AI transformation is everywhere, in boardroom conversations, strategy decks, and technology roadmaps. Yet for all the excitement, many organisations struggle to answer a simple question: what value is AI actually delivering?
That’s where Conny Ploth, VP of Global AI Transformation, brings a refreshingly grounded perspective. In her conversation with Dr. Raoul-Gabriel Urma on Data & AI Mastery, Conny makes one thing clear: if you can’t measure it, don’t start it.
Her experience across financial services, consulting, and impact-driven organisations has shaped a pragmatic approach to AI; one that prioritises people, measurement, and long-term value over hype.
One of Conny’s strongest messages is that AI initiatives often fail before they even begin, because teams skip the basics.
Too many organisations jump straight into pilots and proofs of concept without establishing baselines. Without knowing how a process performs today, it’s impossible to prove that AI has improved it tomorrow.
Measurement isn’t about slowing innovation; it’s about making impact visible. Clear metrics allow leaders to assess progress, justify investment, and decide where to double down or where to stop.
As Conny puts it, transformation doesn’t come from doing more AI projects. It comes from doing the right ones, and knowing whether they work.
Like many previous guests, Conny challenges the idea that AI is primarily a technology problem.
In reality, AI transformation is a mindset and cultural shift. New tools alone don’t change outcomes unless people trust them, understand them, and integrate them into daily work.
That’s why adoption is one of the most overlooked and most important KPIs. You can build technically brilliant models, but if they’re not used, they don’t create value.
Successful organisations focus as much on change management as they do on algorithms: communication, experimentation, and feedback loops that help people feel confident using AI in real decisions.
To cut through complexity, Conny frames AI value around three core levers:
Automation of repetitive tasks can free up time and reduce operational friction. These wins are often the fastest to deliver but only if processes are already well understood.
AI can enable better decision-making, new products, and more personalised customer experiences. These use cases take longer to mature, but often deliver the greatest strategic value.
From faster responses to more relevant services, AI can improve how customers interact with organisations, provided it’s designed with real user needs in mind.
The key is balance: combining quick wins that build momentum with longer-term bets that transform how the business operates.
Another recurring theme is the importance of data quality and governance.
AI systems are only as good as the data they’re built on. Messy, fragmented data leads to unreliable outcomes and in regulated industries, that’s a risk leaders can’t afford.
Conny stresses the need for clear guardrails: knowing who can access what data, for what purpose, and under which controls. Strong governance doesn’t block innovation, it enables it safely.
This becomes even more critical as generative AI lowers the barrier to experimentation. Without clear boundaries, organisations risk scaling confusion rather than value.
AI transformation can’t be delegated entirely to innovation teams nor can it be dictated solely from the top.
Conny advocates a dual approach:
Hackathons, sandboxes, and community-driven learning help employees engage emotionally with AI, not just intellectually. These experiences often reveal the most impactful use cases, grounded in everyday work.
At the same time, leadership must make hard choices. Endless pilots dilute focus. Sustainable transformation comes from fewer, bigger bets that align with business strategy.
Legacy organisations often compare themselves unfavourably to AI-native startups. But Conny offers a more optimistic view.
While startups move fast, large organisations have scale, data, and deep domain expertise. The challenge is creating safe spaces to experiment, environments where people can test ideas without fear of failure.
When learning is experiential rather than theoretical, adoption accelerates. Upskilling becomes less about formal training and more about doing, reflecting, and improving.
The biggest misconception Conny encounters is the belief that AI will magically fix broken processes. In reality, AI amplifies whatever already exists, good or bad.
That’s why transformation must start with clarity:
When those questions are answered upfront, AI becomes a powerful enabler rather than an expensive distraction.
AI transformation isn’t about chasing the latest model. It’s about building organisations that can learn, measure, and adapt continuously.
As Conny Ploth’s experience shows, the organisations that succeed are those that treat AI not as a technology project, but as a measured, people-centred journey toward impact.
_________________
At Cambridge Spark, we help organisations build the data and AI fluency needed to move from experimentation to measurable value, combining technical capability with leadership, culture, and mindset.
👉 Explore how we support AI transformation at cambridgespark.com