SITS26 Recap: Key ITSM, AI, and Service Management Insights from London    

It’s London, baby! The Service Desk and IT Support Show, aka SITS, was held at London Excel last week. SITS is one of my favorite IT service management (ITSM) events, and SITS26 lived up to this. There’s something for everyone, seeing all the latest tech, getting an overview of what’s going on in best-practice land, and catching up with colleagues and friends. This article is about what I learned from this year’s show. 

SITS26 Session: Get Your Data Right Before You Look at AI – Ashwin Ram Ragurthi of ManageEngine

Ashwin began his SITS26 session by discussing the energy surrounding artificial intelligence (AI). He talked about Anara’s Law, which says we tend to overestimate the effect of tech in the short run and underestimate it in the long run, and that’s very appropriate to AIOps. We all want to run AI and get all the results, but we can’t just flip the switch and expect everything to change. We need to understand the tech so that, in the long run, it delivers results.

When looking to implement AI, Aswin suggested these four steps:

1. Sort your data and ensure you have proper categories
2. Look at resolution quality – all knowledge bases should be detailed
3. Metadata enrichment – the process of adding descriptive, contextual, and structured details to existing data or content to make it easier to find, understand, and use.
4. Structured data hygiene – getting your house in order so that AI doesn’t amplify missing or incorrect data.

SITS26 Session: We Need to Change the Way We Do Things – Pat Bolger, Hornbill and  Emily Johnson and Alex Rutter, Warwickshire County Council

Pat started this SITS26 session by discussing the challenges many public sector organizations face. These included:

• The world is changing, and incremental improvements are not as effective as we need them to be in the public sector
• Demands are increasing while budgets are decreasing
• ITSM is no longer a local position; we need to work with a complex service landscape
• Increased focus on efficiency targets – getting the drive for automation and looking at how to reduce demand
• The need to look at how to deal with demand up front and how to reduce it
• We expect AI to assist us, but technology is struggling to keep up – most people have a  better experience at home than at work
• Organizations won’t keep people if we don’t give them a good experience
• If your data’s not in a good state, you’re going to be in a world of trouble
• The problem isn’t people’s effort – it’s the system
• We need to look at value flowing rather than whether people are busy

Pat continued by referencing John Seddon’s research on systems thinking, which provides a set of tools to focus on how work flows through an organization. This enables organizations to study and understand their IT landscape because knowing, not guessing, solves the world’s problems. In the public sector, where organizations struggle with low budgets, high staff turnover,  legacy systems, hybrid environments, and security liabilities, this requires focus. We need a way to settle up, and you do that by getting feedback and iterating your approach based on customer needs.

Emily and Alex continued by talking about Warwickshire County Council’s experience. The organization, with 5,000+ employees across a large geographical area, had to build a process that supported everyone. As well as the industry-specific challenges, Warwickshire also had to deal with:

• Siloed operations.
• Duplicated efforts and tacit knowledge
• A lack of joined-up systems – making it harder to deliver outcomes consistently and at pace.

The goal was to improve delivery and employee experience, and they used systems thinking as a guardrail. They supported a big-picture perspective by digging into the details because you only succeed by bringing your people on the journey.

The aim was to get the right fix, not the quick one, so there were no surprises later. The team used the systems thinking approach to:

• Cultivate strong stakeholder support – exec support was critical, so directors attended workshops and participated with operational colleagues
• Create cultural shift – working with people, not doing things to them
• Have dedicated engagement of teams and service areas
• Not work to assumptions
• Seek absolute clarity
• Not automate broken processes
• Increase team belief by having the whole team work to deliver benefits to this area.

The results included:

• A 92% reduction in onboarding-related emails
• Reduced staffing pressures
• Software asset management (SAM) costs under control
• Building on successes – including that automated provisioning will save 5,000 hours per year, with continual improvement, the two-year forecast is 10,000 hours
• Unified enterprise service management (ESM) – extending ITSM skills across the organization.

Key takeaways:

• Systems thinking – understand a system and the people who use it
• Customer centric
• Global versus local optimization
• Invest in discovery
• Exec sponsorship
• Design end-to-end, not in silos
• Focus on experience and outcomes
• Reality in key performance indicators (KPIs) – a global, not local, approach to remove conflicting KPIs
• Get the process right first – the automation comes next.

SITS26 Session: We Need to Start Thinking of AI Agents As Part of Your Team – Not the Competition – Vijay Rayapati, Atomicwork

AI has evolved from functional models to reasoning models to AI agents. But here’s the thing: AI agents are not the finish line. We need a blended approach that enables AI agents to support their human counterparts. Look at it this way: AI coworkers are multiplying the workforce across the business.

