A number of analyst firms and pundits have declared a list of the megatrends. For examples you can see my blogs:
I have my own list of what I call “disruptive” technologies and trends. This comes from talking to people, observing what companies are doing and what people are saying as well as my own intuition of what we have seen in the past and therefore what seems likely to happen.
I divide my list into 3 categories: 1) Real world disruption 2) Enterprise Disruption and 3) Tech Disruption. For real world disruption we are talking about things that will change the way we do things day to day whether you are a plumber, retired person or high tech worker. For Enterprise disruption I am talking about the trends that will change the way business operates potentially causing some businesses to fail or leap to the front of the pack. For Tech disruption I am talking about technologies that are going to change the way we technologists build, deploy, use technology and lead to the growth of some tech companies.
I have 7 real world megatrends, 4 enterprise megatrends and 5 tech megatrends. Unlike other analysts who take a short term and very qualitative approach to prediction I am going out on a limb to say how big I think these trends are in terms of dollar impact on the economy and how fast the trend will move as well as how far I estimate the trend is along its curve. We start with Tech Megatrends today. Enterprise and Real world megatrends shortly.
Top Tech Megatrends:
PaaS and DevOps. Gartner, Forrestor, IDC all expect PaaS to grow by leaps and bounds in the next 5 years. I also am a big enthusiast. It seems that unlike the last 30 years which have been characterized by a wide plethora of ALM tools and processes adopted by companies the IaaS, open source and other trends are right now driving a consolidation of tools and methods to do PaaS which will make adoption of PaaS and devOps much faster and more complete than previous cycles. The cloud has left all previous ALM and devops tools in the dust so that the few tools created in the last 10 years are dominating the space. Git, Maven, Jenkins, Chef, Puppet, Redmine, Jira, Java or Java based languages, container technologies of various types have become dominant in any modern enterprise development shop these days and there is little resistance I see in people moving to them. This consolidation will lead to wider and faster adoption in my opinion of PaaS.
PaaS can be broken down into various flavors and subtypes. iPaaS, aPaaS, BPMaaS. There are also what I call Application PaaSs, Toy PaaSs, Generic PaaS‘s and Ecosystem PaaS’s. Each of these represents different types of platform technologies used to build different types of applications. I have a blog describing some of these variants on PaaS:
I believe that this is a 100 billion dollar market in 20 years. Gartner predicts 14 billion by 2017. I think PaaS is ultimately a $100B market and we are at 3% of that today and gaining 3% a year so I think it will take a long time to attain it’s ultimate potential but that in 15 years we could see the majority of all applications running and being built in PaaS environments. What will slow down adoption is the existing legacy infrastructure and the need to train many people in PaaS technology.
Today PaaS comes in a number of delivery flavors. Public PaaS’s meaning that a service provider provides not only the technology but the services behind the technology. Private PaaS technology solutions you have to run in a cloud infrastructure that you operate that integrate best practices. Usually on-premise but could be run in a public or private IaaS vendor. I think that over time the complexity of operating a PaaS will cause most companies to switch to the former model but today about half of PaaS is being sold as private PaaS. The trend will eventually move to public PaaS’s.
2) IaaS Commoditization and the end of Enterprise Premise Server Farms
I don’t think all companies will give up owning server farms nor do I think IaaS is the cat’s meow and is painless cheaper and perfect. The reason I am convinced that enterprises will eventually give up owning so much of their own technology is incompetence. I believe most organizations are simply not well equipped to own and manage this equipment and the technology that goes with it. That is not a negative. They shouldn’t. If you sell shoes, concentrate on shoes. The best shoes should win not the shoe company with the best IT.
This movement away from enterprise farms will be slowed by many factors. Enterprise management and workers as well as vendors may push back on too many companies moving too much to the cloud too fast. Government regulation may inhibit some movement.
The lack of standardization in cloud IaaS vendors makes it difficult for some companies to decide what is the best course and will lead to slower adoption. For instance, Google has Google compute cloud and goodle app cloud. Great ideas but nothing like what Amazon does with its AWS service and the various services and tools it offers. SoftLayer has a different set and a number of companies have chosen OpenStack. If standardization wins then OpenStack will end up winning but we see frequently that usually one proprietary non-standard vendor usually keeps a large market share. I am not sure if that is Amazon or Google. They are both very strong.
Other things will slow adoption of IaaS. A large part of enterprise server farms can’t be easily transitioned due to lack of compatible hardware in IaaS vendors as well as prohibitive cost/benefit for doing this. It will therefore take at least 30 years for this transition to happen. However, the size of this is so great (1 trillion or more dollars worldwide in enterprise technology infrastructure) that as slow as this movement is it will be a huge economic and transformative thing. I see that almost all new development will not be implemented on enterprise owned hardware within 5-10 years.
