In a recent response to a posting on this blog, Whit Andrews points out that the shape of the hype cycle is a wave, not a cycle, in that it doesn’t have a loop backward like a true cycle would. This is very true – it’s not typically the technologies themselves that loop around and around. They progress inexorably toward maturity (or obsolescence), albeit at a slower pace that we want or expect. The cycle itself relates to the behavior of people. As individuals, as marketplaces, and industries, we go round and round a cycle of enthusiasm and disillusionment with each new technology or trend.
However, as Whit also accurately points out, there are sometimes technologies that do seem to cycle between the Peak and the Trough. In past research, we have referred to these as Phoenix technologies, defined as follows:
Phoenix technologies cycle through enthusiasm and disillusionment continually, with an evolved interpretation of the concept on each iteration. Agents are a prime example of a Phoenix technology. Agent technology is embedded in certain product classes (for example, network management and comparison shopping), but there are many other capabilities and interpretations of agent functionality that re-emerge year after year. (From “Escaping the Hype Cycle: Dead or Alive?”, Gartner research note from 2002)
One of the graphical challenges we have examined is how best to represent these Phoenixes. We have shown them informally as a loop back from the Trough to the Peak, but with a horizontal axis representing “time”, we don’t want our Phoenixes to be time-travelers as well. In fact they are more accurately represented as a recurring sequence of Peaks and Troughs. We talk in the book about the “double peak” hype cycle, where a significant performance improvement, product repackaging (eg suites) or similarly hyped activity launches the innovation up the slope in a dramatic fashion (with an extra mini-hype cycle), as compared to the usual more subtle and easily missed maturation path. (This also relates to the idea of hype cycle as fractal, which we’ll explore another time). In other cases there are multiple Peaks that each end in a significant Trough – Whit cites interactive video, another example is virtual worlds (VRML around 2007, Second Life in 2007), as well as the intelligent agents we used in the original definition.
This pattern brings to mind the feeling of trying to start a gas-powered lawnmower. A sharp tug on the starting cord triggers a surge of energy which rapidly drops away – either to the satisfying chug of a functioning engine, or the frustrating sound of silence. That’s the perennial question with a Phoenix technology – will this round of hype launch the start of real maturity, or is it another false start? We’ll keep analyzing examples of each to try and answer that question a little better each time.