Innovation Explore

Hype Cycle

The hype cycle traces a technology from the trigger, through inflated expectations and disillusionment, toward broader productive use.

The questionIs the excitement about this technology running ahead of what it can do today, and what does that mean for when we commit?

Model · Jackie Fenn, Gartner, 1995

1
Innovation trigger A breakthrough starts the talk
2
Peak of inflated expectations Hopes run ahead of delivery
3
Trough of disillusionment Early projects disappoint
4
Slope of enlightenment Clearer uses, newer products
5
Plateau of productivity Mainstream use, benefits pay off

Simplified list of the five phases in order. It is not Gartner's chart.

When to use it

A board is asking whether to invest in a technology because everyone is talking about it, and you want to separate what is claimed from what has been delivered.
Several vendors are pitching the same emerging technology, each with a different story about how mature it is.
You are planning pilots and need to decide between moving early, waiting, or watching.
Executives want to know why early projects with a new technology are disappointing, and whether that is a reason to stop.

When to avoid it

You need a forecast of when a technology will mature or what a market will be worth. The curve gives an order of events, not dates or sizes.
The decision is about one vendor or one product. Do due diligence on that product and ignore where the category sits.
You would cite Gartner's placement as evidence without the research behind it. Placements are analyst judgments, and the full reports are paywalled.
The question is how a product spreads through different kinds of buyers, or what sales to expect in each stage. Diffusion of Innovations and the Product Life Cycle fit those better.

Overview

The Hype Cycle describes a common pattern in how people react to a new technology. Publicity and enthusiasm build faster than the technology can deliver, interest collapses when early projects disappoint, and use recovers if the technology proves useful. Consultants use it to start a conversation about timing and risk, meaning whether to move early, wait, or watch.

It is a visual metaphor, not a validated predictive curve. Gartner places each technology by analyst judgment, and published analyses report that few technologies follow the full shape. The vertical axis is expectations, not performance or sales, so a position tells you about attention. Use it to ask what has actually been delivered, and do not use it to estimate dates.

The five phases

Innovation trigger

A breakthrough, a proof of concept, or a product launch starts the conversation. Press interest arrives early, usually before any usable product exists, and commercial viability is unproven. The first version of the curve called this the technology trigger.

Peak of inflated expectations

Publicity produces a few success stories and many failures. Some companies act and most do not. Expectations run ahead of what the technology can deliver, which is the situation the curve is meant to flag.

Trough of disillusionment

Early experiments fail to deliver, interest falls, and weaker suppliers consolidate or exit. Investment continues only if the surviving suppliers improve their products enough to satisfy the first adopters.

Slope of enlightenment

Clearer examples of where the technology helps start to circulate. Second-generation and third-generation products appear, more firms fund pilots, and conservative buyers wait.

Plateau of productivity

Mainstream adoption takes off. Buyers have clearer criteria for judging suppliers, and the technology's benefits are paying off in ordinary use.

What it is not

A schedule. It gives a sequence and not dates, and technologies spend very different amounts of time in each phase. Some never leave one.
A law. One published analysis of Gartner's own curves since 2000 reported that few technologies follow the full shape, and that most of the important technologies adopted since then were not identified early.
A verdict on the technology. A trough placement is not a promise of recovery. An analysis reported by The Economist in 2024 found that most technologies that fell into a trough did not rise again.

The steps

Narrow the technology and the use

State the technology and the specific use at this client, such as automated triage of motor claims, not AI. A category has many uses at different stages, and one dot on a curve misleads.
Collect two kinds of evidence separately

Gather evidence of expectations, such as press volume, vendor claims, funding, and competitor announcements. Then gather evidence of delivery, such as production deployments, measured results, falling prices, and later-generation products. Keep the two lists apart.
Place it provisionally and say why

Use the five descriptions to place the technology, and write down the evidence behind the placement. Where the evidence is mixed, say so. Treat a published Gartner placement as one input, and find out what it rests on.
Choose a stance

Decide whether to move early, take a moderate path, or wait, based on the client's appetite for risk and the cost of being late in its market. Output: a stance and, if it is early or moderate, the scope of a pilot.
Set signals for revisiting

Name the delivery evidence that would change the placement and the stance, and set a date to review it. A placement is a statement about today.

In practice

A regional insurer's board asked whether to build a claims-triage system on a newly fashionable machine-learning approach, because three competitors had announced one. The consultant used the cycle as a set of questions and not as a placement. What had each competitor put into production, and what had they only announced? What had anyone reported about savings, and measured how?

Two of the three announcements turned out to be pilots, and no one had reported production results. The consultant told the board that expectations looked ahead of delivery, and recommended one narrow pilot on motor claims with an exit criterion written in advance. The board scheduled a review in six months, to compare any new delivery evidence against the pilot's own results.

What you leave with

A narrow statement of the technology and the use under consideration.
Two lists of evidence: what is claimed and what has been shown in production.
A provisional placement with the reasons for it and the evidence that would change it.
A stance, early, moderate, or wait, with the pilot scope or watch list and a review date.

Failure points

Placing a whole category

AI, blockchain, and similar labels cover many uses at different stages. One placement for the category tells the client nothing about their own use.

Treating a published placement as evidence

The placement is a Gartner analyst's judgment. Quoting it as proof that a technology has passed its peak adds an authority the method does not have. Ask what evidence supports it.

Reading the trough as a buying signal

The reasoning is that prices are down and hype is gone, so now is the time to buy. Some technologies in the trough recover and some do not, and the position alone does not say which.

Reading the peak as a reason to wait

The client waits for the plateau by default, and its competitors establish a position in the meantime. The curve does not tell you what being late would cost in this market.

Turning the sequence into a timetable

A client plans a budget around plateau in three years. The curve gives no basis for that, and the date becomes a commitment that nobody can defend.

The failures above are usually Framing: the wrong problem was chosen before this framework was applied.

Origin

Jackie Fenn joined Gartner in 1994 and noticed that emerging technologies tended to attract a burst of enthusiasm, then disappointment, then a slower recovery. She labeled five stages on the curve in a 1995 research note that she expected to be a one-off. Clients asked for an update the next year, other Gartner teams built their own, and by 2003 the firm was running an annual special report with many curves. Fenn and Mark Raskino later wrote Mastering the Hype Cycle (Harvard Business Review Press, 2008). The first stage was originally called the technology trigger, and Gartner now calls it the innovation trigger.

Hype Cycle is a registered Gartner trademark, and the research behind each year's curves is reserved for Gartner clients. This note describes the five phases in its own words and does not reproduce the curve. Fenn has since written that the curve is really about how people react to novelty, with the technologies themselves moving toward maturity or obsolescence at their own pace.