AI: adopt or fail

Artificial intelligence (AI) is proving to have the power to transform commerce, says Chloe Hobart. But what do business leaders and consumers think it is, and how is it being used?

There’s lots of talk about AI, but do business leaders and consumers really know what it means and what its applications and constituent parts are?

These questions prompted Wunderman Thompson Commerce to commission research from Censuswide. A total of 1,557 consumers and 610 business leaders were interviewed.

We started by asking leaders in roles focused on digital and ecommerce across a variety of industries in the UK, US and China – how confident they were in their understanding of AI. Confidence is high that they know what AI is and what it does; 83% claim to have a strong understanding. In fact, business leaders are so confident that 89% who have implemented an AI solution consider themselves an “AI leader” in their industry.

How is AI defined?

Perhaps the best definition of true AI is ‘goal-directed adaptive behaviour’. AI is goal-directed, in that it has a clear objective – to maximise fuel efficiency, improve staff wellbeing, and minimise customer churn. An AI solution must have something to optimise for.

Behaviour is about decision-making. AI is responsible for making, or at least supporting, resource allocation decisions. And, crucially, AI is adaptive, in real time and in production (real) environments. It makes a decision, learns if that decision was good or bad and then adjusts itself so that it can make a better decision next time.

True AI is a mixture of many technologies, including descriptive analytics (dashboards, business intelligence), predictive analytics (data science, machine learning) and prescriptive analytics (optimisation, decision science). Each can provide enormous value when used in isolation, but can be truly transformative when used together.

We wanted to get a feel for just how well business leaders understood these different elements. While some of these applications that make up AI solutions are relatively well understood – reporting/dashboards/analytics all ranked highly – others are less so. For example, understanding was weaker around the architecture, implementation and optimisation components.

Adopting AI technology

We also wanted to understand just how widespread AI has been adopted. Our data told us that over three-quarters of businesses (77%) have already adopted or implemented AI solutions in their organisations. However, we suspect that many of these businesses have adopted some form of data analytics and dashboards, or descriptive analytics. Some have implemented data science systems that recognise patterns, forecast demand or generate insights from data (predictive analytics). And fewer again are using optimisation or operations research technology to automate and optimise decision-making and resource allocation (prescriptive analytics).

And across the three countries surveyed, this percentage was relatively consistent, with 79% of Chinese businesses having already adopted AI, 77% in the US, and 74% in the UK. As one might expect, it is the largest businesses – those over 500 employees – whose adoption is highest (92%), while businesses of 1–9 people have an adoption rate of 35%.

With so many businesses having already adopted AI solutions, the next obvious question is how long these solutions have been in place. Across the three countries, the average is 2.64 years, with Chinese businesses being slightly ahead (2.78) versus their US (2.68) and UK (2.45) counterparts.

AI and consumers

What do global consumers believe AI is? For most people it’s the automation of tasks, followed by physical machines and robots, while 20% of consumers understand AI through the prism of the content that they engage with.

But understanding seems to be different across the three countries surveyed. In the UK, 26% of consumers said they didn’t know what AI was, compared to just 5% in China. In China, the highest percentage of consumers felt it was about apps or software (52%), while in the US, the highest percentage thought it was about physical machines or robots (33%).

Our data tells us that 44% of consumers are excited about the prospect of AI, versus just 16% who are not. And what excites them most about companies that use AI is improvements to customer service, including things likes chatbots and call centres (47%). Next up is something that is very dear to many consumers’ hearts – fast delivery. And clearly many consumers believe that AI will be able to help this. And, in third position is a good in-store experience, showing the need for businesses to continue thinking both physically and digitally.

However, concerns about AI do exist. We asked consumers to rate how happy they were for AI to make purchase decisions on their behalf, with 10 being very happy and 1 being not happy at all. The outcome? An average score of 5.41.

Read more from Atticus Journal Volume 27