Instead of simply building AI workflows, organizations are now moving toward deploying an AI workforce comprising AI agents and AI coworkers, powered by functional and reasoning models. What makes an AI workforce possible is the ability to package AI into role-specific “workers” with defined goals, permissions, memory, tools, and guardrails through agent orchestration. These AI coworkers can operate safely within sandboxed environments, use plug-in skills to perform a wide range of tasks, and integrate securely with enterprise systems through growing standards such as MCP and modern access controls.

As a result, organizations are beginning to manage AI in much the same way they manage human teams – defining roles, responsibilities, permissions, performance expectations, and governance. What does this mean for the organization? We’ve gone from IT buying tools and building workflows to IT running the AI workforce.

The final part of Vijay’s session was about how to get started. His top tips for SITS26 attendees were:

• Inventory the jobs, not the tasks – so think Access Management over just “reset all the passwords”
• Pick one role to start with
• Define “what good looks like”
• Name the human manager – where does it go for approval?
• Don’t build AI workflows – deploy the AI workforce.

We Need to Be Better at Understanding Change Impact – Paula Maattanen, GuideVision

Paula started her SITS26 presentation by asking, “What is change?” ITIL defines it as the addition, modification, or removal of anything that could directly impact products or services. Real-life examples include new products, changes in vendors, or the introduction of completely new services.

Paula talked about improving change outcomes by learning from our mistakes. Some of the examples she included were:

• Coke changed its formula in 1985. It was thought this would improve the drink’s taste, but it didn’t account for brand loyalty, and the original version was brought back seven months later.
• The prohibition era. The US government banned booze, but what was the impact? Illegal trade and bars, which actually lead to more criminality and antisocial behaviors.
• Kodak thought people would continue to buy film and print out pictures rather than go to digital cameras.
• Nokia didn’t believe the touchscreen would work, so it lost market share to Apple.

What do these failures have in common? Recurring patterns of overconfidence and poor planning. No risk analysis or user acceptance testing, and ignoring expert advice. Lack of flexibility once problems appear. Plus, my personal favorite: underestimating human behavior. 

The next area Paula looked at was ITSM tool implementation. Beyond time and cost, it’s important to measure what really matters: user experience, employee adoption, service quality, and unintended effects. If you don’t ask enough questions, you won’t get the full user experience. Happy employees result in happy customers.

When considering change, we need to look at the following:

• Business outcomes. When we make the change, what impact will it have, and will we be providing value?
• Define the outcomes – what are we trying to achieve and how do we measure what does good look like?
• People experience and adoption – because if employees find it difficult to use the system, they will use shadow IT and circumvent our IT service desk. The very things we are trying to avoid. 
• Process effectiveness – moving from old processes to redesigned practices.
• Customer and stakeholder impact – because the aim should always be to simplify workflows and reduce complexity.
• Preventing unintended consequences – disruption, resistance, overload, and trade-offs.
• Our change deployment and what best suits our risk appetite? Iterative or big bang? If we do iterative changes, there’s less risk and more time to learn and improve. If we deploy using the big-bang model, we can deploy many changes that are more efficient but also much riskier if things go wrong.

The last part of Paula’s SITS26 session looked at how change relates to the ITIL four dimensions of product and service management:

• Organizations and people – it is always easier for people to have small changes unless they are truly motivated by the benefits.
• Information and technology – consider our rollout approaches. Iterative changes in a DevOps/Agile way yield small, incremental improvements. Alternatively, if everything needs to happen at the same time, go with the big-bang approach.
• Partners and suppliers – most use big bang because, let’s face it, there’s no small way to switch from a major partner.
• Value streams and processes – when making changes, you need to figure out all aspects. Lean into your company culture – political, legal, environmental, social, and technology factors.

Change or chance – it’s up to us. Are these change opportunities, or are they the end of the world for us?

SITS26 Session: We Still Need Structure to Do Asset Management Effectively – Jason Hamer and Sean Auckland, Vodafone

Jason opened this SITS26 presentation by talking about the challenges of managing IT assets without service models. These included:

• The fragmented systems impact
• The lack of end-to-end visibility 
• Governance, risk, and compliance (GRC) gaps
• Operational inefficiencies
• A lack of standards, not understanding business impact, and higher resolution times
• Service modeling – the steps seem simple (business, service and technical layers), but everyone calls their areas a service, and you need a step to define what all these areas need – especially with a business the size of Vodafone.

Jason continued by discussing one of my favorite ITIL principles: start where you are, don’t reinvent the wheel. If you have solid infrastructure data, that’s good, but there can be lots of confusion when addressing service layers for example:

• Technical layer – physical and cloud infrastructure
• Functional layer – logical IT systems that sit on tech services
• Service layer – what the end-user sees and what they actually use these services for. It’s great to understand what the business is using IT for, but we don’t always understand the customer journey.

