Every year Gartner publishes its Hype Cycle. If you’ve been to a business conference, there’s a very good chance you’ve seen it on a slide explaining how some new technology will disrupt the market you’re in (and probably put you out of business).
It shouldn’t be taken too seriously. It’s still a tool worth having, especially when you’re doing strategic planning or key account planning, as long as you know what it can and can’t tell you. Here it is as a set of questions and answers, with a few small animations.
What is the Gartner Hype Cycle?
The Gartner Hype Cycle is a chart of how excitement about a new technology rises and falls over time. Gartner analyst Jackie Fenn drew the first one in 1995, for a one-off research report. Other analysts and clients liked it so much that it became an annual report on emerging technologies (Wikipedia). Fenn and Mark Raskino later wrote a book about it, Mastering the Hype Cycle (2008).
Gartner now produces more than 130 Hype Cycles every year, tracking over 1,900 innovations across industries, functions and regions (Gartner). The ones that get the publicity are the Hype Cycle for Emerging Technologies and, these days, the one for AI.
What are the five stages?
One curve, five stages
- Innovation trigger. A possible breakthrough gets noticed. There are proof-of-concept stories and plenty of media interest, but often no usable product and no proof that anyone will pay for one.
- Peak of inflated expectations. Early success stories, usually alongside a lot of failures. When a technology is at the peak, you tend to hear about it on the nightly news, and soon reality has to kick in.
- Trough of disillusionment. Experiments fail to deliver, interest drops and some of the companies selling the technology fail or give up. Investment carries on only where the survivors fix what went wrong.
- Slope of enlightenment. The benefits become clearer and better understood. Second and third generations of the product appear, and more companies run pilots.
- Plateau of productivity. Mainstream adoption. It becomes clear how to judge one supplier against another, and the technology starts paying off.
Each technology on a Gartner chart also carries a marker for how many years Gartner expects it to take to reach the plateau. I find that marker more useful than where the dot sits.
Does every technology follow the curve?
No.
In 2016, Michael Mullany looked back over twenty years of Gartner’s emerging technology charts. He found that few technologies actually travel through a recognisable hype cycle, and that most of the important technologies adopted since 2000 weren’t spotted early (Mullany). In 2024, The Economist traced breakthrough technologies over time and estimated that only about a fifth go through all the stages. Lots of technologies become widely used without the rollercoaster. And of those that fall into the trough, it estimated that six in ten never rise again.
The trough is where most technologies stay
There are other fair criticisms. It isn’t really a cycle, it isn’t built on a model anyone can test, and the stage names describe a mood rather than a measurement, so two people can look at the same technology and put it in different places (Wikipedia).
So what happened to NFTs?
They make a good test. Gartner put NFTs at the Peak of Inflated Expectations in its 2021 chart for emerging technologies, a few months after Christie’s sold a Beeple NFT for more than $69 million (Decrypt).
Two years later, a study by dappGambl looked at 73,257 NFT collections and found that 69,795 of them had a market value of zero (The Register).
NFTs, from the peak to the trough
So Gartner (correctly!) called the fall. The harder question, and the one the chart can’t answer, is which side of figure 2 NFTs end up on: the four that climb out, or the six that don’t.
Where is AI on the curve now?
Generative AI went into the trough on Gartner’s 2025 Hype Cycle for Artificial Intelligence, as organisations learned what it can and can’t do (Gartner). The same chart put AI agents and AI-ready data at the peak, as its two fastest-moving technologies (Gartner, August 2025).
The 2026 edition came out at the end of August. Gartner sums it up as a change of question: from whether a company can build with AI to whether it can account for what its AI is doing and what it costs. Worries about generative AI have moved from what the models can do to whether the applications are reliable and pay their way, and Gartner picks out AI observability, the tools that spot an AI quietly getting things wrong, as one of the most important entries on the chart (Gartner).
Agents are where hype and reality are furthest apart. Gartner’s first Hype Cycle for Agentic AI, published in April 2026, puts agentic AI at the peak of inflated expectations (Gartner). Only 17% of organisations have deployed AI agents, more than 60% expect to within two years, and Gartner expects more than 40% of agentic AI projects to be cancelled by the end of 2027 because of cost, unclear value or weak risk controls. That’s what the start of the slide into the trough looks like.
AI on the 2025 chart
None of that means generative AI has failed. In the trough the hype has gone, prices and expectations are more sensible, and the useful applications are easier to see. For a buyer, that’s often the calmest time to learn what a technology really does.
When should you commit to a new technology?
Avoid buying in just before a technology falls from the peak. You’ll pay the most, and your capital ends up tied up in something that would have earned more elsewhere.
I’d argue the best time to commit to an emerging technology is when it has climbed some way up the slope of enlightenment. By then you’ve missed most of the volatility, and there’s enough evidence to judge it.
Three moments to commit
The exception is a technology that changes how you compete. There, the trough is the cheapest place to learn, with small pilots and no big contracts, so that you’re ready when it starts to climb.
How do I use it in planning?
Treat it as a prompt for a planning conversation. Four ways I’d put it to work:
- Place your key accounts’ big bets on it. For each important customer, ask where the technologies they’re investing in sit. A customer at the peak will buy quickly and may regret it. A customer in the trough is looking for someone who can make it work.
- Place your own offer on it. If what you sell depends on a technology at the peak, expect your buyers’ mood to fall with it, and plan the conversation you’ll have in the trough.
- Read the years-to-plateau marker. It tells you more about your planning horizon than the position of the dot.
- Check it against your own numbers. Your pipeline, your customers’ budgets and your own pilots are better evidence than anyone’s chart.
For language companies there’s an obvious one to try. Where does AI translation sit for each of your key clients? They won’t all be in the same place, and the conversation is different at each stage.
Remember, though, that the Hype Cycle is someone’s opinion, published by a firm that sells research to the companies on its charts. It may well be written with an agenda in mind. Useful, but never on its own.
Sources: Gartner, Gartner Hype Cycle; Gartner (2025) Gartner Hype Cycle identifies top AI innovations in 2025; Gartner, Enterprise guide to generative AI; Gartner (2026) Hype Cycle for Artificial Intelligence, 2026; Gartner (2026) The latest Gartner Hype Cycle for Artificial Intelligence puts control ahead of capability; Kandaswamy, R. (2026) What the 2026 Hype Cycle for Agentic AI reveals, Gartner; Wikipedia, Gartner hype cycle; Fenn, J. and Raskino, M. (2008) Mastering the Hype Cycle, Harvard Business Press; Mullany, M. (2016) 8 lessons from 20 years of hype cycles; The Economist (2024) Artificial intelligence is losing hype; Decrypt (2021) Gartner Hype Cycle ranks NFTs at peak of inflated expectations; The Register (2023) 95% of NFTs are worthless.