We Need To Talk About IT Analytics
Last week I attended the excellent FutureStack conference. This was the first customer conference by New Relic, the Application Performance Management (APM) and Monitoring company. It was great to see how passionate their customers are and how they realize the strategic importance of monitoring. Well done New Relic! The keynotes and track sessions at this event were great because they did not just focus on technology but addressed the future skills and competencies required for today’s business technology professional.
The main focus of this conference, though, was to unveil New Relic’s Rubicon solution (developed by CEO Lew Cirne during his summer vacation!) which firmly shifts New Relic from being a good APM company into the world of analytics or as they define it – ‘software analytics’. From what I have seen of Rubicon it looks like it has the potential to become a platform for a good analytics strategy. Rubicon, when available next year, will allow the rapid querying of software application centric data via a very easy to learn querying language called NRQL (Lew was creating queries quickly on the fly during his keynotes). The only quick, comparable query language I have seen here is Splunk’s Search Processing Language which is very easy to use with Splunk Enterprise 6. In today’s modern enterprise, applications fuel both internal business capabilities and are increasingly being used to engage customers externally through mobile approaches. As a result, this form of analytics will be critical to forward-thinking organizations. IT Analytics is a main topic of conversation in many of the monitoring and APM briefings and inquiries that I have taken part in this year BUT if you are reviewing solutions or looking to make decisions in regards to your analytics strategy then here are some points to take into consideration based on solutions I have seen:
IT Analytics from a technology perspective. Any analytics solution that you evaluate must be able to store vast amounts of data economically. This means that traditional relational database approaches are not adequate, and non-relational data stores should be the basis of the solution that take into consideration storage and retrieval principles of Hadoop, Cassandra, MapReduce etc. One of the feature requirements is how quickly data can be queried and pulled back from the data store, as analytics is about providing competitive insight and therefore speed is a priority. Linked to this, the query language should be simple and intuitive. New Relic’s and Splunk’s solutions here meet these requirements but you also need to take into account the license costs for each solution. The license costs for Rubicon have not yet been released, but Splunk’s are based on data indexed meaning the more data you index, potentially the higher the license cost. Equally you also need to decide whether your business requires an on-premises or a SaaS (Software As A Service) solution. New Relic’s Rubicon is SaaS based while Splunk Enterprise 6 is on-premises.
IT Analytics from a process perspective. While having a good IT analytics technology solution is no doubt the foundation to an analytics strategy, the real value comes from making the right insight. The IT analytics technology solution will allow the operator to turn data into information but it’s important that you query or ask the right questions. So what initial questions will you ask? What information will potentially show the business value of an IT analytics strategy and show the strengths of the solution? These questions are critical if you are going to achieve Return On Investment and are essential in order to get business support. One of the ways in which New Relic and Splunk are trying to help is by building up collaborative customer communities that include various industry segments. Of course you don’t want to give valuable competitive insight to a competitor though so it’s about achieving the right balance of privacy and collaboration, but value can be realized faster by collaborating in these communities.
IT Analytics from a people perspective. Even if you know what initial questions to ask, how are you going to present it in context of your audience in order to generate knowledge and insight? The standard IT professional way of doing this is to create a visual dashboard. But I believe that this is only the first step because in order to ensure that everyone understands the value in the insights being shown it’s important to wrap around the visual dashboard with a true to life, company/industry specific story thus avoiding any misinterpretation of the information being shown. Only when the information is interpreted correctly by the audience can it be turned into competitive knowledge, insight and be the basis of appropriate actions – the real value of IT analytics.
So, yes, you should be excited by ‘IT Analytics’ but just remember that the solution is only part of the answer. As always any comments or further recommendations are always welcome.