Last year, as part of Leeds Digital Festival, Suze Hawkins, Lead Data Scientist at Hippo, challenged the assumption that AI is a silver bullet, highlighting that strong data foundations, not just advanced models, unlock real value for businesses.

AI isn’t a shortcut to success

It’s easy to believe AI can solve every problem, but the reality is more complex. AI models are only as good as the data they rely on. For instance, Suze shared her experience with IoT sensors, thousands of sensors feeding live data into a system. If some of these sensors are faulty or the data is inaccurate, applying AI without verifying the quality of that data can generate unreliable insights, leading to costly mistakes.

Understanding the difference between AI, generative AI (GenAI), and traditional AI/ML is crucial. While tools like ChatGPT have captured public interest, they represent just one part of a much broader field. Businesses need to move beyond the hype and ensure their data is in order before considering AI solutions.

The reality of data science in business


Real-world data science is messy. Raw data is often incomplete, inconsistent, and spread across multiple systems. A large part of a data scientist’s role involves cleaning and preparing data before it can be used effectively.

But data science isn’t just about technical skills. Communication, business understanding, and stakeholder engagement are equally important. Without a clear grasp of business needs, even the most sophisticated models can end up underutilised or irrelevant. 

The challenge of fragmented data and governance

One of the biggest hurdles businesses face is fragmented data sources and inconsistent formats. Before rushing to deploy AI, businesses must first focus on integrating and standardising their data to create a unified view. Without this foundation, applying AI can amplify errors, leading to misguided insights and costly mistakes.

Suze’s work with companies seeking AI-driven customer insights highlights this issue. In one instance, a company lacked a unified customer view. Rather than applying AI to incomplete and scattered data, the team first focused on building a solid data foundation. Only then could they generate meaningful insights.

When AI isn’t the best tool for the job


AI isn’t always the right answer; sometimes, simple rules and domain expertise can outperform complex models. Clearly defining events and signals can lead to transparent decision-making which may be more appropriate than applying AI to messy, unreliable data.

Without a strong foundation, AI can magnify errors rather than provide clarity. Businesses should prioritise addressing data quality issues before considering AI-driven solutions.

The smarter approach: Data first, AI second


AI should be the final step, not the first. A well-structured data architecture allows businesses to make better decisions—even without AI. Investing in reliable data pipelines, reporting tools, and dashboards lays the groundwork for future AI capabilities. By focusing on data quality first, organisations can unlock real value without unnecessary complexity.

Next steps for businesses:

  • Conduct a data audit before investing in AI to ensure your data is clean, consistent, and accessible.
  • Build internal processes for data validation to ensure reliable data flows into your systems.
  • Focus on upskilling teams in data governance to maintain data quality as your business grows.

AI isn’t a one-size-fits-all solution, and without clarity on your organisation’s readiness for AI or your current capabilities, you risk investing in the wrong areas, missing out on growth opportunities, or failing to align AI initiatives with your organisational strategy.

To help ensure you’re on the right track, our AI maturity assessment tool offers insight into where your organisation stands on its AI journey. It highlights the steps you need to take to progress, providing a clear, actionable path for both organisations just starting out or those looking to scale.

By assessing your AI maturity, you can avoid costly missteps and align your AI efforts with your broader business objectives, setting you up for long-term success.