AI readiness lessons
AI readiness and high performance: our first Showoff Session
On the 9th of July we did something Showoff had never done before.
We put a room full of Connacht business leaders together at Dexcom Stadium, ran a few sharp sessions on getting a business ready for AI, and finished with one of the most respected voices in team-building in world rugby. No product demo. No countdown to a launch. Just an honest conversation about two things that turn out to be the same thing: what it takes to make AI work in a real business, and what it takes to build a team that performs.
This was the first of what we're calling Showoff Sessions. Here's what came out of the room - from the AI readiness framework our Chief Architect walked everyone through, to Stuart Lancaster's keynote on building teams - and why the line between a well-run business and elite sport is far shorter than it sounds.
Why we ran it - and why in a stadium
Most business events follow the same script: a vendor on a stage, a slide deck counting down to a product, and a room half-listening. We wanted the opposite.
So we kept it small - a select group of leaders from across Connacht — and we picked a setting that made the theme obvious. A stadium is a high-performance environment by design. It felt like the right place to talk about how the habits that build a strong company are the same ones that win on the pitch.
The format was simple: a run of practical business sessions, a keynote from Stuart Lancaster, and a panel to finish - with plenty of time afterwards to talk properly. We ended the evening the way you should end anything in Galway, with a walk down Shop Street and good company.
Getting ready for AI
Our CEO Shane Byrne opened the evening — a warm welcome to the room, a quick overview of who Showoff are and what we do, and then he handed over to the first session.
First up was our own Chief Architect, John Kilbride, who took the room through a framework for AI readiness that he was careful to flag up front was not a sales pitch. Judging by the questions afterwards, it landed as exactly that: useful, not salesy.
He opened with a number that tends to stop people. Back in 2024, Gartner forecast that at least 30% of generative AI projects would be abandoned after proof of concept — before anyone saw any value. We're now past that date, and the real figure has come in closer to half. Things have got worse, not better.
His point wasn't that AI doesn't work. It was simpler than that:
Most AI projects fail before the technology is even the problem.
Dirty data. Unclear ownership. Nobody trusting the output, so they quietly go back to the spreadsheet and adoption dies in silence. None of those are AI problems - they're readiness problems.
He made it concrete with two stories the room won't forget. A North American airline's website chatbot once told a grieving customer he could claim a bereavement fare retrospectively - a policy that didn't exist. The bot invented it, and it ended up being argued in front of a tribunal. And an international fast food chain, working with a global system integrator, spent three years testing AI drive-thru ordering across more than 100 US locations before pulling the plug in 2024, after the system started adding nine sweet teas to orders and putting bacon on ice cream. Both looked like AI failures. Both were really readiness failures - bad data in one, the wrong process picked first in the other.
The four pillars of AI readiness
John's framework is refreshingly un-flashy. Readiness comes down to four things - data, process, people and platform - and the honest work is in the order you tackle them.
- Data: Ready data is complete, consistent and connected. His test for the room: if you asked your sales director and your finance director the same question about last quarter's revenue, would you get the same number? In most businesses you wouldn't - and if two people can't agree on last quarter, AI has no chance of acting on it reliably.
- Process: AI amplifies whatever's already there. Automate a broken process and you just break things faster, at scale. Start with the boring, high-volume, low-judgement tasks - not the ambitious headline one.
- People: Someone has to own the outcome, trust the output, and be able to switch it off when it goes wrong. Change management isn't an afterthought - it is the project.
- Platform: This is where Salesforce, Agentforce and Data Cloud sit, and where hidden technical debt surfaces — the custom code and the five-year-old integration that were fine until you tried to build AI on top of them.
The part that got the room quiet was the self-assessment. Score yourself one to five on each pillar, he said - and don't fixate on the total. Your weakest pillar sets the priority. Not your ambition, not your budget. The weakest foundation decides what to fix first.
Salesforce: the model isn't the hard part
Darragh Madden from Salesforce picked up the thread with the company's view of the “agentic enterprise”, and it dovetailed neatly with John's. The argument, in short: almost everyone now has access to capable AI models - that's not where the advantage is. The hard part is making those models work in the context of a real business, and that takes more than a model. It takes the systems around it.
The framing is that the agentic enterprise runs on connected systems rather than one clever model: a system of context that unifies your data, a system of insight (Tableau) that turns it into intelligence, your core system of record where work actually happens, a system of engagement (Slack) where people and agents meet, and a system of agency (Agentforce) that orchestrates it all. Different language, but the same point John had just made - the model is the easy bit, and the foundations underneath it are the work.
The finance lens: what a bank is watching
Denis Ryan from AIB's technology sector team brought the view from the other side of the table. His team partners with tech companies across their whole life - from scaling SMEs through to established players - and the framing that stuck with me was that a bank's role changes as you grow. Early on it's the essentials; through the growth phase it's working capital, trade finance and senior debt; and later, when you're scaling hard or eyeing an exit, it moves to venture debt, development capital and advisory on fundraising, M&A or IPO.
