Smart Shopping: How Is AI Transforming E-Commerce?
[Photo by Andrea De Santis on Unsplash]
While AI and specifically generative AI have catapulted into discourse, so have the ways in which it can benefit businesses and their e-commerce needs. Much like AI can be used to personalise the experience for shoppers, so too can it be utilised to cater to a brand’s specific targets - whether that be trying to convert shoppers, create increased customer loyalty, provide location-specific text generation, and so on.
With an endless amount of possibilities, and as technology advances at an accelerated rate, it’s a given that commerce, specifically e-commerce will be impacted. We wanted to hear about how AI is ever more transformative in the commerce space.
To answer this question, LBB’s Nisna Mahtani asked experts to share their insights with us at LBB. Here’s what they had to say:
Alex Hamilton
Head of innovation at Dentsu Creative UK
When it comes to these large language models in the generative AI trend, they’re cheap to build and can be very personalised. Once they’re in place, they understand everything about the business, you can create copy, image generation, etc. So the challenge for a retailer is where to start – ‘How do we start getting our head around this?’ and ‘Where are we? Where do we begin?’
The other thing is that websites are all quite static, but I think AI can actually add warmth to those experiences. Generally speaking, the digital space is pretty static and built on the same tech – which is a content management system or Salesforce commerce cloud – so I think there's a notion at the moment that AI can make these experiences more conversational and bring more humanity to them.
For example, when you find a website you have to hunt, find the information, scroll and click around, and then you get choice paralysis. If you had to design your kitchen or bathroom manually, that’s tedious, but if you had the chance to go back and forth with AI, upload an image of the space and have it help you design, you can not only have more efficient conversations but more personalised ones. It starts bringing the human touch to life.
An example of this is the Lowe’s retail store that our US team worked on and that was thinking about how to power virtual assets or characters to provide assisted selling. We considered that to the point where we were pairing real people with AI, so that we could record real people and drop them into the website. That way, they could walk around and interact with things – all of these possibilities are quite interesting.
Laura Cullen
Global commerce director at VML
Artificial intelligence is revolutionising the way we deliver commerce, providing personalised recommendations, accelerated content creation, rapid insights generation, and improved customer service – all with efficiencies at scale. Specifically, with the e-commerce space, harnessing generative AI to fuse data on brands, consumers and channels that accelerate the creation of digital commerce content at scale.
While the benefits are vast for brands from search performance to shopper conversion, legal risks accompany AI-generated content that must be mitigated. Instances of brands pushing content directly from third party AI tools that violate e-retail guidelines or neglect legal and regulatory concerns can lead to liability or delisting. Consequently, at VML we focus on best practice discovery, ensuring our AI-powered insights and content creation are complemented by human ingenuity. This allows our teams to deliver higher quality outputs that enhance the customer experience while driving sales growth at scale. Complementing generative AI with curative AI also allows us to evaluate content against brand, accessibility, legal and best practice guidelines which are checked by a human prior to the content’s deployment. These new tools and ways of working are already driving significant time savings in e-commerce content development and deployment of up to 60% for our brands and agency.
Nonetheless, the sheer volume of content produced by generative AI poses infrastructure challenges. Ensuring brands have adequate teams, systems, and processes for content approval, deployment, and maintenance is crucial. Additionally, as AI is increasingly utilised for creativity and personalisation, privacy compliance becomes paramount. Brands must consider the source and use of data, especially with advancing data maturity, to unlock insights and personalised experiences while respecting privacy in a modern era of commerce.
Dane Buchanan
Chief data and analytics officer at M&C Saatchi Performance
When it comes to e-commerce, AI is a double-edged sword. It gives us the ability to personalise customer journeys in ways we couldn’t even imagine a decade ago. However, with the advent of generative AI, the risk is that we may lose the human side of a brand, jeopardising authenticity and entering a world of AI blandness.
While not directly related to e-commerce, behavioural economist Dan Ariely has an interesting experiment live at the moment (https://dan-vs-chatgpt.irrationallyyours.com/). Participants are asked to distinguish between Ariely’s answers and those generated by gen AI, highlighting their preferences and identifying which responses are AI-generated. The results demonstrate a clear preference for the human touch. This underlines that gen AI should be seen as a tactical tool that can enhance and optimise processes and content rather than replace the human element.
The challenge this presents for brands is: how do we capitalise on the advances in AI while maintaining our unique voice and authenticity?
If we take a step back from the world of gen AI. AI/machine learning has been shaping e-commerce for many years across platforms like eBay, Amazon, Alibaba etc. Amazon has brilliantly integrated AI recommendations throughout the consumer journey. However, the recent introduction of AI-generated summary reviews does present a new challenge. These summaries, which are created from real customer reviews, lack the human insight that I require. I find myself ignoring these and sifting through the individual reviews, trusting my own conclusions over the AI’s summary.
