There are rare occasions when technology breaks out of the bonds of the geeks in the basement and impinges on the national consciousness.
The debate around AI is just one of those occasions.
Politicians, business leaders, trade unionists, and media commentators have all weighed in on the subject: AI is going to create jobs; AI is going to destroy jobs, AI is going to make us more efficient, AI is going to make us money…and so on.
Above all, it’s particularly prevalent in business. According to research from Accenture, the use of AI can improve enterprise efficiency by up to 40 percent. No organisation is going to turn away from gains like that!
Naturally, the CRM market hasn’t been immune from these pressures. Adherents of ‘AI-is-the-future’ point out many ways that AI will transform the way businesses interact with their customers, for example by pulling together information and insight on customers from a multitude of sources, websites, and different social media, without any need for human intervention.
But AI offers more than pulling together and making sense of disparate information; it can use a range of different techniques to ascertain a customer’s views and desires to help customise approaches to them.
It sounds like a utopia for marketers: better customer engagement; more sales; more profits with less human intervention.
What’s stopping them?
Sadly, like many utopian dreams, there are some practical matters to deal with. The biggest of them all is the way in which information on customers is scattered around many disparate sources. Enterprises have been used to storing data in corporate silos and pulling it all together is not the most trivial of tasks – companies have rarely been designed to work that way.
And it’s not just a question of simply collating all the information, but also understanding how it’s sorted and what common formats there are. Some of the data could come from financial records, some from email output, some from SQL-based databases, while some could be image or video unstructured data – there’s a wide variety of possibilities and somehow they all these have to be pulled together.
It’s not purely about technology. Companies need to think about how they gather information and how they work together – it may need a completely new mindset. Small businesses understand this instinctively – there’s much more co-operation (and fewer specialist roles), larger organisations are not geared up for this way of working, and each department will often zealously guard its domain.
In an ideal world, companies should set up a cross-functional team to manage the implementation of technology such as CRM (Customer Relationship Management). This CRM team should be working with different departments to work out ways in which they could share skills and data to ensure that everyone is working with a common purpose.
There’s another consideration too: people skilled in AI are really thin on the ground. The use of AI requires some specialist expertise in gathering and interpreting the data. What many organisations mean when they say that they’re using AI is that they’re making use of algorithms – just one part of the AI armoury, but not everything.
In fact, this is one of the issues when it comes to talking about AI. The concept is often confused with the other elements – for example, deep learning, machine learning, and neural networks. Technologists are aware of all the distinctions, but very often business commentators aren’t. There needs to be full comprehension of what all the terms mean when we’re talking about AI engagement.
Of course, AI will have a considerable future when it comes to customer engagement, no-one denies that. But there’s a lot of work to be done first. AI shouldn’t be treated as some sort of magic bullet that will immediately transform company fortunes; we have to be careful that we don’t succumb to the hype too easily.
What companies should be doing is making sure that their CRM systems are configured correctly and are pulling in all the relevant information; businesses may think that they need a magical touch of AI but the answers could be sitting there in their own, existing software.
There’s a long way to go before CRM investment in AI bears fruit. For that to happen, there’s a need to have an enterprise-wide CRM platform (as opposed to a functional/departmental CRM) installed and all data in one place. And, on top of that, for a corporate culture that understands that the days for data silos have passed.
Until that happens, AI is going to be something that generates the column inches in the paper but not a concept that will have an effect on the way that we handle customers. Its day will come – but just not yet.