Start with Progressing Expectations in Advanced IT Analytics: How the Industry is Still Getting it Wrong - Part 1

When I talk with clients, deployments and even when I take time out to read the headlines, I still believe that the industry is confused about how Advanced IT Analytics (AIA) is evolving. Here are three big reasons why, as supported by our current data.

1. AIA cannot be adequately defined by discrete technologies — including Big Data

Industry attempts to define AIA purely in terms of discrete technologies or else simply "big data" may seem reasonable at first glance, but they actually create fictitious barriers. So many different analytic heuristics can come into play from machine learning, to advanced event correlation, if/then change-related analytics, just to name a few.

Similarly, the data sets involved may span everything from events, to time series data, to log files, to spreadsheets, to text and sentiment analysis, and the list goes on. And while the quantity of data is key, thinking of AIA primarily in terms of big data may turn out to be a recipe for failure more often than not. This is an area where relevance, use case and focus seem to count far more than science projects.

2. AIA is not just about operations

The more prevalent nomenclature tends to direct advanced analytics for IT purely as an enabler for operations, whereas the data in this report shows just the opposite is true. AIA as EMA understands it is a unifying layer that may support many IT roles from operations, to ITSM teams, to development, to the IT executive suite, to a growing number of business stakeholders. In fact, the 11 roles targeted by our AIA respondents included 4 domain, 4 cross-domain and 3 business stakeholders on average. Moreover, the IT executive suite led in driving AIA initiatives followed — are you ready for this? — by IT service management teams (ITSM) teams and cloud teams.

3. Service modeling interdependencies and AIA go hand-in-hand

Only 4% of our respondents showed no interest in capturing service models or service modeling interdependencies. The lead sources were:

1. Application discovery and dependency mapping (ADDM) for performance

2. Service modeling dashboard for business impact

3. Service modeling/system provided directly through the tool

4. ADDM for change – tied with a federate configuration management system (CMS)

OK, now think about the industry prejudices in this area. How many industry pundits out there are linking advanced analytics to anything resembling a CMDB? And yet there are many reasons for why this is becoming increasingly critical — and cloud, believe it or not, is one of them (as is also suggested by this research and other data). If you're an old network management hand, you may remember how the advent of topology once worked wonders. Now, it's just beginning to do it all over again in some more advanced environments — in a much more dimensional set of contexts and attributes that hold the potential to link stakeholders, processes, automation routines and business outcomes in a single dynamic fabric.

So Maybe AIA Isn't a Market After All

In the end, I would have to admit that AIA as I view it probably isn't a market in the traditional sense, since market definitions typically require narrow technology parameters for creating discrete buckets for market sizing and contrast. My vision of AIA is rather an arena of fast-growing exploration and invention, in which in-house development is beginning to cede to third-party solutions that can accelerate time to value (also something indicated by our research).

It's still too early to say for sure, but the progress and the opportunities are certainly exciting — and putting artificial boundaries around them isn't going to improve the pace of progress.