However, where I see gen AI making the biggest positive impact is on search, and essentially democratising this across the e-commerce space. For example, Walmart has recently supplemented its search experience with gen AI, giving you a one-stop shop to plan and organise what you need rather than search off the platform and come back looking for specific items. Shopify, Amazon, and Instacart have all introduced similar generative AI search experiences over the last few months. These tools give e-commerce platforms a mechanism to create a direct connection with consumers and increase brand loyalty, and in a world where first party data is king, this is an important asset to cultivate.
Barbara Grabiwoda
Chief strategy officer at Publicis Le Pont and PXP CEE
In today's fragmented e-commerce retail market, navigating data points and consumer behaviour presents a complex challenge for brands. At Publicis, we're harnessing the power of AI to revolutionise commerce across various divisions - research, operational optimisation, and consumer experience. Our approach goes beyond simply replacing humans with machines; instead, we're focused on unlocking new, untapped value by integrating AI capabilities while maintaining a human-first approach.
With AI, we help our clients to streamline the data to make more data-driven decisions.
Through AI-led research, we leverage advanced natural language processing (NLP) and machine learning to reimagine data and provide valuable insights for a personalised consumer experience. This allows us to tailor content strategy and design based on individual shopper needs, patterns, and shopping missions.
Additionally, our Ask Profitero tool, a chat-based AI assistant for commerce analytics, democratises digital shelf optimisation. By simplifying complex data analysis and content generation, this tool provides insights across various metrics including availability, content, placement, pricing, ratings, reviews, and sales data.
Moreover, we're exploring AI's potential to enhance creativity within commerce. By utilising visual gen AI, we can produce backgrounds, textures, and elements to supplement our creative automation processes, leading to a multitude of creative variants without compromising quality.
Furthermore, we have developed proprietary tools for AI-powered copywriting, which offers opportunities for e-commerce-specific text generation, localisation, and adaptation, enhancing communication with consumers on a personalised level.
However, as we embrace AI in commerce, we must also address the associated risks, particularly concerning data and copyright security. Publicis Groupe is committed to ensuring the secure integration of AI into our work through CoreAI, a layer of AI connecting our enterprise knowledge. This initiative aims to supercharge our talents and clients while maintaining robust security measures.
In conclusion, AI is reshaping the landscape of commerce, offering opportunities for personalised experiences, optimised operations, and enhanced creativity. By navigating risks and leveraging AI effectively, we can unlock new possibilities for brands and consumers alike.
Matt Sutherland
Head of tech at True Digital
There is an ever-increasing need for the content and products that we serve to our users to be relevant, and within the e-commerce industry producing relevance is imperative. This need allows for the introduction of AI-powered tools, such as Algolia.
true has been utilising Algolia with great success within our commerce offering to provide AI-generated product recommendations, whether that is similar items, alternatives (when products are out of stock), or highlight trending items which allows for the automation of additional visibility for either seasonal or popular items.
The purpose here is to increase the value of the basket, the lifetime value of the consumer, and the ongoing retention of that user, and we have seen all of this improve, due to increasing the relevance of the content and products that we are serving to end-users.
Algolia also offers AI search capabilities, which are all about prioritising search results based on the understanding of a user, whilst layering in contextual influencers to present relevant results. This is not about returning search results based on alphabetical ordering, or the latest saved article, this is making sure if someone is typing ‘Michael’ into their search box, we know whether they are looking for Michael Jordan, Bublé or Caine and serving them the right content as quickly as possible.
Additionally, Algolia offers advanced AI search capabilities, prioritising search results based on user profiles and contextual influencers. It utilises personalised search to present relevant results, such as prioritising search results for a particular drink (for example) based on user preferences.
The more a user must scroll to the content or product that they are looking for, the more
frustrating the experience. Ensuring consumers reach their end goal, as quickly as possible, enhances experience. AI, operating discreetly and seamlessly, can effectively accomplish this goal.
Karin Libowitzky
Managing director and commerce lead at Accenture Song ASG
AI in e-commerce goes beyond just personalised recommendations and inventory management. It will lead to a renaissance in e-commerce. Creative applications of AI are already redefining shopping experiences, making them more engaging and unique for customers and we have just started to explore the possibilities:
Virtual Try-Ons and AR Shopping enable customers to virtually try on clothes or preview furniture in their space, increasing engagement, and thus reducing return rates.
AI-Powered Chatbots are already evolving into personal shopping assistants, by providing personalised product recommendations and shopping assistance, making the experience feel much more human.
Predictive Personalisation: By analysing customer data, AI predicts interests for tailored shopping experiences, such as suggesting products based on weather conditions or preferred book genres.
Dynamic Pricing and Promotions: AI adjusts prices and promotions in real-time based on various factors, ensuring personalised and relevant deals for customers.
Social Listening for Trend Forecasting: Analysing social media trends, AI enables brands to anticipate and stock trending items, crafting resonant marketing campaigns.
Sustainable Shopping: Promoting eco-friendliness, AI recommends sustainable products and optimises logistics to minimise environmental impact, catering to eco-conscious consumers and thus fueling brand perception positively.