Contrast this with the fact that network effects of the cloud will create a massive new demand for computing resources and for applications so the size of this market is not just “the current IT infrastructure” but the vastly increased infrastructure needed to support bigdata and all the new devices and data about people and things and all the unanticipated new applications of all this data and connectivity. This trend is huge and will stay huge for a long time. I anticipate this could be a multi-trillion dollar industry in 20 years.
Enterprises are rebuilding their infrastructure along API Centric architectures. Some are moving faster than others but the trend is huge and pervasive. There is hardly a company I talk to that isn’t doing this or something like this. Examples like Google, Netflix, Twitter, Salesforce and others are showing that you can make serious money from APIs ($billions). More important is that we are seeing for the first time true reuse from APIs and services meaning higher efficiency and faster time to market. All these things are extremely compelling to any technologist and can be justified to the business on revenue generation or cost savings basis. Lastly API’s are the tools for building mobile applications and it is clear that mobile is a huge megatrend that will change the face of applications away from the desktop more and more. The API refactoring trend also supports social because it is easy to insert the data gathering and analysis needed when using APIs that allow an organization to move to a data driven enterprise. (Another megatrend I will talk about.)
Part of Enterprise Refactoring and the emergence of the services API economy is the Store. The Enterprise Store will become the defacto way we manage technology in enterprises. We have found the Store paradigm to be disruptive with Mobile and with APIs. I expect we will see this happen with all enterprise Apps and Services. The Store provides a place to see and manage all enterprise IT services, data, applications, infrastructure. Check out Store as an early example of this.
The API refactoring / API Economy is a $200 billion market in 20 years. We are 5% of the way into this market today and growing very fast at 3% a year.
4) Open Source
This megatrend is farther along than some of the others but I believe has legs. I believe that this year and last year we saw the widespread acceptance of open source in enterprises at the highest levels. It is true that open source has been in enterprises for more than a decade but now we are seeing organizations as conservative and rich as banks and financial companies, health companies, state governments making a commitment to open source as a framework for their enterprise architecture. I see the cracking of the big enterprise players. Salespeople I know from my past days of enterprise sales are seeing customers delay making these proprietary software license purchase decisions. I hear people in companies saying “I hate x” (replace x with your least favorite proprietary enterprise sales vendor) more and more and looking for some way to avoid the millions and millions of these license costs. Proprietary vendors are engaging in “openwashing” which is a way to say they are open source too even when they aren’t.
This trend is not all cost based. If people were only concerned about cost the megatrend would have no legs. The fact is that in many cases now the best software, the industry standards are based on open source standards, on open source solutions. From operating systems, middleware vendors and database vendors and every kind of enterprise software is being led by open source companies that are pushing the envelope faster than their proprietary competitors.
The reason for faster innovation in many cases at open source companies and projects is because a combination of cross fertilization by companies and inventors able to leverage open source and contribute back improvements faster but also because the cost of sales at open source companies is much lower. A typical enterprise sales software company will spend at least 30% or more of revenue on sales. Open Source companies spend less in many cases (not all).
Overall I am guessing that open source is a $20 billion business and we are 10% into this and will see a rapid increase at 5% a year or 10-15 years to get to $20 billion dollar market.
5) The iPaaS revolution
Integration is a massive business. Most projects in companies have a large fraction which is “integration” combining one system with another.
When we have a service economy, when people have put the services into the cloud we see happening today by the 10s of thousands and these things grow as I expect in tech megatrend 3 we have a completely new way to integrate. Applications become combinations of services and integration is combinations of services. We have new applications which are nothing more than integration and new applications which depend on integration of things we never had before.
There is something called the network effect. The network effect, sometimes called viral is something like Moores law. It says that as things and services and people become connected that new applications and new services become available based on this ability to network all these things together to form ever more valuable services. It is very hard to estimate the size of this but I anticipate that the cloud is going to be mostly about network effects in 10-20 years if not sooner. That means the time to move is now!
The iPaaS revolution is about this network effect and the combination of services to create new services on top of existing services. Uber for instance is an example of a network effect. The ability of smartphones has enabled the cab driver and the passenger to have a much more direct and simpler way to connect and do business. The US economy is managed by the federal government based on statistics that are gathered from surveys and numerous delayed data sources. Imagine if the government actually had accurate data on a minute by minute basis. Could they manage the US economy better? Could they make it easier to identify and help poor people or to otherwise provide services that make the economy function better? What is that worth?
I suggest that the iPaaS revolution and the network effects are worth $1Trillion and that we are only 1% into this market growing at 3% a year. Is that optimistic enough? Am I smoking something or am I being pessimistic?
Things NOT on my Tech “disruptive” technology list may not be on this list because they are on my other lists. For instance, mobile and bigdata are part of Enterprise Disruption, so wait for those blogs before criticizing what’s not on this list. However, feel free to comment on what you think is not included or naivete I may have on what technologies are disruptive, the size of the markets.
The next blog on this topic will be the Enterprise Megatrends.