For more than two decades, Paul Wilkinson has been running a simple exercise at IT service management (ITSM) conferences. He takes the five most commonly chosen ABC cards, the playing cards he developed to surface the human and organizational failures that derail implementations, shuffles them, and asks the audience to rank them in order of relevance. The top card, year after year, shows an IT person sprinting toward a rocket with a floppy disc in hand to deliver navigation software updates. The rocket has already launched.
The card represents a failure to understand business impact and priority. It’s been the most chosen card globally for more than 20 years in a row. People have been asking Paul to update the cartoon and remove the floppy disc, because nobody under 30 knows what one is. He has refused. When he retired four years ago, it was still number one. It still is.
This is Paul’s argument in the second episode of Roman Jouravlev’s Conversations with Giants series. Not that the industry is getting things wrong, but that it’s getting the same things wrong, repeatedly, and that no amount of new frameworks or certifications or technology has changed the underlying pattern. His evidence for this is the ABC cards. His sharpest illustration of it is what happened when he renamed his presentation.
The “ITSM failure” evidence: 20 years of the same cards
The ABC cards stand for attitude, behavior, and culture, and started as cartoons. Paul began drawing them during early ITIL consulting work to capture the failures he kept encountering across organizations: the lack of management commitment, the inability to define service in terms of value, the tendency for every user to classify their request as highest priority. He used them in client presentations and internally at Pink Elephant, where he was a consultant.
The ABC framework itself was identified by someone else. Paul was presenting the cartoons at a conference when an audience member pointed out how cleanly they mapped onto attitude, behavior, and culture as three distinct categories. Paul looked back at the material and realized the person was right. He organized the 52 cartoons by suit: clubs for attitude, diamonds for behavior, spades for culture, hearts for stakeholder types.
What has made the cards useful as a diagnostic tool is precisely the fact that they’re not a questionnaire. Paul’s approach with a skeptical CIO was to put the top five globally chosen cards on the table face up and ask which order they’d put them in. The CIO would rank them based on their organization’s reality. The resulting conversation was more honest than any direct question about issues would have been.
The survey data that sits behind the cards tells the same story as the card choices. Paul analyzed major industry reports, including McKinsey, Gartner, Forrester, State of Agile, and State of DevOps, and identified five capability areas that consistently explain why major transformations fail: strategic fit, leadership, culture, skills, and continual improvement. He built a 13-question survey around those five areas and ran it in 2021. Between 60 and 70 percent of respondents scored poor or weak across all five. He ran it again in 2025. The results hadn’t meaningfully changed.
The demonstration: the shiny new thing
In 2021, Paul decided to do a final circuit of conferences before retiring and submitted an ABC presentation. Organizers told him they’d done ABC ten years ago. Did he have anything new? Something on artificial intelligence (AI), big data, robotics?
He renamed the “ABC failure” presentation The Shiny New Thing That Really Helps and announced it on LinkedIn. The post received five times the engagement of anything else he’d published. Every conference that had previously turned him down invited him to speak. He gave the same presentation he’d been giving for a decade, with the same slides.
Before launching it, he wrote an article predicting what would happen next: tool vendors would start claiming their products were shiny new thing compliant, and experienced practitioners would ask for the badge without attending the course. Both predictions were correct. He started receiving LinkedIn messages asking how to obtain the badge. He only gave it to people he felt had genuinely earned it, then developed a companion badge, the Rusty Ignore, for organizations whose survey results demonstrated they were still doing exactly what the name suggests.
The point isn’t that people are gullible. It’s that the pull toward the new framework or technology is strong enough to override the evidence, and the evidence has been consistent enough, for long enough, that this starts to look like a structural feature of the industry rather than a series of individual mistakes.
The failure parallel: AI
Organizations are demanding AI now, in the same way they once demanded ITIL, without defining what success will look like in six months or asking whether the investment maps to value, outcomes, costs, and risks. The competitive pressure is the same: a competitor has it, therefore, we must have it immediately. A 2025 MIT study found that 95 percent of AI adoptions aren’t delivering the expected value, with the failure reasons clustering around strategic fit, leadership, culture, skills, and continual improvement, which are the same five areas Paul’s shiny new thing survey has been measuring for years.
The floppy-disc card is still the most chosen globally. The survey results haven’t shifted in four years. The ABC presentation got more traction when it was renamed than it ever did under its own name. These aren’t separate observations. They’re the same observation.
The games: making the argument experiential
The simulation games Paul developed with Jan Schilt, beginning with Apollo 13: An ITSM Case Experience and expanded to Challenge of Egypt, 2020, Grab a Pizza, and Marslander, built on the same premise: that people understand and retain things differently when they experience the consequences of their decisions rather than having the theory explained to them.
