How, why, and when AI & machine learning will transform organizations

Written by: Nick Patience, 451 Research Founder and Research Vice President – Software

AI – and machine learning in particular – is set to be the most transformative technology existing over the next decade, but it’s only just getting started. 

We have been covering AI and machine learning at 451 Research since we started the company in 2000. Back then, we were primarily focused on text analytics around use cases in government intelligence scenarios and, in the future, in the legal industry. We branched out from there as additional information types – such as audio speech, images and video – became viable data types from which machine learning can extract insights. To enhance our coverage of this vast and evolving space, we launched our inaugural survey on the topic: Voice of the Enterprise: AI & Machine Learning survey – Adoption, Drivers and Stakeholders 2018

Our new survey brings together use cases, business benefits, barriers to adoption and information about how who is influencing and ultimately deciding when, how and why to adopt machine learning. Given it is an omni-purpose technology, it is not surprising that use cases for machine learning are spread across all industry verticals and all layers of organizations. Practically everybody surveyed – 97% – believe AI will have an impact on society and 75% of this group think it will have a moderate to significant impact within the next two years. The question is how much impact and where will it be felt?

PR Graphic VotE AI 2018 2 1
Industry Optimism
 
Respondents are broadly optimistic, with 69% of them believing AI will have a mostly or somewhat positive impact on society. The much-vaunted threat of job losses doesn’t seem to concern people that much. While 62% think AI will have some type of impact on their organization, only 7% of them think it will be negative, suggesting respondents aren’t worried about AI replacing their jobs in the near term. 

VotE AI 2018 blog image 2
As the domain gets closer to the individual situation, we find people becoming more ambivalent about the effect of AI. Our survey found only 36% or respondents believe it will have a significant or moderate impact and 40% are unsure whether the impact will be positive or negative.
 
Enterprise adoption
 
Although we are in the very early phases of machine learning adoption – probably in the flat bit at the start of an S curve by our reckoning – we found plenty of organizations already developing or deploying machine learning. Some 36% of respondents are currently either developing or deploying machine learning. Early adopters could be developing it themselves with in-house talent, using some sort of third party to develop it for them or buying applications with machine learning already built in – security tools of various types being a good example of the last of those. 
 
Exactly how they are developing and deploying is predictably varied, with no single approach dominating. Some respondents are using cloud-based tools and third-party systems, but a healthy portion are looking to buy applications with machine learning built in while others are looking to build in-house. Similarly, several different execution venues are available for both development and deployment, as are a proliferous set of tools. Cloud-based platforms are the most popular, with almost half using them, but breaking that down, we see where machine learning is being developed or deployed varies by the status of the initiative. For example, those that say they plan to use machine learning in the future are more inclined to choose cloud platforms than those already deploying machine learning. 
 
This survey paints an overall picture of what is going on in small to large enterprises in North America, Europe and Asia-Pacific. Future surveys will focus sharply on use cases and business benefits and delve even more deeply into the infrastructure changes that are required or already happening to make the promise of AI and machine learning attainable for all organizations.

Learn more about Voice of the Enterprise.
Written by: Nick Patience, 451 Research Founder and Research Vice President – Software
5335 Hits

The Trouble with Cloud “Repatriation”

Key analysts: Liam Eagle - Research Manager, Voice of the Enterprise: Cloud, Hosting & Managed Services, and Melanie Posey - Research Vice President and General Manager, Voice of the Enterprise

We have all known someone who regretted getting a tattoo. A seemingly permanent choice they loved ten years ago, is now covered up or removed. But who knows where they will be in another 10 years. In a nutshell, this analogy describes the phenomenon many IT analysts called “cloud repatriation” – the shift of workloads from public cloud to local infrastructure environments.

We have also called this phenomenon “cloud repatriation,” though not without some internal healthy debates about the validity of the phrase (yes – it rivaled the infamous “what color is the dress” debate for us). The debate focused on the word “repatriation” and how it suggests a transition to a permanent state of being. That doesn’t accurately describe an enterprise’s relationship with IT infrastructure. Much like our acquaintance full of tattoo regret, the priorities, needs and sometimes the entire business model of an enterprise changes and makes that new infrastructure less effective or appropriate than it was first deployed.
The Trouble with Cloud Repatriation fig 1The term “repatriation” takes a lot for granted. It assumes a permanent outcome or “permanent residency,” if you will. It assumes an improvement of some kind. And it assumes some failure to deliver on the part of public cloud. But what about the rolling back, for instance, of a failed cloud migration project? Is that repatriation? What if they try again in a few months?

