Tag Archives: data goveranance

AI, ITSM

Five critical steps for making a good AI/ITSM decision

Leave a comment
Share twitterlinkedinmail

There is no question that AI-enabled technologies have the potential for significant positive impact for organizations overall, and for ITSM specifically.  This recent TechTarget article highlights a number of positive business impacts resulting from the adoption of AI-enabled technologies, such as new capabilities and business model expansion, better quality, more innovation, and personalized customer services and experiences.

New and existing ITSM-related vendors are rushing into the space with solutions like AI-powered automation, conversational AI, intelligent chatbots, predictive analytics, and agentic AI (A web search on these terms will return numerous examples!).

And we’re only scratching the surface.  New AI-enabled capabilities are on the horizon, such as:

  • AI agents capable of executing discrete tasks independently based on personal preferences or providing customer service without requiring specific prompts.[i]
  • AI-powered cybersecurity in the form of automated, near-constant backup procedures and AI tools for managing sensitive data to enhance data protection and resilience.[ii]
  • Small Language Models (SLM) that aim to optimize models for existing use cases. SLMs can be trained on smaller, highly curated data sets to solve specific problems, rather than act on general queries (like Large Language Models).[iii]

But just because these rapidly-evolving technologies represent the “latest shiny new thing that really helps” (a tip of the cap to Paul Wilkinson) doesn’t mean that you should succumb to the fear of missing out by just “doing something”. In my experience, a new technology alone rarely (if ever) solves a business challenge.  When it comes to technology investments, it’s better to make a good, informed decision, based upon the unique needs and challenges faced by your organization.

Yet, AI-related technologies can have and are having a significant positive impact on ITSM environments. Many organizations are already benefitting from the use of AI-enabled chatbots, automated ticket management, and service request automation.

The pressure to introduce AI-enabled capabilities to ITSM implementations is real. But which tools?  What capabilities?  How can one decide?

Five critical steps

Here are my five critical steps to making a good AI/ITSM decision.

  • Define overarching goals for using AI within ITSM. It’s easy to become captivated by the latest products and features, especially in today’s AI/ITSM market frenzy. But chasing new products and features usually results in a short-sighted approach to technology adoption that will likely not meet longer term goals and needs. AI within ITSM should not be approached as a point solution; rather, AI should be considered within the broader perspective of ITSM. How will adding AI capabilities address current challenges?  How will adding AI enable the organization to realize future ITSM objectives? Defining overarching goals for AI in ITSM – in business terms – ensures that broader perspective .  Defining overarching goals also establishes the foundation for measuring AI/ITSM success.
  • Conduct a SWOT analysis of the ITSM environment. Conducting a SWOT analysis identifies a company’s internal strengths and weaknesses, as well as external threats and opportunities. Understanding an organization’s ITSM SWOT identifies the critical factors that must be considered before developing an AI strategy. A SWOT is a good way to understand an organization’s readiness and ability to take on an AI initiative.  Having the right stakeholders participate is critical to the success of a SWOT. Include stakeholders (especially non-IT colleagues) that have an interest in both ITSM and in AI capabilities and use.  Include stakeholders that will freely share thoughts and ideas and have a pragmatic understanding of organizational issues and challenges.
  • Develop the AI strategy. What is the approach for bringing in AI into your service management implementation? An effective AI strategy is not about finding places to “plug-in” an AI solution. It’s about understanding the organizational change, data, skills, budget, and infrastructure that will be needed for successfully utilizing AI technologies within the ITSM environment to help achieve the organization’s mission, vision, and goals.  Use the results of the ITSM SWOT as an input to this strategy.
  • Define evaluation criteria. The next step is to define the criteria by which potential AI solutions will be assessed. Defining this criteria up-front helps prevent falling victim to ‘shiny object syndrome’ and identify the solution that is best for your organization. As part of that criteria, consider the solutions alignment with the AI/ITSM strategy, costs (initial, ongoing, and cost effectiveness), the effectiveness of the solution to leverage issues identified in the SWOT, and how the solution enables the pursuit of potential future opportunities.
  • Develop and present the business case. Gaining and maintaining the commitment of senior management is critical for success.  When a potential solution is found, develop and present the business case for that solution. Discuss the technical and cultural challenges that come with AI adoption. Discuss the opportunities that AI with ITSM will provide.  Discuss how a solution will address SWOT and align with the AI strategy.  Discuss the benefits of implementing the solution , how risks will be optimized, and how success will be measured.  Discuss the consequences of doing nothing. Most importantly, ask for management commitment.

Cautions

Before moving forward with introducing AI within an ITSM environment, here are some cautions of which to be aware.