The next part of the session focused on capabilities. For each IT system, the team reviewed the capabilities it delivered. The target benefits were:

• Operational benefits – faster resolution and root-cause analysis (RCA), stronger GRC posture, smarter decisions and cost control, and stronger customer focus.
• Personal benefits – one example was change management: how do we correctly quantify risk? Understanding the end-to-end system will give you that clarity.
• Service modeling and enhancing observability – they wanted to look at AIOps.

Other areas that were looked at included:

• How failure events normally create incidents.
• How having an end-to-end service model gave Vodafone the ability to drive events into a singular ticket situation, akin to a more swarm-like collaboration
• Allowing AI OPs to understand the impact on the business rather than just the technical components.

Sean then shared how Vodafone used BMC Helix for their service modeling journey, leveraging Helix for ITSM and Discovery to create blueprints and templates that standardized how IT infrastructure and services were discovered and modeled.

The challenge was that assets were spread across other areas and the configuration management database (CMDB), and a different tool was used for SAM, desktops, and cloud assets. This caused fragmented data and governance gaps. Sean talked about how the team needed to federate all the information for improved quality and visibility. The next step was to pull data from federated sources into a single data lake and visualize that GRC requirements could be met and the minimal viable product (MVP) delivered.

The results were high-quality data that was secure by design, and the establishment of a solid foundation with trusted data that improved support for GRC, increased visibility and control, and enabled collaboration and business alignment.

SITS26 Session: ITSM, DEX, and Discovery Work Brilliantly Together – Alex Ares, Priory Group

Alex’s SITS26 session was about unifying ITSM, digital employee experience (DEX), and discovery to achieve better business outcomes. He discussed the challenge of modernization –  the organization had grown over ten years, and people were working really hard, but things were fragmented. Priory Group wanted to redevelop people, processes and the tool so the team could evolve toward their goals.

Healthcare faces specific challenges, including patient ratios and staff safety. Providing care is everything. Alex noted that having a joined-up approach is essential – service management cannot operate in a silo if it wants to solve problems and improve patient care.

The final part of the session was a discussion of the best ways to put people at the center of everything. One of my favorite takeaways was how Priory Group used its service desk as a sort of resource bench, so if a space opens up in a more technical team, that gap can be filled by a member of the desk, as they’ve worked closely enough together that they’ve already been trained for the role. This keeps talent in-house and supports career progression – everyone wins.

SITS26 Session: We Need to Make AI More Human Centric – Mark Basham, Hit Global

Mark’s SITS26 session was all about humanizing AI and what happens when experience debt meets AI. He reflected on how, while he wouldn’t change ITIL 4, he would look at human experience, especially through the lens of hidden friction, behavioral impact, environmental impact, and how services are actually experienced.

Mark continued by noting that we’ve never been trained to deliver customer experience and asked, “Just why do so many AI projects struggle?” The answer is normally a combination of unclear problem definitions, poor data quality, lack of change management, over-reliance on tech not designed for humans, and most AI not delivering the expected outcomes.

Enter Human Centered Design (HCD) – putting people at the heart of the process, which creates meaningful and useful solutions. Behind every experience is technology, and he used a Multi-Factor Authentication (MFA) example to explain this. By and large, end-users don’t love MFA. It’s important, but it’s a lot. The reality is, if you’re at home and understand the tech, it’s OK. But what happens if you’re a front-line worker with a huge queue of people to look after? Not so much.

We need to balance IT’s ability to deliver and support services against customer experience. Combining ITSM with HCD is not a perfect fit because, while ITSM has to operate within constraints, HCD prefers to ignore them and focus on the best possible outcomes. What does balance look like? The double diamond framework: maintain and run, then review with the business.

The final part of the presentation looked at how experience is the new currency in ITSM. IT leaders are now demanding services that are user-focused and experience-driven. Hence, we need to have user experience (UX) front and center. People think AI will fix everything, but let’s look at our CMDBs and our service catalogs in the real world. Are they perfect? Probably not. AI doesn’t fix the system. It amplifies it, and we need to put time and effort into putting our house in order.

Humanising AI starts by redesigning the system for people, then AI can scale what really matters. We need to view services through the lens of users and speak in business terms to make real change.

So that was my take on SITS28. What did you think? Please let me know in the comments.

Further Reading

Vawns Murphy

Senior ITSM Consultant at i3Works

Vawns Murphy holds qualifications in ITIL V2 Manager (red badge) and ITIL V3 Expert (purple badge), and also has an SDI Managers certificate. Plus she holds further qualifications in COBIT, ISO 20000, SAM, PRINCE2, and Microsoft. In addition, she is an author of itSMF UK collateral on Service Transition, Software Asset Management, Problem Management and the "How to do CCRM" book. She was also a reviewer for the Service Transition ITIL 3 2011 publication.

In addition to her day job as a Senior ITSM Consultant at i3Works, she is also an Associate Analyst at ITSM.tools.