Before looking at where we are with artificial intelligence (AI) in ecommerce, it’s worth looking at some definitions. AI might seem to have emerged in the last couple of years but has, in fact, been around for decades.
Artificial intelligence or AI covers a broad range of ‘smart’ technology that can respond to enquiries and solve problems like humans. The concept has been around for over 70 years and has developed by using models of how our brains work. The term can apply to almost any technology that imitates human intelligence like Siri, Cortana or Alexa Voice, for example.
Machine learning or ML is a subset of AI. It enables machines to identify patterns and learn from data to make predictions and choices without further programming. ML can track online behaviour, for example, and use recent purchases, preferred colours and spending to recommend other products.
There are some interesting and exciting opportunities to use AI and ML in ecommerce. They can improve customer experiences, enhance personalisation, predict online behaviour and generate content.
However, alongside these opportunities, it’s also important to retain human engagement and oversight. Unintentional bias can lead to confusing and inconsistent responses and even unfairness. Maintaining ecommerce security and data protection is also essential as huge sets of data are interrogated by AI engines. The key is to maintain the right balance between automation and human interaction.
Practical applications of AI in ecommerce
We already use AI in ecommerce much more than we realise. Automated learning and modelling can improve online searching and buying. Based on near real-time information customers can receive information about product trends, pricing and optimum delivery routes. Businesses can automate processes, make better decisions and predict future demand more accurately. AI-enabled dynamic pricing can even change product prices based on current levels of supply and demand.
Online searches
Seven out of 10 people expect more personalised experiences and get frustrated when they aren’t provided. Machine learning has improved search results, personalisation and recommendations. Adobe Sensei and Klevu are examples of ML search and recommendation solutions based on customer preferences, interests, and online behaviour. Klevu’s AI chat tool MOI is accessible from a website’s search box. It offers ‘conversational commerce merchandising’ with an easy-to-use dashboard for fine-tuning.
Content creation
As many businesses aim to do more with less, in-house teams have become smaller. Generative AI or GenAI tools like ChatGPT can support content creation and improve accessibility. For example, product managers can input specifications and guidance on the right tone of voice for AI–generated product descriptions. Other tools like Perplexity AI answer questions and provide citations for blog content. Especially helpful for large fashion catalogues, Adobe Firefly generates images to reduce the need for new product photography. However, a human touch is still needed to check, review and amend content before use.
Customer service
Chatbots can support simple transactions around the clock while your sales support teams focus on more complex queries. They work well for frequently asked questions but are not a risk-free option. For example, Air Canada was challenged after its chatbot advised a customer they could claim a refund. In fact, this was against company policy, so the customer later received compensation. It raises some interesting questions about the legal position of chatbot responses.
Search engine optimisation
Alt text for visual content improves accessibility but is underused to improve search engine results. It’s a ranking factor because it helps search engines understand website content. However, most website homepages have accessibility errors and missing alt text is one of the top problems. Creating good alt text is time consuming if your site has hundreds or thousands of images. AI computer vision image-recognition algorithms reduce the manual effort, freeing up time for other valuable activities.
Google’s latest feature, AI Overviews, organises answers to open-ended search questions. Like a personal assistant, it can help with planning and research. The AI information cards provided include links to products and content publishers. Specific products and images appear rather than links to categories, so content optimisation at every level is becoming more important. Overview responses appear before the more familiar organic search results which will move further down the screen. High quality content and product details are essential to appear in the Overviews and maintain organic site traffic.
Marketing
In a survey of retailers, improved personalisation was a priority but only 15% said they have fully implemented it across channels. The Adobe Experience Platform uses AI and ML for sophisticated personalisation and customised content. Big data from purchase histories and customer interactions improves segmentation and messaging. It can increase both revenue and customer retention by 10% to 15%.
Sales processes can also become more efficient when you use AI and ML to analyse cart abandonment and automate follow-up messaging. You can also encourage customers to move through your sales funnel with simple chatbot messages.
Overall, a third of businesses say they are already using generative AI in marketing and sales. Many have set up dedicated teams and budgets to integrate AI into their development plans. Half are focusing on customer experience as the core of their AI strategy. As ecommerce businesses look to do more with less AI will become an integral part of the solution.
Where is AI going?
The overall trend for AI and ML applications in ecommerce is to improve capacity, reliability and accuracy. Businesses must determine the role they can play in product innovation and understanding customer behaviour, for example. They will need new skill sets to successfully capture the transformative power of this technology. Capability building must feature in their strategic plans.
Despite exciting new possibilities, it’s essential to set clear business goals for the use of AI and ML in ecommerce. Monitoring and measurement will still provide the ultimate evidence of effectiveness.
Partners of choice
Some of the benefits of AI are fundamental. They include improved customer experiences, better search engine results and streamlined business processes.
Using our tried and tested discovery process we can identify where AI and ML could make the greatest difference for your ecommerce business. We can help you target your marketing more accurately, reach your ideal customers more easily and increase customer retention. Supported by our extensive experience and expertise, you can implement powerful and future-proofed solutions.
As Partners with Adobe and Klevu, we constantly explore AI and ML innovations and opportunities that can help your ecommerce business grow. Get in touch today.