The Apollo 13 game came from a keynote at a Pink Elephant conference in America. James Lovell, the commander of the Apollo 13 mission, was speaking to a room of 2,000 ITIL-certified professionals. The room was completely silent. As Paul listened, he found himself mentally mapping every element of the story onto ITSM practice. Problem management, availability management, configuration management, change management. The crisis that nearly killed the Apollo 13 crew was, in structural terms, a service management failure and a recovery.
He began building cards: configuration cards for the rocket, incident cards, problem cards. He had no game engine, just material. The structure came from Schilt, who had extensive experience turning concepts into multi-round game dynamics. They prototyped Apollo 13 at an ITIL conference in the Netherlands. After that, it ran at 10 to 15 locations worldwide every week at its peak. NASA itself played it.
Paul’s frustration is consistent with the broader argument. When simulation games were offered as an additional day alongside foundation training, organizations frequently declined, preferring to use the time on exam preparation instead. The certificate mattered more than what participants could take back and apply. And when CIOs did commission the games, they often didn’t return at the end of the day to review the list of identified improvements and commit to acting on any of them. People had spent a day identifying what needed to change. Nobody with the authority to allocate resources showed up to say what would happen next.
The games were designed to develop judgment, collaboration, and the ability to translate learning into practice. They were often used instead as team-building exercises or as a box to tick at a motivational event. This is the same pattern in a different form.
The real work of behavior change in addressing “ITSM failure”
Paul’s view on what the industry consistently underestimates is straightforward: behavior change is the hardest thing there is, and almost nobody treats it that way. Frameworks, processes, certifications, and tools are all means to an end: getting people to do things differently. That end requires time, patience, iteration, failure, coaching, and feedback. It’s not achievable through a project plan or a go-live date.
The World Economic Forum’s Future of Jobs Report 2025 identifies analytical thinking, creative thinking, resilience, and collaboration as critical skills for 2030. Paul’s point is that these are the same skills the industry was identifying as critical 20 or 30 years ago. This is either evidence that the industry has consistently known what matters and consistently failed to act on it, or that the skills are genuinely difficult to develop and an effective approach hasn’t been found. Probably both.
The story Paul finds most instructive came from a CIO at a large telecom organization who attended one of ten Apollo 13 sessions run for their organization. They were the most critical person in the room, skeptical of the exercise’s value and the relevance of talking to users. Paul asked them when they’d last spoken directly to a user, not the business, a user. The answer was that they had not.
About six months later, the CIO called to say that the session had saved the company two million euros. The CFO had wanted to spend that amount on a new CRM system. The CIO went and spoke to the users of the existing system first. Everything they said they needed from a new system was already in the old one. Nobody had trained them to use those features. They sat with them and showed them. The two million euros stayed where it was.
The CIO had been the hardest person in the room to reach. They’d gone away, applied one specific thing, and it had made a material difference. The learning had landed despite the resistance, which is perhaps the most honest version of what behavior change can look like.
What to do with this
Paul’s argument isn’t just a diagnosis. The evidence points to a consistent set of things organizations get wrong, which also points to what needs to change.
Define success before you start. Whether the initiative is an ITIL adoption, an experience level agreement (XLA) program, or an AI deployment, the failure to define measurable outcomes upfront is the single most reliable predictor of a failed investment. Value, outcomes, costs, and risks need to be articulated in concrete terms before anyone signs off on anything.
Treat the five capability areas as a checklist, not an afterthought
Strategic fit, leadership, culture, skills, and continual improvement aren’t soft factors. They’re the factors that determine whether any technical or process change sticks. If leadership isn’t genuinely committed, if the cultural environment penalizes failure rather than learning from it, if skills development is treated as an optional extra, the implementation will fail regardless of how good the framework is.
Make leadership follow-through visible
One of Paul’s specific frustrations with the simulation games was that CIOs would commission a day of intensive learning, compile a list of identified improvements, and then not return to say what would happen to them. If the people with authority to allocate resources don’t demonstrate commitment where it matters, the signal to everyone else is that none of this is really serious.
Use tools that open conversations rather than close them
The ABC cards work because they’re not a questionnaire. They’re a mirror. Any diagnostic approach that gets people talking honestly about what’s happening in their organization is more valuable than one that produces a score and a report.
Stop treating the certificate as the goal
Paul’s sharpest observation is that the moment passing the exam becomes more important than applying the learning, the whole exercise becomes self-defeating. This applies to ITIL, AI training programs, and any formal learning pathway. The question isn’t whether someone passed. It’s what they changed when they got back to their desk.
ITSM failure: Will things change?
The floppy-disc card has been the most chosen card globally for over 20 years. Until organizations stop running toward the rocket after it’s already launched, it probably will be for another 20.
Sophie Danby
Sophie is a freelance ITSM marketing consultant, helping ITSM solution vendors to develop and implement effective marketing strategies.
She covers both traditional areas of marketing (such as advertising, trade shows, and events) and digital marketing (such as video, social media, and email marketing). She is also a trained editor.
