Data and analytics: are you asking the right questions?

The foundation of successful AI implementation lies in the quality of questions we ask and the data we feed them, not the sophistication of the algorithms.

The journey from scepticism to understanding

Brian’s reflection on his journey from scepticism to understanding felt very familiar:

“My path into AI began around 2012, well before the current wave of generative AI captured public attention. At that time, machine learning felt limited – impressive for number crunching, but constrained by data availability and narrow applications. The breakthrough moment came when I encountered early discussions about systems that could engage conversationally, concepts that would eventually manifest as the generative AI tools we know today.

“When ChatGPT launched, those first few weeks were genuinely unsettling. The realisation that this system was demonstrably more capable than me in many areas forced a reckoning with assumptions about expertise and value. Rather than retreating into pessimism, I chose to explore what this meant for the future of work and how individuals and organisations could adapt. That exploration became a four-year learning journey, ultimately crystallising into The AI Lead and a focus on helping others navigate this transformation".

The challenge of data drag

While working with organisations across various sectors, one pattern consistently emerges… the technology exists to do remarkable things, but the foundational elements – processes, information architecture and mindset – lag significantly behind. Brian calls this phenomenon ‘data drag’ – the friction created by unknown data, inadequate processes and lack of contextual information that prevents AI from delivering its potential value.

Data behaves like air in most organisations. Nobody truly owns it, everyone complains when it’s unavailable and, because it’s theoretically everyone's responsibility, it becomes nobody’s priority.

This creates a particularly challenging dynamic when attempting to implement AI solutions that depend entirely on quality data inputs. No data, no AI. It's that straightforward.

The irony is stark. Executives are eager to invest in AI transformation and deploy thousands of agents across their organisations, but they haven’t established what those agents should accomplish or what value they might create.

Meanwhile, shadow AI usage proliferates throughout these same organisations. Just as shadow IT emerged when official technology solutions couldn’t meet user needs, employees are incorporating AI tools into their workflows regardless of official policy.

The question isn’t whether people are using AI – they are. The question is whether organisations are providing the infrastructure and guidance to make that usage productive and secure.

Beyond the dashboard: creating speed layers

Traditional business intelligence relies heavily on dashboards and static reporting, often delivering information 30 to 60 days after the fact. While these tools remain valuable for tracking established metrics, they create a significant gap between knowing and doing.

Generative AI offers the possibility of closing that gap through what we term ‘speed layers’ – systems that synthesise data in real-time and provide contextual insights that enable immediate action.

The goal isn’t to replace dashboards or eliminate existing processes, but to reduce complexity and lower the barriers to taking informed action. Instead of waiting for quarterly reports or submitting requests for custom analyses, decision-makers can engage directly with their organisation’s data through natural language queries.

They can explore relationships between different data points, test hypotheses and receive synthesised insights that would previously have required significant technical resources to generate.

This shift represents a fundamental change in how we think about AI’s role. Rather than viewing it as an output creator that works independently, successful organisations are deploying AI as an input enhancer that provides better information to human decision-makers.

The technology excels at structuring information, synthesising across data points and presenting contextually relevant insights, but the critical thinking and strategic decisions remain firmly in human hands.

The organisational reckoning

Perhaps the most significant challenge facing organisations isn’t technological, but structural. Traditional organisational hierarchies assume that expertise lives in silos and that cross-functional collaboration requires formal processes and meetings.

AI doesn’t recognise these boundaries. It draws connections across departments, challenges assumptions embedded in workflows and often suggests actions that require individuals to step outside their defined roles.

This creates profound tension. Employees find themselves receiving insights and recommendations that extend beyond their job descriptions. They may identify opportunities that require collaboration with colleagues in other departments or propose solutions that challenge established processes.

In risk-averse organisations, this leads to frustration and eventual abandonment of AI tools. The technology becomes an expensive curiosity, rather than a transformative capability.

The organisations that are succeeding take a different approach. They create innovation pods or small teams with explicit permission to experiment and prototype rapidly. These groups operate outside traditional constraints with modest budgets but significant autonomy. Their mandate is to demonstrate what’s possible, rather than to scale solutions immediately.

This parallels historical periods of rapid innovation. The Renaissance emerged not from rigid institutional structures but from the convergence of different perspectives, the breakdown of traditional boundaries and the freedom to experiment across disciplines. We’re experiencing a similar phenomenon today where the most significant breakthroughs come from individuals and small teams willing to work across traditional boundaries.

The curiosity imperative

The single most important capability for navigating this transition is curiosity. Organisations must shift from demanding immediate answers to fostering thoughtful inquiry. Instead of asking ‘When will it be done?’ and ‘What’s the answer?’, leaders need to ask ‘What happened here?’, ‘Why did this occur?’, and ‘What do you think we should explore?’.

This isn’t merely about being more patient or collaborative. It’s about recognising that the problems worth solving today require genuine exploration, rather than the application of established procedures. The value of AI lies its capacity to reveal patterns, connections and possibilities that weren’t previously visible.

Natural language as a gateway

One of the most under-appreciated aspects of current AI capabilities is natural language interaction. This isn’t simply about convenience. it’s about democratising access to insights that were previously locked behind technical barriers. When anyone in an organisation can ask sophisticated questions about their data using everyday language, it fundamentally changes who can contribute to strategic thinking.

The key distinction lies in context. Consumer AI tools have impressive capabilities, but they lack the specific context of your organisation’s data, processes and objectives. Enterprise AI implementations that combine natural language interfaces with proprietary data create competitive advantages that compound over time.

Moving forward

The organisations that will thrive in the AI era are those that solve the foundational challenges first. They invest in data quality and accessibility. They create cultural conditions that reward curiosity and cross-functional collaboration. They establish small-scale experiments that demonstrate value before attempting large-scale deployments.

Most importantly, they recognise that AI is not a destination but a capability – one that requires ongoing development and integration into how they operate. The question isn’t whether to embrace AI, but how quickly they can build the foundations that make AI adoption successful.

The technology exists. The opportunity is enormous. The question remains whether organisations are asking the right questions to unlock that potential.

Those that are will find themselves with significant advantages in an increasingly competitive landscape. Those that aren’t will discover that their competitors are asking better questions and getting better answers.