But doesn´t the application of a responsible AI framework 'eat' into creativity?
There is no doubt that AI is already fostering creativity by providing new tools and methods for creative expression. Artists and creators can leverage AI to explore ideas and produce works in ways that were not possible before.
Here comes the ‘but’: The implementation of a responsible AI framework isn't optional; it's mandatory. By mandating a responsible approach, we can harness the benefits of AI while minimising risks to society and individual rights. This isn't about stifling innovation or creativity; it's about ensuring that technological advancement proceeds with caution, ethical considerations, and respect for human values.
Emma Murray
UX designer at Build in Amsterdam
AI has been quietly revolutionising our online experiences for years. Spotify's personalised playlists, TikTok's endless stream of engaging content, Instagram's tailored feeds, and Netflix's spot-on recommendations are just the tip of the iceberg. But we're now moving to a place where this tech isn't just for the big players anymore; it will be everywhere, giving every kind of company the ability to offer highly relevant content.
The real transformation, we believe, lies in the production process. AI is set to streamline how companies integrate their operations, analyse vast amounts of data, and connect those insights to their content strategies. This will not only make content more pertinent but really accelerate the pace of creation and curation. The era of automated content generation is upon us, but so is the need for meticulous curation.
However, as everyone gets on this AI train, standing out will require more than just feeding people what they like. It'll be about having a distinct voice and brand that doesn't feel like it's just echoing the echo chamber.
Now, everyone's buzzing about chat interfaces as the next big thing. But let's not forget, computing started with text input interfaces but quickly moved to graphical interfaces because they were easier to use. Betting everything on chat feels like missing the point. The future's not just about sticking with one format; it's about staying flexible and being ready to adapt.
Rupert Woodward
Managing director, data and performance at Iris
AI has the potential to transform how we do business and live our lives. This is especially true in the world of commerce, where AI has been working behind the scenes to curate better on-site search results, surface suggested products and drive in-market users to site through performance media like Google shopping. The likes of Shein use AI to trawl huge volumes of social and scraped web data to identify fashion trends, getting products manufactured and on sale in a matter of days to capitalise on these trends.
The rise of generative AI adds a further dimension to this, giving retailers the ability to create the high-volume, low creativity content that powers commerce - such as product and category descriptions and longer-form buying guides. At Iris, we are using a ‘human in the loop’ approach to Gen AI. The technology is not yet at a point where it can create copy for brands that delivers on tone of voice requirements without supervision, but it adds significant value to the work we do for clients by saving many hours of manual time on tasks like optimising meta data, structuring articles and building negative keyword lists for brands. This frees up time for our people to focus on the value-additive work which really drives our clients’ businesses forward.
The flip side of generative AI is its inevitable use by bad actors to flood search results with content and create fake product reviews - both the fake written reviews we are all familiar with and deep fake video reviews by AI-generated influencers which have the potential to do huge harm to brands. At Iris, we are helping brands to navigate this fast-moving, challenging landscape, ensuring we are pushing the boundaries and delivering the value that AI can bring, whilst doing so in a transparent and responsible way that matches individual clients’ appetites and needs.
James Shepherd
Co-founder and managing director at Atomic Altitude
Retailers are increasingly integrating AI into their operations to enhance efficiency, improve customer experience, and drive sales. AI provides clear benefits to both retailers and consumers. However, there are clear disadvantages too. Whatever the advantages and disadvantages, AI is here to stay, and its use will only become more prevalent.
All the advantages stem from the way AI enhances data collection, sorting and deployment. AI analyses and sorts data at great speed to provide greater efficiency, personalised recommendations, improved customer engagement and increased sales.
Because of AI, decision making to improve everything from inventory management, price optimisation to supply chain management now happens at a speed and accuracy we’ve not seen before. Predicative analytics can signpost future trends, customer behaviour, and market demand, helping retailers make informed decisions and stay ahead of the competition.
For consumers, more relevant inventory recommendations, cheaper prices, and really clever tools like visual search, where customers can upload an image to find similar products, enhancing the shopping experience and product discovery.
The disadvantages are just as clear. The initial investment required to integrate AI systems is significant. Technology, infrastructure, and training can all be cost-prohibitive; the amount of data being analysed will present clear data privacy concerns; and the dependency or over reliance on technology leaves retailers exposed to cyber-attack and technical failures.
Retailers, and in fact all businesses, are of course accelerating AI integration for one key reason. AI-driven automation reduces staff costs, minimises errors, and optimises resource allocation, resulting in significant cost savings. While AI can clearly deliver efficiencies for retailers and can enhance the customer experience, we can’t ignore that it’s increasingly responsible for one of the greatest cost efficiencies for any business - staff. Automation driven by AI will lead to job displacement, particularly those involved in repetitive or low-skilled tasks. I’d like to think this presents a moral conundrum, but it would be naïve to think businesses won’t take every opportunity to reduce the need for people where AI can achieve the same, and often better, results for a far lower cost.
