How to use AI in e-commerce

Technological innovation is no longer optional: in an effort to optimize customer experiences and increase e-commerce results, we have become slaves to the algorithm. Fortunately, artificial intelligence opportunities abound and new technologies can be implemented relatively easily.

Accenture Interactive's Fjord identified seven trends that will impact your business this year. We are becoming Slaves to the algorithm, as in the new marketing environment, algorithms are taking on the role of gatekeeper between consumers and brands. They are indifferent to all branding efforts that still influence consumers’ buying decisions. How can you adjust your marketing strategy to embed this trend when all visual and empathy-reliant signs are removed?

This article examines how artificial intelligence will impact e-commerce and digital marketing. What technological options are there? And how can you best adopt these technologies yourself? Imagine how you can benefit from AI in e-commerce.

Why the use of AI has become inevitable

In an ever-changing world, innovation is key to staying relevant. And today’s world is changing quickly indeed. Not only are retailers, manufacturers and digital start-ups locked in a constant competition for customers, there’s even talk of a Fourth Industrial Revolution. One in which there is no longer a physical place where a brand might exist and come to life for its target audience. One in which a fast succession of technological breakthroughs enabled frontrunners to change the retail game in its entirety.

Failure to keep up will lead to a rapid decline in business. Now more than ever, e-commerce is an innovator’s game. 

Amazon Go, Ebay’s ShopBot and other genius applications are now offering a seemingly endless amount of opportunities to ease their customers’ journeys, relying heavily on artificial intelligence solutions. Their purpose: to create a more personal shopping experience for their consumers. As contradictory as this may sound, that is exactly what AI is doing. Through the use of algorithms, the possibilities for personalized marketing in e-commerce have greatly increased – but we will get to that in a moment.  

Over the last few years, more and more marketing touch points have been created. Ranging from stores to websites and from chatbots to voice assistants, consumers use these (and many more!) contact points at all times of day and at all possible locations. Moreover, they expect a consistent and personalized experience, and preferably one in which all information is updated instantly throughout all communication channels.

To watch this video you need to accept marketing cookies

Both the diversity of these touch points and the amount of data that is generated necessitate the use of artificial intelligence (AI). Manually, it will simply be impossible to keep up.

How AI can lead to more profitable customer experiences

Now that artificial intelligence is making waves like never before, how can you use it to create a more personalized, more profitable customer experience?

Use AI to predict sales or services

First of all, machine learning, which is an important part of artificial intelligence, enables the analysis of large amounts of customer details. Instead of painstakingly working your way through massive data sets, CMOs can now establish patterns much more easily. Patterns that, more often than not, would have previously remained hidden. As proofs Netflix, by predicting what their viewers want, before they know themselves.

The takeaway message, of course, being that once you automate the collection and interpretation of information, you can make much more informed predictions with regard to marketing campaigns.

Why the Most Human Approach to Digital Identity Might Not Be What You Expect - Automation & RPA Visual - by Accenture

Use AI to provide better services at lower cost

Another way to create more personalized marketing opportunities is by using AI to eliminate redundancies and automate processes. Chatbots, for example, can help businesses save on customer service costs by:

Not only might this lead to cost reductions of up to 30 percent, it will also lead to a more customer-centric approach. After all, the freed-up agents will now be able to focus on what they do best: get in touch with their customers in person.

It is worth adding that, as the system is self-learning, its answers will reflect human conversation more and more over time. Already, people can become confused when conversing with a chatbot, as illustrated by the much-asked question: ‘Are you a bot?’

Use AI to increase customer satisfaction and sales

Apart from offering customer service, artificial intelligence can enable conversational commerce. This is real-time, human-style interaction between a consumer and a bot, messenger or voice-chat. Using automated and intelligent services, they ask questions to gain a better understanding of the consumer’s intent, followed by personalized recommendations.

Unsurprisingly, this leads to higher customer satisfaction and, consequently, more sales. Ebay’s ShopBot, for example, offers curated deals, based on the information provided during GIF-enhanced quizzes.

Starbucks’ chatbot, which can be found in the company’s app, makes it easy to place an order, whether you prefer using voice commands or text messaging. The chatbot will tell you when your order will be ready and what the total costs are.

A similar service is offered by Pizza Hut, which has launched a Facebook Messenger and Twitter app taking users’ orders and tracking delivery. Customers can also re-order their favorite pizzas, ask questions, and see current deals.

Another good example is 1-800-FLOWERS. Their Facebook chatbot shows pictures of various flower arrangements and make suggestions based on conversational cues, enabling customers to order flowers all over the world.

AI in e-commerce - COBE chatbot - Accenture

Use AI to scale personalized content

Instead of manually writing algorithms or simply guessing what customers would like to see, machine learning and intelligent analytics can be used to provide up-to-date and personalized content, and to scale this to the vast amount of touch points, channels, brands and countries.

Optimized shopping experiences used to be an area where only giants like Amazon excelled. With the advent of leading platforms such as Adobe Experience Cloud and Salesforce Einstein, regular enterprises have the possibilities to leap-frog and become at par, or even surpass, the experience offered by Amazon.

Use AI to optimize customer searches and recommendations

But what if a consumer is looking for specific information? Do we then resort to ‘old-fashioned’ search boxes? Of course not. Nowadays, artificial intelligence helps customers to find what they are looking for in a much more natural way.

Go Find, for example, invites its users to take a picture, based on which it returns similar results, including price ranges and (online) stores. Amazon’s Echo Look uses a camera, combined with artificial intelligence, to become your personal assistant that recommends which looks suits you best.

And based on typed queries and external data sources, Bloomreach uses its understanding of user intent to automatically find the right products and create landing pages for specific categories. Even before marketers realize they are needed.

Use AI to automate back-office processes and reduce manual tasks

A final example of how artificial intelligence can be used in e-commerce, is the manual mapping of products and the management of product data. Tasks like these are time-consuming and error-prone. Automation can be a great help.

But there is more to this. While image recognition might be a great help in product categorization, Pinterest uses the technology to help its users search for a specific item within a photo.

Automating back-office processes have enabled website creation tools to build websites based solely on questions they ask and answers their clients provide.

In short, AI in e-commerce is needed in today’s business environment. It is also an excellent way to generate cost reductions. But most importantly, it offers enormous innovative opportunities. That said, the question arises: where do you start?

From blueprints, to design thinking, to prototype development

We have said it before and we will say it again: it all starts with a strategy. First, you need to identify where automation and intelligence have the highest potential.

One way of doing this, is by using Accenture’s Capability Blueprints and Design Thinking Services. The blueprints enable the mapping of processes and subsequent identification of AI-related opportunities. Design thinking will shed light on which activities will add the most value. For example, the blueprint might identify customer service and design thinking will point at the deflection of calls to a digital assistant.

Now, a prototype can be developed and marketed. When automation begins to generate value, the money ought to be reinvested into additional AI technologies, thus scaling up gradually.

AI is a necessity for survival

We agree, the message is strong: AI is a necessity if you wish to remain relevant in the future and maintain your competitive edge. But as much as we are slaves to the algorithm, we can use it to our advantage.  Moreover, we can start small, by identifying opportunities and initiating projects that can be scaled up when proven successful. The time is now! Where will you start?

Author: Mehmet Olmez