HypeCycle

I have found the following model useful for describing the productivity of technology.

My source for the model is the Gartner Group and I share it using the exact headings that they use. Gartner also use symbols for the techologies that show its productivity plateau's timeframe, such as less that 2 years, 3-5 years, 5 - 10 years, greater than 10 years, and obsolete before plateau.

If I map technologies, such as object oriented programming, java, c# and so on, they follow the curve.

I would love to hear where you would put technologies on this model. If there is sufficient data, I'll be happy to draw a chart and post it of the AYE participants view of today's technologies.

I supsect that the Gartner Group borrowed this model for someone. If you know their source, please share it.

SteveSmith 2003.09.05

They've left out the interesting part - when the "trigger" takes place. There are lots of would-be triggers that don't result in this cycle - fuel cells, for example, have been known for well over 100 years. Every few years, we hear that NOW is the time when they're going to take off, but so far it hasn't happened. Yet fuel cells are growing in use - not, I think, by this curve.

I'd say this curve is followed by those technologies that follow this curve, and not by those that don't - which is most of them. Heck, 100 years ago, radium was going to be the home-heating source of the future. So, Steve, what use is this curve? -- JerryWeinberg 2003.09.05

I hypothesize two processes going on in technology adoption. Approximately:

A social process and
A delivery process

The former is about people, their expectations, and what they imagine then learn about what a technology can do. The latter is about making that technology available.

Using a technology widely usually (always?) involves other technologies. If all the other technologies are in place when the "new" technology is discovered, I suspect the social discovery and learning process dominates, producing a curve like the above. This assumes the new technology does something of use, and probably some other things as well. If all the other technologies for delivery and use are not in place when the "new" technology is discovered, I suspect that a curve like the above does not happen.

Taking fuel cells as an example, they have been "known" for a long, long time. Fuel cells might provide a real benefit in many applications if they were available. Any portable, electrically powered device is a candidate for a fuel cell application. Fuel cells are attractive because their energy density can be higher than batteries, for example. Broad adoption of fuel cells requires several supporting technologies that are non-trivial. At least these:

Fuel-cell catalysts are easily poisoned, so require a solution to fuels with impurities.
Fuel-cell fuels have distribution and safety challenges.

I suspect that many technologies that follow the bouncy curve above aren't so much "discovered" but "rediscovered", with more of the preconditions for their adoption in place. See: _As We May Think_ by Vannevar Bush as an example of a technology that was discovered, then waited around for a long time before it was possible to make it kind of real.

--- JimBullock, 2003.09.05 (Edits.)

Jerry, The hype cycle explores both the development and marketing side of a technology. It usefulness is its ability to reveal where a technology is in the cycle.

I often find it difficult to effectively set the expectations of the people in an organization whose opinions count the most about new technology. The hype cycle is one of the models in my tool bag to support an argument about whether a technology is ready for deployment.

For instance, utility computing is a "new" technology touted by IBM. The concept is that a customer can manage and bill out their IT infrastructure just like a power company manages and bills electricity. Yeah, I know: It's an old computing concept. But it's new to a whole new generation.

The utility model doesn't work at this point in time. It might work many years from now after the instrumentation of today's open systems finally catches up with the drive for functionality.

The hype cycle helps me put an argument like the above into the perspective of the development and marketing of other technologies. For instance, I can compare utility computing with JAVA or another technology that has familiarity for that person.

SteveSmith 2003.09.07

What an interesting general purpose graph.

My first response was ... Oh look, the output of a system tuned for quarter wave decay for a step change on the input.

Then I read the labels on the x-axis, and they reminded me of the parts of the Jeannie Duck's Change Curve (Stagnation, Preparation, Implementation, Determination, and Fruition).

This of course reminded me of the Satir Change Model (late status quo, foreign element, chaos, transforming idea, integration and practice, new status quo).

What the Gartner graph leaves out is all the messy stuff (wildly fluctuating values). It also applies only to those technologies that "make it". Those that don't continue downard through the trough of despair, past the x-axis, only to be seen again in the list of SilverBulletsThatMisFired. DonGray 2003.09.16

Never mind fuel cells, the graph works perfectly for skateboards and Razor scooters!

RobWyatt 2003.9.16

Don Wrote: ... Oh look, the output of a system tuned for quarter wave decay for a step change on the input.

Yup. It's a flaw in the education system, actually. There's all that time spent doing differential equations in math, or getting drilled in "dynamics" in physics, if those are things you studied. That isn't such a useful set of skills, actually. Most of us won't often need to tune a bridge spar, so the span doesn't flap it's way into a bay like the Tacoma Narrows bridge did.

What is useful, I think, is being able to see the data and recognize that there's a system there. Once you go looking for a system, you can learn some interesting things.

At one job, I went and had lunch in the company cafeteria - rare for one of the technology types. I sat at a table full of folks from one business department - unheard of. Now, the idea was to learn some stuff, which I did. The other idea was to get some idea how the company worked. It turns out that within 45 minutes of lunch one of the IT old-guard happened by, just to say "hi" and somehow managed to work in to the conversation my choice of lunch companions. The new IT management didn't become aware of my little social event until I told them, then asked "Why?" and looked scared.

Such interesting information. With one little experiment, I know there are at least two communication systems, and that socializing with our customers is a perturbing input. I have some idea of the time domain behavior of one of those systems. I have actually modeled the dynamics of some social systems in companies. You can often do this well enough to win a small bet.

Why, in a dynamic, active world, do we so often seem stuck on static models, and thinking like we are the only actor making choices?

-- JimBullock, 2003.09.17

Why, in a dynamic, active world, do we so often seem stuck on static models, and thinking like we are the only actor making choices?

<mock horror>You mean someone else is making choices too?</mock horror>
This list isn't all inclusive, but for starters:

- It's easy, and I can blame someone else (because they did something too) if things don't turn out "right".
- We're taught simple linear cause/effect thinking in school.
- We're not aware there are other options.
- It's always been this way.
- We tried that once and it didnt' work.
- ???

DonGray 2003.09.18

Updated: Thursday, September 18, 2003