Want to Know if Your Data Function is Successful? Focus on the A:I Ratios

While organisations around the world are publicly scrambling to understand how generative artificial intelligence will transform their business, in the background similar questions are being asked about how past investments in data capabilities have made an impact. Before there was the hype around Chief AI Officers, there were board discussions about the need for Chief Data Officers. Now many organisations are rightly asking what return they got on that investment.

This is not idle speculation. After all, not only are their many parallels between becoming a data-driven organisation and one that leverages artificial intelligence at scale, but data foundations are critical to succeeding with artificial intelligence. Given the cost pressures many organisations are facing, CFOs are on the hunt for budgets that can be trimmed and the business case for new investments needs to more rigorous than that during previous cycles.

With so many different metrics and frameworks, how should business leaders evaluate whether their data and analytics functions have been successful?

From my perspective, as someone who has held both commercial and data leadership roles, there are two straightforward metrics that can be used. To understand what those are, we need to consider the journey of investment to business impact: Data Infrastructure (people + platforms) → Data Assets (e.g. single customer view) → Data-driven Insights (e.g. next best action) → Data-driven Actions (e.g. personalised offer driving higher average order value).

With this framework, the first metric is the ratio of Assets to Infrastructure (A:I). On numerous occasions I have seen organisations make large investments in data teams and analytics platforms and still struggle to unify their customer data or create consistency in their sales numbers. To be blunt: unless you are using your infrastructure investments to serve up high quality data assets, what was the point of that investment? Technology for technology's sake is a distraction.

The second A:I metric is the ratio between Actions to Insights. In other words, how effective is the organisation at using insights surfaced from their data to drive meaningful action. Examples include using data to improve how the organisation segments customers, places orders from their factories, or decides to launch new products. What matters is that insights are not just intellectual curiosities, but meaningful findings that end up driving the business forward.

Now one might ask: what about the ratio of data insights to data assets? Surely that matters too? The challenge is that finding insights can be a matter of luck as much as skill. Sometimes high quality data assets end up telling you things you already know – not something insightful and actionable. Conversely, a messy and partial data set could hold the key to a powerful insight into customer behaviour. One simply does not know beforehand, which means the ratio is less reliable.

The important thing to point out is that the business impact of the A:I ratios is multiplicative. After all, if your business functions are amazing at turning insights into actions, but the vast majority of your infrastructure investment does not produce usable assets, what is the point? Given the focus on AI, data leadership is arguably more important than ever, but only where investments and insights are leveraged to create high-value assets and actions. Otherwise, it is just noise.

– Ryan

Cover image by DALL·E.