Why AI and big data still fail in business and how leaders can get it right

You need to answer the question ‘what’s in it for me’ to become a data-driven organisation

Jack Lampka is an AI and data strategist who helps organisations translate complex data into clear, practical business decisions. Drawing on senior leadership experience across global enterprises, including as head of data science at MSD, he is widely recognised for his ability to explain how data and analytics can be applied in ways that deliver measurable value.

In this exclusive interview with the London Keynote Speakers Agency, Jack Lampkashares his insights on why AI initiatives fail, how businesses can build stronger data cultures and what it truly takes to scale data-driven decision making.

You’ve stated that around 70% of AI projects fail. From your experience, what is the single most common factor behind those failures?

Them. People. People developing AI solutions and people using them. Whenever I talk about the topic, I always ask the audience what they think the main reason is for failures of AI projects.

And we basically have three components of success: people, data and technology. And when I ask the audience, most of the time, actually all of the time, most people say it’s the people. People are the biggest driver of success, but also failure of AI projects and that includes people using AI solution and people developing.

How do you cut through the hype and demystify AI so they can focus on what actually matters?

AI is less complicated than most people would like you to believe. At the end, AI analytics is just math. These are mathematical models that try to replicate, to model their reality as close as possible. So, it’s math. Math may be complex, but at the end it’s just numbers, it’s just data.

A second component when it comes to thinking about AI technology, you should not think as a business user about the technology. That comes only last. First, you start with business needs. You identify what data potentially is needed to address those business needs and only then comes technology.

Sometimes it’s AI that would be used to address the business needs. Sometimes it could be simple dashboards that could be sufficient to address your business requirements. So don’t start with technology and don’t think AI is as complicated as most people would like it to.

You often use the concept of a ‘data village’. What does this mean in practice and why is it critical for organisations looking to extract real value from data?

I use the term data village to describe the different skill sets that are required in order to be successful with data and AI. And it’s not only the data scientists that you need who develop these cool solutions, but also you need data analysts, data engineers, you need visualisation experts, data stewards.

So, it’s a big set of skill sets required, basically everything along the data analytics value chain and I call that the data village. Another component of that skill set is data product marketing. So that’s now a part of that data village in order to be successful with data in any organisation.

What are the most persistent misconceptions businesses hold about big data and how do those misunderstandings hold back data-driven decision making?

So, most businesses think you need big data in order to become a data-driven organisation. But at the end, it always starts with business needs. You need to identify what business hypothesis you have that could potentially be answered with data.

Whether that data is big or small, that comes next. First, identify the business needs. And sometimes small data is beautiful; sometimes small data is sufficient to address that business question. So don’t worry whether it’s big or small, just start with the business needs.

How can businesses cultivate a data mindset among non-technical employees?

People are the biggest driver of AI success or AI failure in any organisation. That includes people using data and AI solutions. And for that, in order for those to be used, you need to have the right data mindset level.

And if you don’t, you need to implement a data literacy programme to bring that mindset level. You need to identify where it is today and where you want to be in the future and identify a bottom-up data literacy programme to get.

You frequently emphasise the question ‘What’s in it for me’? Why is this mindset so central to successful data and AI adoption?

When you think about yourself, when you buy a product, when you buy a smartphone or choose a vacation spot or decide on a job offer, you think about what’s in it for me.

You never ask that explicitly, but you think about why should I buy this product, why should I go on that vacation, why should I take this job offer. So at the end, you think about for every product you purchase or select, you think about the question, what’s in it for me.

Why should that be different for data products? Every user of data and AI is asking him or herself the question, what’s in it for me. They never ask that explicitly, but if you don’t answer that question, at best your business colleagues will not use any of those data and AI solutions and at worst they would just pretend they do.

So you need to answer that question, what’s in it for me, in order to become a data-driven organisation.

About Jack Lampka

Known as a big data speaker who cuts through hype, Jack focuses on the human, cultural and strategic foundations of successful data-driven organisations. His work centres on data literacy, decision-making and ensuring technology supports real business needs rather than leading them.

This exclusive interview with Jack Lampka was conducted by Tabish Ali of the Motivational Speakers Agency.