without changing anyone's behavior. The fidget spinner in 2017 is the textbook case.
Trend: what people do, value or expect has changed, and that change survives after the novelty wears off.

Step 2: only then, what size?

Postma and Papp use duration:

Microtrend, up to 5 years. New events, innovative products and services, emerging practices. Most of what you'll spot in daily life lives here.
Mesotrend, 5 to 10 years. Changes that cut across a whole domain: an industry, a discipline, a market. (You'll also see this middle level called "macro" in other sources.)
Megatrend, 10 years or more. Global forces with breadth and deep effects. John Naisbitt popularized the term with Megatrends (1982), and the European Commission's Competence Centre on Foresight keeps a public Megatrends Hub written in plain language.

One honest warning about those numbers: they're a convention in the foresight literature, not a measurement. Other authors cut the ranges differently. When you classify, cite the range you're using.

A note on the Hype Cycle

A movement is signed. A trend isn't.

This is the distinction people get wrong most often, and it's the most useful line in the whole topic.

The four kinds of drivers

A trend doesn't float on its own. Something pushes it. Drivers usually fall into four groups:

Technological. Something became possible. The useful question isn't "does the technology exist?" but "since when, and for whom, is it within reach?"
Cultural. What people value, or what an action means, has changed. This is the hardest one to date and the one that explains the most.
Economic. Buying power, relative prices or business models changed. This is often the real driver behind trends that get told as cultural stories.
Regulatory and environmental. A law, a deadline, a shortage. These come with a date built in, which makes them the easiest to cite well.

What a documented trend includes

If you want to write up a trend so someone else can check your reasoning, it needs six parts:

Name. Short and specific. If you need a paragraph to name it, you don't understand it yet.
Definition. One sentence that says what changed, in whom, and compared to what earlier state.
Drivers. At least two of the four types, each written with a subject, a verb and a date.
Manifestations. Three or more, from industries that don't copy each other. If all three come from fashion, you haven't proved much.
Adoption. Who adopted it first and who is joining now. Everett Rogers' adopter categories (innovators, early adopters, early majority, late majority, laggards) are handy here.
Argued classification. Both steps: first whether it's a trend at all, then its level, with the duration range you're citing.

A Trend Is a Category, Not a Compliment

Four definitions that don't line up

Try it on one example

Classify in two steps, not four boxes

Step 1: is it a trend at all?

Step 2: only then, what size?

A note on the Hype Cycle

A movement is signed. A trend isn't.

The four kinds of drivers

What a documented trend includes

Further reading

A Trend Is a Category, Not a Compliment

Four definitions that don't line up

Try it on one example

Classify in two steps, not four boxes

Step 1: is it a trend at all?

Step 2: only then, what size?

A note on the Hype Cycle

A movement is signed. A trend isn't.

The four kinds of drivers

What a documented trend includes

Further reading

A Trend Is a Category, Not a Compliment

Four definitions that don't line up

Try it on one example

Classify in two steps, not four boxes

Step 1: is it a trend at all?

Step 2: only then, what size?

A note on the Hype Cycle

A movement is signed. A trend isn't.

The four kinds of drivers

What a documented trend includes

Further reading