Mixed Reality

Gartner Hype-Cycle: Everything You Need To Know

Understand the Hype-Cycle plot; how it is drafted, how accurate has it been in the past and how useful it is in offering practical advice to investors, entrepreneurs and consumers.

As per an FBI estimate annual non-health related insurance frauds amount to around $40 billion and insurance companies, at present, rely on a combination of claim analysis methods, computer applications and even private detectives to investigate fraudulent claims. Meanwhile, a maturing tech called Emotional-AI claims to detect frauds using the audio analysis of the claim calls.

Mysterious? Magical? Useful?

Emerging techs sometimes look like that - like superpowers; they excite and surprise us bringing many such questions in our minds. They make promises about taking us places, helping us meet new people, ideat, imagine and create things - faster, easier. If such an emerging tech catches media attention; magazines publish articles on them, news anchors talk of them, films flaunt them with a vivid imagination - while we wait in anticipation and discuss them with awe.

However, when it comes to these techs, like many other things, our fundamental challenge lies in foreseeing the constantly shaping future. How useful is the new tech? Is it a worthy investment? What do its performance trends look like? Will it substitute the use of an older tech or create new uses or even new needs?

Introduction to Hype-Cycle

New technologies hold new promises. Media, entrepreneurs and investors, are always on the lookout to identify and capture a nascent technology and try and repackage it into a news story or into a viable business idea, essentially creating hype around emerging tech in the market. Investments are made, further research starts, enthusiasts develop prototypes, while discussions on tech forums and word-of-mouth do their rounds. Technological innovation is one of the key competitive differentiators, and it is not surprising to see why predictions about the commercial viability of an emerging tech run wild.

What returns are on the chart? Can it be mass-produced? How long before it's adopted by the buyers to solve real problems and provide new opportunities to businesses?

What is Hype-Cycle?

The Gartner hype cycle gives us a graphical representation of the life of upcoming technologies in terms of their impact on businesses, people and society in the coming five-ten years. Gartner says it profiles and closely follows more than 2000 technologies to narrow down upon a succinct set of must-know emerging technologies and trends. It tries to answer questions like - How and if at all, new technology will mature in the coming time.

In the past 25 years, since Gartner is publishing this annual graph, there have been many hits and misses. In this post, we’ll try to understand what Gartner is trying to tell us and analyse if it is useful in providing a roadmap for business opportunities and applications or is a misleading and unscientific predictor of an inherently unpredictable future.

To begin, these are the stages an emerging tech goes through before realising its true potential.

Hype-Cycle-General

5-Stages of the Hype-Cycle

  1. Innovation Trigger: A public announcement or demo of a tech triggers the cycle, hopes and expectations around an array of new possibilities arise swiftly. (“AI is an exceptional breakthrough!”)
  2. Peak of Inflated Expectations: Fervour around the tech crescendos, deafening hype leads to unrealistic expectations and hasty investments. (AI will transform our lives!”)
  3. Trough of Disillusionment: Reality seeps in, investments fail and expectations fall. Market plunges further due to negative press. (“AI isn’t that great.”)
  4. Slope of Enlightenment: With steady improvements, tech contextualises and use-cases develop, expectations rise around what is achievable. (“AI can be useful in these cases...”)
  5. Plateau of Productivity: Techs underlying value translates into realistic applications and best practices are established. (“AI is commonplace. This is how I use it.”)

The Hits - When the Cycle gave useful predictions

In cases where the techs have loosely followed the Hype-cycle predictions, there were certain returns for early adopters and investors who could foresee their potential. Out of more than 200 new techs identified only a handful like - cloud computing, 3D printing, Natural Language Search - were identified earlier and travelled somewhat predictably on the graph’s path.

4G, for example, appeared around 2005. At that time it was rising on the hype cycle, peaking its hype in 2015. Five years later now, 4G is common with many network providers offering the service across wide geographies. 4G, we can say, is at the plateau of its productivity and is making way for 5G in the coming future. Similarly, cloud computing that peaked around 2009 is now a trillion-dollar industry. NFC peaked around 2011, now with advances in chip tech in bank cards, it is nearing mass adoption with the global market expected to reach more than $150 billion by 2023. 3D printing peaked in 2012 but failed in mass adoption by direct consumers. However, do not put aside the hype-cycle yet as 3D printing’s enterprise use now amounts to $10 billion.

Even though the cycle wasn’t precisely spot-on, when it came to identifying techs at the right time or giving them their warranted attention, many discoveries while looking transitory turned out to be foundational for the future generation of techs and related applications. We had cases where one tech reshaped into several other leaving lessons for them. For example, when speech recognition and generation was featured in the first hype-cycle, as climbing the plateau of productivity, it wasn’t much evolved. With time, the developed-in complexity to deliver near-human performance - after advances in machine learning and AI. As a result, as per 2018 hype-cycle, it is still advancing on the plateau of productivity and is scheduled for mass adoption in coming 2-3 years. In a similar manner, NoSQL gave way to MongoDB, Cassandra etc.

Take the case of Open Source as a licensing model peaked in 2003 disappearing later and appearing with different names. OSI resulted in widely beneficial changes like community code sharing, of commoditization infrastructure applications and cloud computing. Similarly, data analysis featured as data mining in the 90s, changing to analytics in 2000s and finally, to accommodate the increasing scale and scope of data, to big data in 2010.

The Misses - When to ignore the predictions

Even though useful in cases, Gartner’s Hype-Cycle has been criticised for several reasons - firstly in the way it is drafted, and secondly for inconsistencies between cycles as techs do not match well with their uptake in practice.

