AI in ITSM: Data is the key to success
Research firm, Gartner Inc., reports that up to 85% of enterprise AI initiatives fail. The main reason: poor data quality. Followed by poorly defined target use cases and lack of AI skill and understanding within the organization.
Whether it’s routing tickets, managing SLAs, or deflecting support tickets, every ITSM and Enterprise Service Management (ESM) capability depends on a reliable foundation of clean, categorized, and complete information.
When records are fragmented (or knowledge is outdated or missing), any AI experience that is driven by this data becomes unreliable—often frustrating users rather than empowering them.
Success starts with data
A successful AI initiative starts with data—of the right quantity and quality. A strong data foundation enables:
- Higher accuracy outputs and a reduction in bias.
- Improved decision-making based on factual insights.
- Increased trust in AI outputs among stakeholders.
- Faster recognition of patterns and identification of opportunities.
When stakeholders see that AI outputs are based on clean, verified data, trust increases and resistance diminishes.
From 70% to 97%: Building a strong foundation
70% data coverage is considered the industry benchmark required to deploy an agentic AI. But the 70% threshold still leaves a lot of room for hallucination, misinformation, and missed automation opportunities. In short, more data coverage means more of your AI-powered use cases can be implemented—and perform at a level that meets expectations.
At Hornbill, we’ve created custom machine learning tools that focus on the quality of the underlying service management data. These tools are a catalyst for driving data coverage from 70% to 97%—creating a more stable data foundation which powers high-performing, context-aware AI solutions.
How to become AI-ready: Data driven practices
| 1. Data quality audits |
Begin by identifying gaps in record completeness, categorization, and accuracy. Filling gaps and improving data integrity leads to:
|
| 2. Knowledge management reviews |
Audit your knowledge base to:
Use generative AI to create new knowledge where gaps exist, either autonomously or with human-in-the-loop quality assurance. Human-in-the-loop quality assurance means your subject matter experts review and verify the knowledge content before it is used as content to enable self-service and power AI. |
| 3. Automation opportunity mapping |
Analyze current processes to find automation-ready use case candidates. Some may not require AI, but efficiency can be gained through traditional process and task automation. For scenarios that do require AI to operate, machine learning can rapidly highlight automation scenarios that manual reviews would miss. |
| Next: From foundation to best practice |
With data in place, organizations can confidently adopt best practices for AI integration into service ecosystems. These practices—supported by the latest global research—will help organizations deploy AI responsibly, reduce risk, and maximize value for employees and customers alike. |
Tools to help you find and resolve data issues
Our AI Flight Check service is a solution to the challenges of AI data readiness—designed to accelerate and smooth-out the path to adoption by taking the complexity and hard work out of assessing and cleaning your service management data.
The FlightCheck service is enabled by our consultants and a set of purpose-built tools. These tools automate the process of analysing and assessing your data—telling you precisely where there are data quality anomalies and concerns that could compromise the output of Hornbill's proprietary ML models for service management.
That means you don’t need to manually trawl through your data to find and eliminate issues (which could take weeks or months). Our consultants will run the tools, analyse the data to pinpoint specific problems, and work with you to quickly resolve them. During this process, your data remains secure and won’t leave your Hornbill instance.
Next: From foundation to best practice
With data in place, organizations can confidently adopt best practices for AI integration into service ecosystems. These practices—supported by the latest global research—will help organizations deploy AI responsibly, reduce risk, and maximize value for employees and customers alike.
Read this report to find out more about the importance of good data for an AITSM initiative:
Find out more about data readiness for AI