  • Good AI will not fix bad ITSM. The adoption of AI technologies can enable and enhance ITSM capabilities. However,  AI is not a “magic wand” that solves issues like poor process design, inadequate service management governance, and ineffective measurement and reporting.
  • Don’t overlook data quality and governance. Many organizations have data quality and data governance challenges. AI needs data – lots of it – and that data must be accurate, reliable, and trustworthy. Data quality and governance is not just a challenge for ITSM, it is an organizational problem.
  • Is there an ITSM strategy? Many organizations are not achieving the full potential of ITSM adoption. Rather than applying ITSM holistically, many implementations have only focused ITSM implementation on IT operational issues, and not on how ITSM enables business outcomes. Without an overarching ITSM strategy, AI investments risk becoming short-sighted and expensive point solutions that do not address business needs.

Augmenting the ITSM environment with the right AI capabilities can be a huge benefit for the organization, ITSM, and the employees of an organization.  But introducing AI within ITSM is not a decision to be taken lightly. Taking a systemic approach to identifying, justifying, and selecting solutions sets the right expectations with stakeholders and helps ensure successful introduction of ITSM with AI.

[i] https://www.uc.edu/news/articles/2025/01/innovation-experts-predict-top-tech-trends-for-2025.html , Retrieved January 2024.

[ii] Ibid.

[iii] https://www2.deloitte.com/content/dam/insights/articles/us187540_tech-trends-2025/DI_Tech-trends-2025.pdf , Retrieved January 2024.

 

 

 

 

 

 

 

Share twitterlinkedinmail
AI, ITSM

Three AI truths with IT Service Management

Leave a comment
Share twitterlinkedinmail

There’s no question that introducing AI capabilities can have a dramatic impact on IT Service Management (ITSM). Done well, AI adoption will free up ITSM professionals to do the work for which humans are uniquely qualified, like critical thinking, contextual understanding, and creative problem-solving. Furthermore, AI will enable organizations to realize many of the theoretical benefits of ITSM. For example, the use of AI and machine learning can leverage comprehensive in-depth data, not just a small recent sampling, for cause analysis, problem detection, and impact determination of problems. Another example is the use of AI can increase the data of the IT environment and automate the remediation of incidents.

But AI is not a “magic wand” for ITSM.

Before introducing AI capabilities into ITSM, organizations must first consider these three AI truths.

Truth #1 – AI needs good data

For the use of AI to be effective, it needs data. Lots of data. But, if that data is inaccurate, lacks integrity, or is not trustworthy, then the use of AI will only produce inaccurate or poor results.

Data quality is an issue that many organizations will have to tackle before realizing the complete benefits of introducing AI to their ITSM implementations. These means that organizations will have to step up their technology and data governance posture. According to this recent Privacera article, a fundamental principle of data governance is having a high-quality, trusted data source.  Having trusted data sources enables capabilities like ITSM to make accurate and reliable decisions regarding service management issues. But if the data sources used by ITSM tools contain data that is unregulated, the ability to automate responses is significantly hindered.

Truth #2 – AI doesn’t mean process design goes away

The need for effective ITSM processes and procedures doesn’t go away with AI adoption. Machine learning can be used to detect data patterns to understand what was done to resolve an issue. But what machine learning doesn’t do is determine if what is being done is the best approach. Machine learning doesn’t consider organizational goals and objectives with the adoption of ITSM. Machine learning cannot determine what processes are missing or need improvement to gain needed effectiveness and efficiency with ITSM.

Truth #3 – AI doesn’t replace knowledge

“Reducing cost”, often in the form of headcount reductions,  is frequently used as the justification for AI investment, as the use of AI will enable ITSM activities to be automated. And it’s true – many of the ITSM activities currently performed by humans can and should be replaced with AI-enabled capabilities, such as the automated fulfilment of service requests, automated response to incidents, and problem data analysis. But one of the hidden costs of using AI to justify headcount reductions is the form of knowledge loss – the knowledge inside people’s heads walks out the door when their positions are eliminated. And this is the knowledge that is critical for training the chatbots, developing the LLMs needed, and to the continual improvement of AI and ITSM.

While AI can provide the “how” for “what” needs to be done, it cannot answer the “why” it needs to be done.

Good Governance facilitates AI-enabled ITSM

Without governance,  AI can do some serious damage, not just with ITSM, but to the organization. As the role of IT organizations shifts from being data owners (often by default) to being data custodians, having well defined and enforced policies regarding data governance is critical. This means that the frequently found approach to governance consisting of an IT track and a corporate track is becoming untenable. As organizational processes and workflows become increasingly automated, enabled by AI capabilities, governance must become cross-functional[i] , with sales, marketing, HR, IT, and other organizational functions all involved. Organizations must consider and address data-related issues such as compliance with data privacy laws, ethical data use,  data security,  data management, and more.

