How long has our industry been talking about the volume of data potentially trapped in IT service management (ITSM) or enterprise service management (ESM) tools? There are many different flavors of this, including:
- Tool-related issues – where you just can’t get what you want or need out of your ITSM tool. This can also result from implementation decisions.
- Information or data is collected, but the reason why is unknown – for example, a field is added to a form, but nothing is done with what’s captured.
- Inconsistency – while information and data are collected, it can’t easily be aggregated and analyzed.
I’m sure there are many other reasons why we consistently talk of the data “trapped” in ITSM tools, but the collection with no purpose or that is underutilized stands out as an opportunity to use artificial intelligence (AI) to understand and improve more.
Unused ITSM tool data
The issue is not so much that “unused” ITSM/ESM data is collected. It’s more that it’s collected without a purpose, or without understanding its potential purpose(s). This is likely collected because someone, somewhere, said to do so; it might even be mentioned in industry best practices as a “good thing to do.”
Effort goes into collecting the data, but it ends up in the ITSM tool equivalent of a black hole. Likely lost forever. That is until AI arrives from stage left to help out – more on this later.
One of the most obvious examples of this was the heavy investment in configuration management databases (CMDBs) two decades ago, when the purpose was seen as “configuration management,” not the collection and management of configuration item (CI) attributes to help other ITSM needs such as incident management, change enablement, problem management, etc. Sadly, this issue can still happen – data is collected but never used (or never used to add value).
Underutilized ITSM tool data
Continuing from the above, a good example is CMDB fields that get created but are insufficiently used. Yes, even high-quality CMDB data can be underutilized. For example, aggregated “end of life” data might not be used to drive or inform technology refreshes or contract renewals. It may only ever be checked to prove why a particular device is underperforming.
Another example is that incident management and service request management ticket fields are captured for categorization but never analyzed downstream. IT service desk setups can easily over- or under-collect data, and the populated fields don’t always map cleanly to the metrics anyone is actually tracking.
Finally, free-text resolution notes and closure comments get typed into nearly every ticket but might not get “mined.” Thankfully, AI is now helping with exploiting this information/data more – perhaps as part of knowledge management.
Should your IT organization collect more or less?
This is very much about focusing on “what matters most”:
- What needs to be addressed or improved in your organization?
- What data is available to help with this?
This need to improve in a specific area is a good place to start when looking for new datapoints to capture. But, as with any collection process, the cost of collection shouldn’t outweigh the benefits.
A good example is experience management. An IT service desk already collects a lot of data (and likely faces more performance metrics than other IT capabilities). However, very little of what’s already collected (about IT service desk performance) can inform insights into employee experiences and opportunities to improve. There’s likely a need to collect more data.
“Not more datapoints!” I hear you cry. Don’t worry, you can always stop collecting some data. Quite simply, focus on what’s collected but not used. It should either be used (to add defined value) or not collected at all. Collecting data just in case you need it later is not a great idea.
Using service data beyond the obvious
The data available in ITSM tools can serve a wide range of IT and business purposes; i.e. the data ITSM tools collect for service delivery and support can have a second life when other departments start pulling from the same platform.
For example, identifying underutilized services or services in need of improvement. Or, in ESM scenarios, different business functions might handle the same issues inconsistently; for example, HR and Facilities might handle changes to employee roles differently.
Returning to the CMDB, configuration and asset records might support incident management and change enablement work in IT. But the same records answer questions finance cares about: What software is actually licensed versus paid for? Or governance teams: What needs to be replaced before it fails an audit?
AI will help with this issue
Returning to the opening paragraph, we have long bemoaned that ITSM tools contain so much data that isn’t exploited.
As mentioned earlier, AI will help.
Just as AI is currently used to optimize ITSM processes, it can also help you understand what you collect is used for. It can spot “wasted” data and recommend additional use cases, cessations, or even new collection opportunities.
Organizations should be doing this, but many won’t be (yet), likely with process optimization as the primary use case for AI’s ability to assess and recommend improvements. Please stop for a minute to consider whether “data optimization” might be just as valuable.
ITSM Tool Data FAQs
It covers a few different issues: tool limitations that stop you getting data out, what’s collected without anyone knowing why, and data that’s collected but is too inconsistent to aggregate or analyze.
Unused data is collected and then never touched again, effectively lost. Underutilized data is being used, just not to its full potential, like aggregated end-of-life data that only gets checked to explain a device’s poor performance rather than driving technology refresh or renewal decisions.
Two decades of heavy CMDB investment framed the purpose as “configuration management” alone, rather than as a source of configuration item data to support incident management, change enablement, problem management, and other ITSM needs. That gap between collection and use can still show up today.
It depends on what needs improving and what would actually help. Data collected “just in case” isn’t a good idea. The better approach is to either put existing data to defined use or stop collecting it.
Yes. Configuration and asset records that support IT incident and change work can also answer questions finance cares about, like what software is licensed versus paid for, or what governance teams need replaced before an audit.
AI can identify data that’s going to waste, recommend new use cases for what’s already being collected, suggest stopping collection where there’s no value, and flag new datapoints worth collecting, alongside its more established role in process optimization.
Stephen Mann
Principal Analyst and Content Director at the ITSM-focused industry analyst firm ITSM.tools. Also an independent IT and IT service management marketing content creator, and a frequent blogger, writer, and presenter on the challenges and opportunities for IT service management professionals.
Previously held positions in IT research and analysis (at IT industry analyst firms Ovum and Forrester and the UK Post Office), IT service management consultancy, enterprise IT service desk and IT service management, IT asset management, innovation and creativity facilitation, project management, finance consultancy, internal audit, and product marketing for a SaaS IT service management technology vendor.
