With infrastructure-as-a-service enjoying a price war of sorts thanks to Amazon’s continuous price improvements, there is talk that Moore’s Law has given way to Bezos’s Law. Call it what you want but we have have had a deeper disease in our industry for a while now as I wrote in a complete chapter in The New Polymath five years ago
“The reality is that while the technology industry proudly talks about Moore ’ s Law, the majority of the industry has not been subject to it for decades. Most technology investment these days is not in hardware but in software and services — costs that are more closely tied to labor economics, not trends in transistor design and scale manufacturing. In telcos, pricing
has often been driven by regulatory influence and, in many countries, monopolistic positioning.”
“At least Intel benefited us all by “ killing its own children ” and unleashing generation after generation of chips. Most other tech companies pay lip service to new products — managers of their incumbent products undermine the new ones. The industry has validated over and over the “ Innovator’s Dilemma ” that Clayton Christensen made famous. He pointed out the phenomenon where many large, successful companies are turned off by the initially poor economics of disruptive technologies (even those from their own labs) and so they fight them. They pamper their older children, and allow them to kill their younger siblings.”
“The technology industry is replete with examples:
- IBM announced “ On - Demand ” services in 2001 in response to growing customer clamor that IT had become too fixed an expense. But it isn ’ t IBM but cloud computing vendors, which we discuss in Chapter 18 , that are delivering small, bite - size, variable units of technology procurement and provisioning.
- Major telcos such as Verizon, AT & T, and Sprint sued Vonage, claiming prior art around VoIP dating back to the 1990s. But in spite of claiming prior art, they were not very interested in, and then not very successful at, offering VoIP to their customers. In fact, many have fought providers
like Skype from offering VoIP on their mobile networks.
- Accenture (in a predecessor entity, Arthur Andersen) pioneered a software development center in the Philippines in the mid - 1980s. It did not scale that low - cost pool for years and did so years later only in response to the global delivery model that firms from India and younger U.S. companies, such as Cognizant (see Chapter 6 ), showcased to Western customers.
- SAP and Oracle are both years late in offering Software - as - a - Service (SaaS). In fact, it could be said that Larry Ellison, CEO of Oracle, smartly invested in two SaaS start - ups — salesforce.com and NetSuite — because he realized starting them as projects within Oracle would not generate much success.”
Rather than emulating Bezos I see many vendors try to explain why the AWS model is flawed. No retail customer wants to be on (competitor) Amazon’s infrastructure. No auditor will sign off on Amazon’s inadequate data location transparency. Amazon’s data centers do not use clean fuel. And on and on.
So, improve on Amazon’s model, and even charge more for that. But not 10X, and like Amazon be prepared to pass along improvements several times a year.
Let’s not reminisce about Moore’s Law or idly talk about Bezos' Law. Let’s all strive to deliver to them.