A couple of years ago, I proposed the 10 Laws of Cloudonomics, an attempt to address the economics of cloud computing. Beyond the cloud, though, neo-classical economics has been augmented in recent years by behavioral economics, where psychological anomalies modulate pure rationality. Indeed, purchasing and adoption of clouds are no less subject to human behavior, hence I offer the 10 Laws of Behavioral Cloudonomics.
Under expected utility and rational choice theories, Homo sapiens is assumed to weigh probability-adjusted costs and benefits in a cold and calculating fashion and select the option with the highest expected payoff. However, Nobel laureate Daniel Kahneman and others have demonstrated that the real world is different, that the Earth is not Vulcan. Although not their terms, one might describe human behavior as lazy, hazy and crazy. Lazy, as in minimizing physical, cognitive, emotional and real dollar costs; hazy, as in using heuristics or rules of thumb rather than precise calculations; and crazy, or, as MIT’s Dan Ariely says in his book of the same name, “Predictably Irrational.”
With that in mind, I’d like to review some cognitive biases most relevant to cloud computing. These should be of interest to cloud service providers, as they can be intangible barriers to cloud computing acceptance, and to customers, who can recognize these behaviors and moderate their impact. The interests of both will be discussed in great detail later this month at the GigaOM Network’s annual cloud computing conference, Structure, where I will be hosting a panel titled “Cloudonomics: The Value of the Cloud.”
- Risk and Loss Aversion — There are emotional and perceptual asymmetries between losses and gains. A loss is more painful than a commensurate gain is pleasurable: losing a $10 bill can be more irritating than finding one is joyful. Certainly CIOs must exercise due diligence regarding proposed cloud initiatives, but should also be aware that these asymmetries may cause some concerns to be overweighted relative to benefits such as total cost reduction and enhanced agility.
- Flat-Rate Bias — One effect of loss aversion is that consumers often prefer flat-rate plans (PDF) even when pay-per-use would cost less. With flat rates or up-front capital expenditures, the charges are never in doubt. The cloud’s pay-per-use pricing for on-demand resources typically reduces total cost while enhancing scalability, but the pleasure of a dollar saved may not outweigh the fear of loss from auto-scaling gone awry. Flat-rate buckets, monitoring and reporting, and auto-scaling policy management with maximum capacity limits can help.
- Need for Control and Autonomy — Author David Rock reports that people have a deep-seated need for control over their environment (PDF), or they may exhibit “learned helplessness” and shortened life spans (PDF). Owning a data center may provide a perception of greater control over assets, making dashboards, portals, transparent policies and fine-grained management essential for cloud providers. Autonomy —“I can do it myself!”— is an important driver for cloud computing. Developers can autonomously procure infrastructure resources, and platform services enable the democratization of IT.
- Fear of Change — Rock also observes that people are often uncomfortable with uncertainty and therefore fear change. The cloud offers not just new technologies but new business and organizational models. Consequently, overcoming the inertia of the traditional owned asset model may require free trials, unalterable and explicit privacy policies, and/or multiyear price guarantees.
- The Endowment Effect — People value goods that they already own more than they would pay to acquire them. Ariely showed that for the same hard-to-acquire Duke basketball tickets, students were willing to pay up to about $170, but weren’t willing to sell them for less than $2,400. Add in the choice-supportive bias, which rationalizes selected options and discounts unselected ones, and a stubborn fondness for existing IT technology and organization assets can be understood.
- The Status Quo Bias and Escalation of Commitment — Moreover, we tend to prefer things the way they’ve always been, and invest additional amounts in past strategies which we have pursued. Again, this can lead to inertia slowing the adoption of new approaches.
- Hyperbolic Discounts and Instant Gratification — People tend to discount future risks and benefits hyperbolically, that is, more steeply than accounting texts teach: a chocolate chip cookie is much more valuable now than in an hour. This is good for the cloud, which promises instant gratification via on-demand services. Moreover, the “pain” of payment is deferred, thus discounted.
- The Zero-Price Effect -– Ariely argues that zero is special. People would rather receive a free $10 gift certificate (a $10 gain) than pay $7 for a $20 one ($13 gain). This also benefits the cloud, since up-front costs are typically eliminated.
- Need for Status — Rock points out that humans and other social primates have exquisite, fine-grained status detectors. For cloud adoption, the status associated with managing a large IT organization with a substantial asset base needs to be replaced by the status accruing to being perceived as an innovator through the use of cloud services.
- Paradox of Choice — Web commerce has enabled a shift from a few big hits to a long tail of boundless choice. However, too many choices may cause paralysis by analysis, reducing sales. Fewer bundles, rather than unlimited configurability, may best boost service provider revenues.
Behavioral Cloudonomics — the intersection of psychology, economics, and the cloud — can help us to make the most of the appeal of the cloud and understand and address barriers to adoption.