An effective approach to governance enables organizations to define their digital strategy[ii] to maximize the business benefits of data assets and technology-focused initiatives. A digital strategy produces a blueprint for building the next version of the business, creating a bigger, broader picture of available options and down-line benefits.[iii] Creating a successful digital strategy requires an organization to carefully evaluate its systems and processes, including ITSM processes. And as ITSM processes are re-imagined for use across the enterprise in support of organizational value streams, effective governance becomes essential.

Getting ready for AI-enabled ITSM

What are some of the first steps organizations should take to get ready for AI-enabled ITSM?

  • Formalize continual improvement. One of the most important practices of an effective ITSM implementation is continual improvement. As organizations are continually evolving and changing, continual improvement ensures that ITSM practices evolve right alongside those business changes. And just like service management, AI adoption is not an “implement and forget”; in fact, AI will absolutely fail without formal continual improvement.
  • Answer the “why”. To say that there is so much hype around the use of AI within ITSM would be an understatement. Before jumping into AI, first develop and gain approval of the business case for using AI within ITSM. How will success be determined and measured? What opportunities for innovation will emerge by relieving people from performing those tedious and monotonous tasks associated with the current ITSM environment? What returns will the organization realize from the use of AI within ITSM? What new business or IT opportunities may be available because of the use of AI within ITSM? A good business case establishes good expectations for the organization regarding AI and ITSM.
  • Begin thinking about how AI can be leveraged by ITSM process designs. As discussed in this recent HBR.org article, AI will bring new capabilities to business (and ITSM) processes. With these new capabilities, organizations will need to rethink what tasks are needed, who will do those tasks, and the frequency that those tasks will be performed. The use of AI will enable organizations to rethink their ITSM processes from an end-to-end perspective, considering what tasks should be performed by people and what tasks should be performed by machines.

The concept of augmenting ITSM with AI is a “no-brainer”.  However, success with AI in an ITSM environment requires a lot of up-front thought, good process design, solid business justification, and considering these three AI truths.

[i] https://2021.ai/ai-governance-impact-on-business-functions

[ii] https://www.techtarget.com/searchcio/definition/digital-strategy

[iii] Ibid.

Share twitterlinkedinmail
AI, ITSM

The 3 Pillars of Success for AI-enabled Service Management

Leave a comment
Share twitterlinkedinmail

In her book[i], Dr. Kavita Ganesan suggests that any AI adoption be evaluated using three pillars:

  • Model success – Is the AI model performing at an acceptable level in development and production? (In other words, the model performs at the required levels of accuracy, execution time, and other factors.)
  • Business success – Is AI meeting organizational objectives?
  • User success – Are users satisfied with the AI solution and perceive it to be a valid solution?

Many organizations are rushing to incorporate AI-enabled technologies to improve their service management capabilities. AI technologies, such as AI-assistants, chatbots, intelligent process automation, generative AI, and more, can provide a next-level set of capabilities for service management. But are these organizations’ service management practices positioned to fully take advantage of these new capabilities?

Let’s be clear – AI is not a “magic wand.”  AI is a technology. And like any other technology, there are factors that must be addressed if an organization is to realize the benefits that AI can bring to service management.

First, AI needs data – and lots of it. The effectiveness of AI depends on the quantity, quality, relevancy, and timeliness of the data being used by the AI models and algorithms. Any limitations in the data being used by AI will be reflected in the outputs produced by AI – and the use of those outputs by service management processes. The old axiom remains true – garbage in will result in garbage out.

AI cannot be a solution looking for a problem. Just because AI is a “hot topic” now doesn’t mean that it is the solution for every business challenge – especially service management issues. If issues like ineffective workflows, undefined services, poorly defined measures, lack of continual improvement practices, or the absence of high-quality data already exist within the service management environment, the introduction of AI will only exasperate those issues.

Lastly, the use of  good organizational change management practices is critical. There is a lot of FUD (Fear, Uncertainty, and Doubt) surrounding the introduction of AI[ii] within organizations. Yes, there will be impacts to how humans work and interact with technology, but for whatever reason, there is a heightened fear associated with AI-adoption within service management.

Applying the 3 pillars for AI success to AI-enabled Service Management

Before rushing into incorporating an AI solution with a service management environment, let’s adapt and apply Ganesan’s three pillars for success with AI-enabled service management.

