Skip to main content

2018: Innovation Hits the Enterprise

By January 25, 2018Article

Judith Hurwitz knows a technological innovation is only as good as the business value it creates. So when the woman who literally wrote the book on cloud computing (and many other enterprise technologies, including  Cognitive Computing and Big Data Analytics and Smart or Lucky? How Technology Leaders Turn Chance into Success ) issues her predictions for the year ahead, wise vendors around the world take note.

As president and CEO of Hurwitz & Associates, Hurwitz and her team of strategic research experts have spent years working closely with leading enterprise customers and translating their needs into insight for the tech industry. So what does the market predict for 2018?

“If 2017 was the year of massive hype around artificial intelligence, machine learning, security, and the cloud – 2018 will be the year where enterprises will begin to absorb these technologies,” says Hurwitz, who believes smart business leaders are now taking complex emerging technologies, putting them together, and building actionable strategies for success. (Click here to read more predictions for the software industry in 2018.)

SandHill.com spoke to Hurwitz about her five forecasts for enterprise innovation in 2018 and the trends supporting each one.

1.  The private cloud is a real thing and customers are beginning to transform their data centers into private cloud services. This does not mean that the data center will disappear. Instead, the data center will focus on particular high-end applications and workloads where there are a lot of dependencies. The private cloud is becoming a software-defined infrastructure that can leverage existing computing resources in innovative ways. The private cloud is a way of architecting the data center so that some data and workloads can operate within the firewall. The private cloud architecture will be standards-based and will leverage services such as microservices and containers and well-defined APIs.

“I think there are two issues at play driving the use of private clouds. The first is that organizations are beginning to see the downsides of public clouds in terms of cost and manageability. Too many companies had the mistaken impression that if they used a public cloud, they would be off the hook in terms of responsibility for compliance, security, and other risks. They assumed these issues would be the vendor’s problem and that the vendor would take care of them. Unfortunately, this is not true. When an issue arises with third-party services, the enterprise has to take responsibility. Running a private cloud eliminates this illusion of protection.

Second, more companies are now looking to cloud services as a profit center. And if this is the case, they must manage their cloud diligently to ensure profitability. With a software-defined infrastructure, enterprise customers can build those cloud services into a private cloud that really does act like a public cloud in terms of capabilities and automation. So I think we’re going to find a bifurcation of public and private cloud services where with something like email, there is no benefit to building an in-house system over using Gmail. But for specialized services, customers want to keep them behind the firewall, with specialized service-level agreements. When a private cloud service goes down, the enterprise is held accountable. A customer might lose $5 million if the service is down for 15 minutes. No public cloud provider will assume that risk, therefore, the answer is a private cloud.”  

2. A new generation of hardware is being architected to support emerging Artificial Intelligence workloads. General-purpose hardware will still have its place, but the innovation will be focused on developing hardware that is designed to the most complex and compute-intensive workloads. 

“If a company has a good level of maturity, then they are looking at very sophisticated machine learning and similar technologies. They need a lot of power so they are looking at GPUs and other ways to make things faster. This isn’t going to be the case for every enterprise but for those in need of speed, buyers can now find it embedded into a solution. I’m seeing vendors of GPU systems partnering with companies like Dell, IBM, and others to provide this much-needed speed in their solutions.”

3. Cloud infrastructure services are increasingly maturing so that the battle among cloud providers will be around the ability to provide hybrid cloud management, data management, security, compliance, and service level agreements. Up until now cloud vendors have differentiated themselves based on price and speed. In order to survive, cloud suppliers will have to move to a higher level of service and manageability based on enterprise requirements. 

“As cloud computing strategies become more mature, we are seeing people grasp the reality that they are not managing one service or one specific environment. Rather many enterprises have 15 different SaaS services, a bunch of cloud infrastructure services, and even more things to manage in the data center. The users? They just think of this as computing. When management looks across all these demands, multi-cloud management becomes the solution – and with it comes a heightened need for better service and manageability.”

4. Augmented intelligence will become the most important transformation in the movement towards artificial intelligence. This new generation of systems will be truly a collaboration between machines and humans. Augmented intelligence emulates and extends human intelligence with data to find the hidden meaning in the data. Augmented intelligence opens the potential to understand hidden data in context with solving business problems.

“In the early days of enterprise AI, I heard customers thinking all they had to do was put in a bot or create a machine learning algorithm and let it go off and have fun. Now everyone knows there are some manual processes and simpler models that are fine to run on their own but the real big value comes from the collaboration of the individual subject matter expert and machine learning and cognitive technology. Our human brains can decipher a lot but if there is too much information in disparate forms, we don’t have the time or ability to figure out the answer to a certain question. Humans can look at a machine learning solution and think, “That doesn’t sound right, let me add a piece of data.” This augmented scenario better resemble the nuances of real life. We are definitely starting this strategy at work in the enterprise with two types of partners: emerging companies that have a base technology that isn’t fully baked yet, and also large vendors providing really amazing solutions, rather than just tools.”

5. Enterprise innovation adoption is happening faster than ever. 

“I do believe enterprise adoption of new technologies is happening more rapidly in the cloud computing era. Take an early innovation like refrigeration. The first practical use of refrigeration – not just an icebox filled with ice – came in the 1750s but it took 200 years for it to become price competitive and practical enough to become ubiquitous. Fax technology also appeared about 100 years before it became mainstream. And even the concept of service orientation has been around for 50-60 years.

The difference today is the existing cloud infrastructure. Vendors can invent solutions by pulling essential elements together from the cloud, tap into its critical mass of users, arrive at a competitive price point, and deliver a market-ready product in record time. In prior waves of enterprise computing, a vendor would have had to invest in the infrastructure, marketing and training before they could determine if the innovation was good enough to take off.”

 

Clare Christopher is editor of SandHill.com

Copy link
Powered by Social Snap