In our Voice of the Enterprise (VoTE) Cloud Transformation, Organizational Dynamics 2017 survey, 34% of respondents said they had moved their workloads from a public cloud to a private environment (cloud or otherwise). Notably, when we asked them to cite their reasons for the move, many of those matched the reasons we know businesses ultimately decide to shift to the public cloud in the first place: performance/availability issues, high cost, latency issues, security and more.
The Trouble with Cloud Repatriation fig 2
These reasons appeared in our survey again when we asked respondents about their reasons for using multiple infrastructure environments to operate individual workloads. Forty-seven (47%) percent of them said improving performance/availability was one of the main reasons for leveraging multiple infrastructure environments.

What does this mean? For us at 451 Research, we believe this phenomenon we are all hinting at isn’t a repatriation or a reverse migration, but a cloud evolution. Our data shows more businesses see the value of a hybrid IT strategy. In fact, 58% of VotE survey respondents said that they are “moving toward a hybrid IT environment that leverages both on-premises systems and off-premises cloud/hosted resources in an integrated fashion.” Hybrid IT does offer the opportunity to build a framework for workload portability and mobility to match the ever-changing needs (or tastes, looking back at the tattoo analogy) of an organization. While this strategy isn’t the solution for all IT organizations, it is true that VMs are rarely stagnant because the best execution venue (our term for “home”) changes based on the available resources and shifting requirements around issues like performance, security and availability.

So, here is our problem: how will we rename cloud repatriation? We have played with many ideas like workload (re)balancing and liquid workloads (no, we are not suggesting you water your workload like a plant – water and electronics still do not mix), but those don’t sit well with us either. Tweet us your ideas for a new term, one that fits the "fluid" nature of the best execution venue for workloads and hybrid IT. 

5963 Hits

IoT = IT + OT: Solving the Enterprise IoT Equation

The Internet of Things (IoT) represents an enormous opportunity for technology vendors. But selling IoT outcomes – in particular into industries that have deep industrial or operational technology (OT) roots – is a fundamental departure from the traditional IT sale.

Vendors face a more complex solution sale that goes beyond the usual IT buyer and involves multiple stakeholders that sit at the crossroads of IT, OT and strategic business agendas. For the first time ever, 451 Research launched a Voice of the Enterprise (VoTE) IoT: OT Stakeholder Perspective survey in addition to the quarterly VotE IoT: survey of IT decision-makers. Together, these two surveys provide complementary insights and guidance into the IT and OT view of enterprise IoT deployments, budgets, spending, technology priorities, vendor choices and more.

Join 451 Research’s Rich Karpinski, Research Director for Voice of the Enterprise – IoT, to see highlights from this first-ever VotE IoT: OT Stakeholder Perspective survey and learn about:

• How to best work with OT and IT buyers
• The key OT-heavy verticals driving industrial IoT, including manufacturing, utilities, transportation, oil and gas and smart cities
• How to target the very specific OT-centric use cases that are driving IoT deployments and spending in those verticals today – and in the future
• How to position IoT products and services effectively with OT decision-makers – what they want and need from both their IT- and OT-centric vendors
• Where and how OT and IT are collaborating most effectively on IoT – and where and why they are in active conflict (and how you can help)

3024 Hits

Infographic: Changing Information Security Spending Landscape

The Voice of the Enterprise (VotE): Information Security, Budgets and Outlook survey represents the second in-depth look at security budgeting. A year-by-year comparison and predictive look into the future, shows security budgets increasing but also coming under various forms of stress that are currently shifting spending allocations and portend future changes. These shifts are taking the form of increased spending on software- over hardware-based security solutions, opex over capex and endpoint-based solutions over network-based ones. Interested in learning more about Voice of the Enterprise? Download the full infographic, or click here for more information.
7981 Hits

Featured Data: Top three storage pain points for enterprise customers

2017 04 05 COTW

Voice of the Enterprise: Storage survey respondents indicate that handling data growth, meeting disaster-recovery requirements and high costs are the top three storage pain points. But other pain points, most notably growth from new applications, are up markedly YoY. Nearly one in five respondents cite managing data stored in third-party cloud environments as a top pain point.

The core challenges in storage are shifting; public cloud is beginning to assume the burden of inexorable data growth, and the emphasis is moving to other areas, such as effectively managing data and storage across both on- and off-premises locations. Organizations are increasingly moving to hybrid cloud environments, bringing new providers into the mix (chiefly public cloud providers). The core role of IT infrastructure managers is shifting away from adding capacity toward more effectively managing it. Some storage managers are now charged with handling an organization's cloud storage capabilities – which is often a steep learning curve, and one that incumbent vendors can help alleviate by adding more comprehensive cloud-based integration to their offerings.

The VotE: Storage, Budgets and Outlook study represents more than 1,000 completed surveys from pre-qualified IT decision-makers about adoption levels of various cloud storage technologies, and the impact of public cloud and SaaS strategies on enterprise storage.

If you're interested in reading more of our research, apply for a trial here.
7241 Hits