The first pillar is business success. How do current service management capabilities support business outcomes and enable value realization? How will the introduction of AI capabilities further enhance the realization of the outcomes and value delivered by service management? If the answers to the above questions aren’t clear, revisiting some foundational elements of service management is in order. Consider the following:

  • Have IT services been defined, agreed, documented, and measured in terms of business value, business outcomes, and the costs and risks associated with the delivery of services? Many IT organizations have defined what they call “services” in terms of
    • what goods and products (like laptops and smart devices) are provided
    • the service actions (like password resets) a service desk will perform, and
    • procedures for gaining access to digital resources (like a cloud-based resource or a shared drive).

Not only does this approach inhibit a mutual understanding of the vital role of technology in business success, but it also commoditizes what IT does. Secondly, this approach fails to establish business-oriented measures regarding results and value.

  • Are non-IT colleagues named as service owners? Are these non-IT colleagues actively involved in the delivery and support of services? This is a significant issue for many service management implementations. In many organizations, IT personnel, not non-IT colleagues, have taken on the role of service owner – the person that is accountable for a service meeting its objectives and delivering the required business outcomes and value. The service owner is critical to understanding what is needed and importantly, how business outcomes and value are realized and should be measured.
  • How might AI adoption enable organizations to consider service management practices that would enhance their business? For example, better service portfolio management would enable better utilization of and data-driven investments in services and technology.

The next pillar is employee success. Frequently (and counterintuitively!), service management practices have been designed and implemented with IT and not the IT service consumer in mind. As a result, interacting with the service desk or a self-service portal can be an exercise in frustration due to the over-technical nature of those interactions. Consider:

  • How might the introduction of AI result in friction-free interactions with services and the fulfillment of service requests? How might AI personalize end-user interactions with service management practices? Consider how AI could shift the burden of interacting with service management practices from the end-user to a personalized and proactive AI-enabled capability.
  • How might the introduction of the AI model result in friction-free interactions with supporting IT services? If consuming IT services present challenges to end-users, it can also be challenging for those that deliver and support those services. Will AI-capabilities enable service management practices to shift from a reactive to proactive stance by identifying and eliminating causes of incidents before they occur? Will AI-capabilities enable better issue resolution by suggesting potential solutions to IT technicians?
  • How might the introduction of AI enable employees to make better, data-driven decisions based on relevant, timely, and accurate knowledge? Knowledge management is among the most significant challenges of a service management implementation, as knowledge is ever evolving and continually being created, revised, and applied. AI may provide a solution – this blog explores how Generative AI could provide organizations (not just IT) with the capability of harnessing its collective knowledge.

The final pillar is AI / service management model success. Frankly, many service management challenges can be resolved through continual improvement activities. Some issues may be resolved through the application of effective and efficient automation. Questions to consider include:

  • How might AI adoption result in better and proactive detection and resolution of issues before those issues impact the organization? How might AI adoption result in improved change implementations through better testing or confirmation of positive business results?
  • Is there sufficient, good-quality data to enable AI-driven service management actions? If AI models are not supplied with sufficient, good-quality data, the results from the model will be suboptimal at best – or worse, just flat-out wrong.
  • What is the required level of accuracy for the model? A “100% accurate” model may be too costly to achieve and maintain; a “75% accurate” model may be perceived as a failure.

Get ready for AI-enabled service management

The introduction of AI to a service management environment can be a game-changer on many levels. Here are four steps to get ready:

  • Make the business case for introducing AI to service management. Think strategically about AI , service management, and how the combination of AI and service management will help the organization achieve its mission, vision, and goals.
  • Communicate, communicate, communicate. The mention of AI adoption may cause concerns among employees. Start open conversations regarding AI-enhanced service management capabilities, incorporate feedback, and proactively address concerns.
  • Identify and define success measures. The mere implementation of AI capabilities within service management is not an indicator of success. Define how the benefits articulated in the business case will be captured, measured, and reported.
  • Begin data governance now. The success of any AI initiative depends on the availability of good quality data. If service management is to leverage AI capabilities, the data being captured must be of good quality. Define and publicize data quality standards for service management practices and ensure compliance through periodic audits.

The introduction of good AI capabilities will not fix bad service management. Applying the three pillars described above will ensure successful introduction of AI capabilities resulting in next-level service management practices for any organization.

Is your service management approach “AI-ready”? An assessment by Tedder Consulting will identify any foundational gaps so your service management environment is “AI-ready”.  Contact Tedder Consulting today for more information!

[i] Ganesan, Dr. Kavita. “The Business Case for AI: A Leader’s Guide to AI Strategies, Best Practices & Real-World Applications”.  Opinois Analytics Publishing, 2002.

[ii] https://www.forbes.com/sites/jenniferfolsom/2024/03/28/meet-your-newest-co-worker-ai  Retrieved April 2024.

Share twitterlinkedinmail