One of the people I've reference numerous times when talking about cloud computing and its inevitable adoption is Clayton Christensen. Not because he is a cloud expert, but because he has written about similar disruptive shifts in the steel industry (by mini-mills) and in the hard drive product space, in his book: The Innovator's Dilemma.
CIO magazine had an interview with him recently that reinforces the assumptions and comments I've made before. He states:
"Cloud computing-any computing over the Internet-just isn't as good as enterprise computing. It's not as secure, not as fast, not as reliable as your internal network. But like all disruptions, it's getting bigger and better. As it does, it pulls applications, one by one, out of the corporate network into its world. CIOs have to manage that."
Cloud is likely to be in for a rough year (as it gets pushed into more an more areas where it can fail) but that's what it takes to become strong enough to meet organizations needs
In this month's Wired magazine there is an article titled: The Good Enough Revolution that discusses how new markets are taken over by products that are "good enough" and undercut technically better approaches. The examples discussed are the Flip camera undermining the camcorder market, or MP3 taking on the recording industry and even the Predator drone replacing manned aircraft.
These products grab market share because they make our selection standards shift and what was important before may no longer be a deciding factor in making the decision. Clayton Christensen in his book The Innovator's Dilemma talked about this phenomenon (Joe Hill did a blog entry on it a while back). Jerry Pournelle has also been talking about selecting "good enough" products in his Chaos Manor blog for almost as long as I can remember.
A reason this is important to IT organizations today is the whole concept of Cloud techniques, including SaaS. Granted there are many scenarios where an in-house or custom developed solution would be better, but the decision criteria may have shifted so those "better" characteristics are irrelevant. We need to focus on "What does the business want vs. what do they need?" and how do the characteristics of the product or service address them. For example: A cloud provider may not be able to guarantee anything close to 100% uptime from end-to-end for a transaction, although they may be willing to do this within their four walls. Is this guarantee good enough? If not are there alternatives?
I was in a Azure meeting in Dallas the other day and when I brought this issue up, a person in the audience said "No one can guarantee end-to-end!". I totally disagree. We've done it for years, although not in a cloud context.
I'd say organizations need to get used to the "good enough" thought process and be flexible about their critical characteristics. The way they have always measured success may no longer be what is most important to the business. This concept has been around for a very long time and is not a fad but the future.
In his two highly-acclaimed books, The Innovator's Dilemma and The Innovator's Solution, Clayton Christensen (and co-author Michael Raynor) identifies three strategies to create new-business growth: the sustaining strategy, the low-end disruption, and the new market disruption. The following diagram (based on Figure 2-3 in The Innovator's Solution) shows these strategies in relationship to performance, customer, and business model.
The sustaining strategy offers better products to an established market. It tries to maintain current competitive advantages by exploiting existing processes and cost structure to improve performance on attributes valued by the most demanding customers. These customers are also the most profitable and are willing to pay more for these improvements. This strategy is often followed by existing market leaders to maximize the value of their franchise.
The low-end disruption strategy addresses over-served customers with a lower cost business model. It attacks the market leaders from below by offering performance that is good enough, to low-end customers who are currently over-served, at prices substantially lower than the mainstream leaders. This business model requires new operating or financial efficiencies to generate acceptable returns on these low-margin customers. Market leaders are often blind to this strategy, making it attractive to challengers.
The new-market disruption strategy targets non-consuming customers or occasions by offering improved performance on new attributes (often simplicity and convenience). This business model requires margins on lower price per unit sold and at smaller production volumes, at least initially. The economics make it difficult for existing market leaders to follow this strategy.
Both of Christensen's books provide lots of examples to illustrate the general patterns and principles he identifies.
Like Drucker's seven sources model, these growth strategies focus on the innovation-commercialization side of the Rogers' innovation diffusion processes. Christensen discusses customers and their reasons to buy in order to determine appropriate business strategies for the seller of innovative solutions. He describes optimum architectures for innovators to use in the different stages of market development. In taking the innovator's perspective, Christensen is true to his books' titles.