The Cloud is a key component in the five pillars that allow companies to scale in unprecedented ways. Companies like Amazon scale their services at remarkable speed. Did you know it took Amazon.comm only 120 days to become the largest online music retailer. It took Apple only 9 months to hit 1 billion downloads from its Apps Store. Cloud represents a tipping point in corporate capability – but when is it appropriate and when not?
Neither Apple nor Amazon’s case, above, were true Cloud applications. Both Apple and Amazon had to make the investments in server capacity that allowed these remarkable landmarks to be reached. Imagine if the Cloud had been available – the investment cost would have been a lot lower and services could have been ramped up even more quickly.
Without the Cloud companies like Apple and Amazon had to invest in enough server capacity and the associated software to meet peaks in demand. In designing a system like the Apps Store or in Amazon’s case music retailing, you need to plan in the bursts of interest that are typical of web behaviour.
We are in a very “bursty” economy where people flock to certain sites on the back of news references, launch announcements, or say, a new game that goes into a store. Companies that don’t invest for the peak are going to disappoint vocal customers, which could have repercussions for sales.
There’s a great series of articles over on Forbes right now written by Darcy Travlos, an investor, on how the Cloud ecosystem is shaping up. She divides it into four segments:
The first category is the suppliers, such as semis, that contribute to the second category, the large-scale server/storage manufacturers. On top of the servers sit the third category of managed services and the fourth category of platforms/applications.
Rather than repeat what Darcy is saying, we recommend you take a read. Her points are interesting because she’s describing how difficult it is to design Cloud services in a way that secures sustainable profits – except in the fourth of her categories, applications. In her account Cloud looks pretty fragile and that reflects some of the conversations we’ve had with CIOs. We doubt Cloud will quickly take over from internal infrastructure, but it is being used to pioneer new services (see this example at Dell). And that’s the point of an elastic enterprise.