During the past few days I had the fun and privilege of spending time at two ServiceNow user groups here in Ohio (I recommend not visiting Cleveland in July – the hail is brutal). After the time of learning about ServiceNow, seeing cool features planned in future releases, and generally talking with my fellow fans and cohorts of the product (along with enjoying the fine beverages served during happy hour), I was talking to a friend and colleague about the future of ITSM tools and I had a chance to brainstorm what the next five to ten years may look like. I certainly won’t make the claim as being someone that can successfully predict the future, but I came to the conclusion that the next “big” thing isn’t just a new tool or technology, and it’s not even automating our tools or technology, but rather what I’m waiting to see is some intelligence in a tool. And not only intelligence, but something that can actually improve itself over time.
I’d like to take a step back and give a picture of where we are today. IT is going through some fantastic changes and in the ITSM industry, there’s a player called ServiceNow that is currently the 800 pound gorilla in the market. The tool came from humble beginnings in the mid-2000′s in which no one had a product in the market that was using web 2.0 technologies. Sure, Remedy, CA and HP were all big players, but their tools were built on proprietary languages and still had to be hosted onsite. So when ServiceNow hit the market and not only was a cloud-based solution, but was also easy to develop for as a platform, it’s no wonder the product has grown so quickly and by so much. Getting back to today, ServiceNow is now (that’s a lot of “nows”) positioning itself to be an enterprise platform for applications used by people outside of IT. In other words, it’s going away from just being another ITSM tool and the recent marketing material suggests that’s where Fred and Frank want to take the product.
Moving ahead slightly into the future, I’m going to guess that we’ll eventually see another product come out that, like ServiceNow, will start small, be built on the recent common technologies, will have the potential to be a platform as well as be something that has an ease of development, will incorporate best practice ITSM processes, and will be inexpensive for IT departments to not only purchase, but also to implement. If such a tool also includes a social media and mobile platform, then it’s sure to disrupt ServiceNow’s market in ITSM. But such a product will only disrupt, not dethrone. And there is where the conversation over dinner turned into answering the question….what will dethrone ServiceNow in ITSM? And please keep in mind, I specifically put “in ITSM” for the very good reason that ServiceNow is looking to be an enterprise platform and competing on such a level will probably take a force of nature. But change the scope to only IT and the core processes, and the field of competition will change as well.
In the IT world we’ve had a few interesting themes that I’d like to bring to attention. First, analytics. How to take data (and probably a lot of it) and find meaningful patterns, preferably to help make intelligent decisions. Second, artificial intelligence. With AI, we want to essentially automate finding the patterns and making the decisions. I think the best quote I’ve heard in the context of this blog posting is “Artificial intelligence is common sense wrapped around an algorithm.” – Christopher Dancy. With that in mind I hope you know where I’m going with the “evolution” in ITSM tools…automated intelligence.
I describe the idea (and I’m sure other people want this and have stated it, so I won’t say I’m the first to think of it) of automated intelligence as that of an IT product that can take in vast amounts of data, or inputs, and over time start to identify patterns to the inputs that leads to high probably of correct outputs. In other words, it’s something that just learns how people in an organization operate and eventually starts to learn what they want. Of course, there would need to be some kind of algorithm that needs to rank the probably of outputs, and human intervention is a necessity to ensure people have input into the process.
Now, does any of this sound familiar to you? I hope so, because Google is already providing such a service on their webpage. The Google search takes input, and based on several pieces of data (your location, search history, page ranking, etc.) it gives results based on what you want to find. In the ITSM world it wouldn’t be as simple as displaying web pages, but in our IT world each request really only provides one of a small handful of results; someone needs something fixed, someone needs information, or someone needs something new. Those three items can be translated to creation of an incident ticket, direction to a knowledge article, or creation of a service request. Could it get more complex? Absolutely, but for purposes of this blog post I’d rather keep it simple.
And this is where Google has the lead; they already have the experience and technology for being able to take input and provide a ranking of outputs. Not only do they have the experience, but they have the experience based on the human practice of typing searches into a free-text format. When it comes to the ITSM world, they would even have an easier time with providing correct results as several pieces of data would be in a structured environment and easy for an algorithm to search.
So Google has the advantage, but why even try to go with automated intelligence? My thinking is simple: The goal is to have a way to have an IT tool improve itself, and with improvement based on the patterns and behaviors of the environment. As an example of how such a system could work, let me lay out a scenario of a high priority incident. It has information such as the location, CI, key words in the description, etc., etc. Eventually that information provided on the incident form creates the actual P1 incident and during its life it gets passed around between groups, until something major goes wrong and the (hopefully they have one) incident manager opens a bridge to resolve the incident. Fast forward to weeks or months into the future and another incidents gets created matching most of the input from the previous incident. The person created the incident saves it with a P1 and assigns it to some group as the previous ticket. At this point, an intelligent system would match that inputs are similar, and immediately prompt the incident manager that a similar ticket was created that required a bridge to be opened for resolution. In this case, a previous experience matched to an outcome would help automate matching a similar experience to the same outcome. Ironically, it’s the exact same way that humans work where experience influences our behavior.
When you consider how Google works, it just takes input and matches it to output based on previous experience, and often by using the search experience of others. Us humans work the same way, so why can’t our tools do it? Sure, the algorithms will be complex and will always need tweaking, and it’s something that will build value over time, but why can’t it work? We’re simply dealing with inputs and outputs and matching them from previous experience. If Google can become a multibillion dollar company from the concept, certainly the ITSM industry is ready for a tool that does it as well.