The AI thread ran through his section too. His own organisation runs its mission-critical services at 99.99% availability and has stood up a dedicated resilience function to take a more predictive, intelligence-led approach to cyber and fraud. It was a grounded reminder that, in a regulated business, AI has to stand up to the same scrutiny as any other decision - which is really the same readiness point, seen through a risk lens.
Stuart Lancaster on building teams that last
After the business sessions, the evening built to its keynote. It came from Stuart Lancaster - now head coach at Connacht, and one of the most respected voices on team-building in world rugby. If you only know the headline of his career, you know the hard part too: he was head coach of England into a home World Cup in 2015, and that tournament ended badly and publicly. What he did next is the more interesting story. He rebuilt, as a coach at Leinster, and turned a period most people would rather forget into the foundation of everything he's known for now.
That arc - a farming background, teaching, the academy, England, a very public failure, and a rebuild - was the setup for a deceptively simple question he put to the room: what have I actually learned?
Culture is the leader's job
His first point was blunt. Your most important job as a leader is to drive the culture - the environment. He reckons roughly 70% of behaviour in a team is set by the environment it sits in, so if you want different behaviour, fix the environment first. And culture doesn't hold itself up - it drifts the moment you stop actively sustaining it. In his words, it's not a book or a lecture that makes the difference. It's the leader.
The shift he kept returning to was from “prescribe and control” to “acknowledge, create and empower” - moving a team from being directed to genuinely owning the thing. Anyone who has tried to move a business from command-and-control to something more grown-up will know how hard, and how worth it, that is.
It starts with trust
He walked through the model that underpins long-term high-performing teams, and it's one plenty of business leaders will recognise: trust at the base, and then — only once that's real — the ability to have constructive conflict, genuine commitment, real accountability, and finally attention to results. Skip the trust and the rest is hollow.
It was the same failure mode John had put to the room earlier in the evening, just on the data side: without trust in the output, people quietly disengage and go back to what they know. Trust at the base, or nothing holds. Same lesson, different pitch.
The part that landed hardest was on identity. People need to belong - to know what their “tribe” is - and a team's values are simply an expression of that identity. When it's missing, you get what he called the identity vacuum, and the “silent dance” that comes with it: quiet, tense meetings on the surface, and a lot of unspoken activity in cliques underneath. Most of the room had clearly sat in that meeting at some point.
Five lessons that travel
He closed with five he'd take into any team, in any field:
- Build a culture of growth from within.
- Develop leadership throughout the organisation - help people find their voice.
- Understand the “performance clock” - sustained high performance demands change, not comfort.
- Always be learning, especially from failure.
- Take the lessons and move on - “SUMO”, as he put it.
And a last line that needs no translation into business: there is no better place to be than on a winning team.
The panel that closed the night
We finished with a panel - Stuart joined by former Ireland international Ian Madigan, our own CEO Shane Byrne, and former Kilkenny hurler Richie Hogan. Three very different sporting backgrounds, one shared theme. Ian in particular is on an interesting road: an international rugby career behind him and a career in enterprise software ahead of him. That shift from elite sport into business is a good reminder that the skills that matter most travel well - discipline, coachability, and the ability to perform when it counts. None of that is unique to sport. It's just more visible there.
What we're taking from the first Session
If there was one thread running through the whole evening, it was this: high performance isn't a trick. Whether you're getting a business ready for AI or rebuilding a rugby side, it comes down to the same fundamentals - trust first, an honest look at where you stand, solid foundations, and the discipline to keep doing the unglamorous things well. John said it about data; Stuart said it about culture. It's the same sentence.
For us, the data version of that is the part we can help with. The path to AI is paved with data, as we tend to put it - and the teams that get real value from it over the next few years won't be the ones with the flashiest tools. They'll be the ones who did the groundwork first.
A grounded start
Our first Showoff Session did exactly what we hoped it would. It put good people in a room, kept the vendor noise out, and got to something real about what high performance takes - in business and in sport.
The two worlds rhyme more than we tend to admit. Both reward trust, clear roles, honest self-assessment, strong foundations, and a refusal to skip the boring work. In our world, that boring work is your data - and it's the single biggest thing standing between most businesses and the AI outcomes they're being promised.
Thank you to Stuart, to all our speakers, to everyone who came, and to Galway for having us. This was the first Session. It won't be the last.
Ready to do the groundwork?
If John's framework struck a nerve - and judging by the room, it did for a few people - that's the exact work we do. Showoff runs an AI Readiness assessment before any build: a workshop with your team, an org health check to find the technical debt before it bites, and a prioritised roadmap that starts with your weakest pillar. Or start with his one honest question - what's the one process in your business that would change everything if it ran itself? - and get in touch.