EMA has just completed a groundbreaking (the word is apt here) research project assessing just how IT organizations are seeking to invest in, optimize, integrate and prioritize use cases for what we call "Advanced Operations Analytics" or "AOA." AOA is our term for "big data for IT" which others have termed "operations analytics" and which EMA initially described as "advanced performance analytics".
In reality all these terms, including AOA frankly, leave something to be desired. First of all, AOA isn't limited to operations and in fact our data shows that it's far more often driven from the executive suite or from a cross-domain cloud team than from operations as a whole. AOA benefits many stakeholders outside of operations including non-IT stakeholders, such as online operations and business planning. Moreover, AOA is not just about "performance" as our first coinage somewhat underhandedly proposed. (We knew better then, but we just milked the term "performance" for as many variations as we could get away with.)
This research — fresh from August data collection — included 250 respondents with 150 in the US and 100 in Europe (evenly divided across England, Germany and France). We required that all respondents be involved with cross-domain analytics for IT either as stakeholders or implementers, and required as well that the AOA investments be either already deployed, or actively in deployment (where most respondents were), or in PoC, or at minimum in planning mode. This worked well and about 70% were either fully deployed or actively in deployment.
I'll be giving a webinar with more details and grit than I can share here — so let me pause to provide you with a link for the invite: Advanced Operations Analytics and Beyond - A Look at Real Adoptions and Real Plans in the Real World
We asked about heuristics — and data mining came in with the lead, but real-time predictive and predictive trending were only a few percentage points behind. The general array of heuristics of interest, however, underscored the fact that AOA is not about a single technology, but multiple options that come together in various ways and design points. In fact, the average respondent indicated about three separate choices for AOA heuristics — from the above categories to others such as anomaly detection, correlation, machine learning, and if/then change impact analysis.
What does this mean, really? Another way of looking at the many faces of AOA is use cases and benefits, which I'll examine in more detail in my next blog. But for now suffice it to say that these ranged from security, to ITSM/Service Desk, to software development, to application management, to change and configuration management, to IT asset management and financial optimization. And this list is still far from complete.
The point about AOA is that once you find a way to pull data together into a common resource — whether it's an aggregated data store provided by IT or a data store integrated into the vendor's own AOA solution set — there's A LOT you can begin to do with it. Just like CMDB Systems, AOA helps to break down IT silos and juice forward more effective, collective decision-making. With AOA, interdependencies across cost, performance, change, capacity optimization and business impact can all potentially be viewed with a common modular eye.
This is one reason why the IT executive suite is so keen on AOA as the data shows. AOA is potentially transformative; gets astonishingly good marks in supporting the move to cloud and agile; and helps to clarify IT business value as well as manage costs if used well.
A look back at the AOA market before it emerged into prominence in the last 18 months would have begun with innovators like Netuitive and ProactiveNet (now a part of BMC), with advanced learning combined with eclectic data collection embedded in a single tool. As our data now shows, the current trend is to move toward aggregated data stores, often with some in-house (versus commercial, third-party) ownership — in large part because AOA is still effectively brand new. But innovators such as our research sponsors — in alphabetical order, AppDynamics, AppEnsure, HP, IBM, Prelert and Splunk — all reflect evolving, though different, approaches to the AOA opportunity. And they are not at all alone. There are more than 20 vendors currently exploring AOA at some level — and the list is sure to expand significantly in the coming years.
We'll be following up with more insight on AOA use cases, stakeholders and benefits, as well as more insight into AOA design points and what you might want to look for before investing. But I'd like to end with one data point from the research. In the grand, progressive, AOA landscape, data collection still favors, guess what, spreadsheets! Spreadsheets were the number one source, followed by performance-related events and configuration data. Log files and access logs came in next. So, as the cliché goes, "some things never change" — or when they do, it seems to occur more on a paleontological than a high-tech time scale.
But even with this reptilian blemish, I think I can comfortably predict that AOA is here — and here to stay.
In other words, it's finally time to make way for the mammals.