Many things around us move in cycles, gases in the atmosphere, moon’s phases, seasons and so on. These things go in repetition with time but techs do not. Critics hence question the hype ‘cycles’ nomenclature. Techs might have a lifetime and they might advance with years but they do not go back to the stage where they started, to follow the same evolutionary steps again.

Similarly, other terms used in the cycle i.e. to label various stages can often be misleading to people who are looking for practical advice on adopting a tech.

Should they just watch a tech when it is at Peak of Expectation or wait for it till it overcomes the Trough of Disillusionment and starts climbing the Slope of Enlightenment. Is a position on the Plateau of Productivity a confirmation for the tech to be a productive investment now? Looking at the terminology and associated possibilities we can see that the hype-cycle while tracing the journey of different techs on a common differential plot doesn’t offer an actionable perspective for the stakeholders who are watching the tech.

Moreover, if one can say that the cycle offers fair advice to business investors on where to invest or not - the investment decisions vary depending upon the risk appetite of the investor and ever-changing market conditions. Big companies can afford to invest in R&D and reap the benefits of being early adopters, while smaller ones, with limited finances, will wait and watch for prototypes and proven results to come out.

Not just at the surface level of nomenclature, if we look closer, at the mathematics of the plot, we’ll find more inconsistencies.

The Hype-cycle is a combination of two-curves - a ‘hype-level’ curve that looks like a distribution-curve and a ‘tech-maturity’ curve that follows an s-curve. The hype-level is a depiction of human attitude towards new things, more specifically innovation, more than it is of any particular tech itself. It is a curve of expectations. On the other hand, s-curve is a curve of adoption. It says that a new tech takes time to be understood and leveraged, once it matures - it rises in performance, it comes to a plateau depending upon nature and limitations of the tech.

The-hype-cycle-and-its-stages-indicators-9

Gartner combines these two curves to give us the Hype-cycle curve but doesn't provide a mathematical relationship between the two. However, both the curves measure separate phenomenon and unless there is a proven mathematical relation between them, they cannot be simply added, as they depict different variables. Maybe as a consequence of this internal inconsistency, Gartner does not provide a joint mathematical formula for the combined plot.

If we dig deeper and look at the various featured techs and how their predictions faired in the coming years we’ll have a better idea on how to read the plot i.e. what to take seriously and what to ignore conveniently.

About one-fourth of the featured techs, when they appear to stray from the plotline, were discarded. 802.16 WiMax, for example, was a competitor to LTE for the 4th generation cellular standard. It first appeared on the Hype Cycle in 2005 at peak hype, then was relegated to the trough of disillusionment in 2006 before disappearing eventually. Similarly, desktop Linux for business first appeared at peak hype in 2003 and moved towards the trough by 2005. However, Linux never caught up to windows as a mainstream desktop OS, while VMware enabled users to run Linux like an app on windows, pushing it out of the hype-cycle after 2006. Mesh networking first appeared in 2003, and then featured in 9 out of next 11 year’s graphs, but eventually disappeared and never moved past the trough of disillusionment.

Looking further inconsistencies make themselves more apparent. Of the 250+ techs featured in the past two decades around 65 appear only once never to reappear again like Social TV (2011), Truth Verification (2004), Folksonomies (2006), Expertise Location (2007) and others. Some survive today but not the way they were predicted. Take the case of Podcasts which peaked in 2005. A promising prospect then, now 15 years later it is still in its nascent stages, yet to be adopted by the masses. Some techs enter too early and keep receding into the future like for example, Quantum Computing was featured in the 2000s, is still a decade away now. Others techs, like Call centres, IT Outsourcing and ERP entered in 2003-04 directly in the Trough of Disillusionment, too late to have any impact on the users or businesses.

2019

Conclusion

While one can appreciate Gartner for taking a realistic approach with their plot by factoring in in human nature, more precisely our reaction to innovations, we should also realize that it is an extremely difficult task to quantify, measure and compare it. Questions on the interaction of tech with a societies’ socio-political scenarios, its readiness to embrace not only economic but social and technical changes introduced by the tech, are difficult if not impossible to plot on a graph.

Looking at Gartner’s data sources we can understand this better. Gartner says that in order to plot these expectations it surveys the managers of companies and vendors developing these techs - due to their enormous stake at succeeding, they have the tendency to overstate the potential of their tech. A survey of present or potential users would be more accurate, but at an earlier stage, reliable identification of who they might be is difficult.

It is important to understand what the Hype-cycle stands for and is trying to tell us year on year. Treating it like an oracle and having unrealistic expectations will only lead to wastage of resources and efforts and their diversion from where they are more urgently needed. We need to learn to differentiate between ‘hype’ and actual usage. The hype is created - by media, news articles, online forums and so on, one might Google search the term, read about it but never use it.

Even though Gartner’s Hype-cycle has weak theoretical foundations and uses unreliable data to produce blurry predictions we can appreciate it for the great insight it provides into the history of various techs. Looking back one could see the wastage of resources that went behind a failing tech but one could also take a moment and appreciate the progress we’ve made. Our labour has brought us to a wonderland of techs - of computers that can ‘almost’ read our minds, of experiences that immerse most of our senses, of cars that can drive themselves and data that promises to save us from our impending doom. While the plot talks about ‘hype’ and predicts the future we can say that the difference between hype and its conversion into real usage is opportunity and courage.