Believe it or not, but the biggest thing holding AI back is people. Not so much due to AI replacing jobs, but more about the fact that in our jobs, people tend to create siloes – areas they like to own, be responsible for, and ultimately get the credit for when things go well – to be fair, there’s not much wrong with that, but it does have a negative by-product.

As powerful as modern tools might be in providing data-driven insight, it’s the raw materials that make the biggest difference – the source data. And when that might be held across different locations, a single view becomes a challenge.

This is what holds AI back. When customer preferences and prior interactions aren’t joined up properly, how can you hope to address and anticipate their needs adequately?

In a report late last year, Deloitte argued that for a company to be able to use AI to its full potential, they need to fundamentally rethink the way humans and machines interact with each other. But where do businesses start?

Break the silo habit

We’re at a tipping point when it comes to data. Many businesses have more than they know what to do with – typically because it isn’t properly organised, updated or accessible.

Businesses need to resist the temptation to hoard data (which could well breach local privacy laws) and instead spend more time evaluating what data they really need to collect and curate to deliver great customer experience; they need a true data strategy.

They also need to show their customers how they benefit from their data being used. When this is not clear, trust in that brand is at risk.

What will separate the successful brands from the strugglers will be their ability to navigate the complexity that data presents. They’ll need to effectively demonstrate that smarter use of data can bring value to their customers and improve their experience.

Perhaps the biggest of all complexity challenges is presented by disconnected, siloed data and technology. With a data strategy set to identify the right data for success, the prime objective has to be, to connect that data, literally or virtually. Which business wants to plan for success on a disconnected, broken understanding of their customers?

The problem is, as businesses evolve, they’ve added telemarketing systems to CRM systems, to email, social, mobile and more.

Often essential evolutions, but they must be connected, to enable businesses to plan, forecast and nurture both proactive and responsive customer experiences that result in customer and business growth. Of course, it also helps with measurement too.

It is imperative for businesses to have effective identity management solutions and a strategy for integrating technology at the data layer, i.e. not for the sake of connections but for the sake of customer understanding. Done correctly, overall data protection compliance is improved by making sure data is accurate and up to date.

When AI comes to town

Artificial intelligence will be incredibly powerful for the future of customer interactions. It makes analysis and forecasting possible at real-time rates impossible for human-led analytics to achieve.

Whether it’s bringing siloed datasets together, crunching data for complex research projects, or just figuring out how to make smart devices more useful, AI has a wealth of potential.

Paradoxically, moving the needle here requires humanity. What this means is businesses need to determine AI’s remit and give it the opportunity to get it right.

AI gives us potential, but only people can give it purpose and, part of that means giving AI the data necessary to understand customers. Never forget that the first person benefitting from having a good understanding of the customer is the customer themselves.

When customers know their data will be managed and used in a fair and compliant way to genuinely improve their experiences, we know they are more willing to share it.

Data-enabled convenience, relevant recommendations or tangible offers are essential for building repeat business and loyalty. Once in motion, the ideal is a virtuous cycle, with data gathered from interactions being used to make future experiences even richer.

AI needs the right fuel to succeed

Most organisations can make huge improvements to their data-driven operations by eradicating or at least reducing silos. AI is the shiny new thing that promises so much. But as is often the case, it’s behind the scenes that has the biggest impact.

Only by having a unified approach to data alongside a robust data strategy to protect and manage it, can data’s potential be unlocked to provide value for all.

Jed Mole is Acxiom CMO.