- When it comes to enterprise data and analytics investments among enterprise IT buyers, recent GlobalData research points to a future dominated by all things cloudy.
- But how are buyers prioritizing specific areas of investment such as workloads, including BI, data warehousing, and AI? We dive into a new survey of more than 3,000 IT practitioners to find out.
As an industry analyst, when a new survey arrives at your doorstep, you greet it as you would the arrival of an old friend who’s been oversees exploring other cultures and climes. After the customary hugs and hellos, it’s instantly down to business. Where did you go? What did you see and learn? And so it was with our newly returned 2018 Global IT Customer Insight survey. Now in its sixth iteration, this annual survey plumbs the international depths of the IT buyer landscape to discover current and future buying priorities across a wide array of investment areas.
Needless to say, with 3,249 respondents at hand, spanning all major company sizes, geographies, and vertical markets, this global survey stands as a significant helpmate, particularly in helping we analysts validate (or disprove) our first-person view of market transitions.
Case in point: cloud migration. 2018 spending on big data and AI platforms both saw a combined increase of more than an 11% on average, versus the diminutive spending increase of only 2% revealed in last year’s survey. Big data platforms on their own saw a massive 19% increase in spending over the past year, with 69% of IT budget holders setting money aside for data at scale across the enterprise. It is an obvious, foregone conclusion that the cloud is the way forward for IT buyers and technology vendors alike.
What may not be so obvious, though, is how IT buyers prioritize these different workloads that are ‘destined’ for the cloud. Over the coming months, we will dig into our new data set and discuss these priorities in detail, but for now, here are a few self-evident conclusions and some initial thoughts thereon.
First and foremost, the cloud will dominate future investment priorities across the board for all major areas of investment that concern big data and analytics, as you can see in Figure 1. We actually expect a 13% increase in investments among those currently leveraging the cloud, meaning the rate of investment in the cloud will only pick up momentum throughout the year. The management of huge amounts of data (a la data warehousing) still favors on-premises deployment models.
This is also true for workloads that concern potentially sensitive data, as with customer marketing and customer relationship management solutions, perhaps also a reaction to recent data protection legislation such as GDPR. For these areas, on-premises deployments continue to be the status quo. But, for both BI and big data solutions, it appears current investment priorities are equally split between cloud and premises. This tracks with the slow rate at which traditional BI players have migrated their on-premises software to the cloud.
Lastly, and perhaps most interestingly, despite the heavy data storage and processing requirements associated with AI implementations, the cloud is now and will be the dominant target for the foreseeable future. This clearly points to the fact that the highly specialized nature of AI’s demanding infrastructure requirements favors economies of scale – something currently best consumed via global cloud platforms from providers like Google, Amazon, IBM, SAP, and Microsoft.
These vendors have invested significantly in hosting and operationalizing AI frameworks such as TensorFlow, Caffe, Theano, and Torch. They have made available to their customers’ sizable banks of on-demand AI hardware (e.g., Google TPUs, NVIDIA GPUs), which are tuned to supporting AI demands such as predictive model training. Obviously, unlike the mainframe and high-performance computing (HPC) of the 90s, the best place to get high-end processing power is on the web.
Next up, we’ll look at how future cloud buying strategies differ between those prioritizing either cloud or premises today.