Harnessing AI for Retail Transformation: Insights from Re:Tech Disrupt’s Track 2
The retail and ecommerce environment has changed drastically with the onset of artificial intelligence (AI), a word that was once the future, but has now become the very, very real present. As such, Track #2 of Re:Tech Disrupt focused on the impact of data, AI and generative AI.
AI is not a new technology, it has been around for decades, but has only matured technologically today, with many advanced use cases. We explored the possibilities of how to use it actionably in the future, alongside real life examples of how brands and retailers use it across the board from operations to marketing.
Thank you to everyone who attended, and especially thanks to our speakers and sponsors. Here are some key takeaways from the panels, fireside chat, keynote and roundtable discussions.
Missed this one? We also have 2 other retail tech events coming up, check them out and register here.
Panel 1: AI for Revolutionizing Customer Experiences
Yaron Jacobs (Triumph), Yana Routshtein (Walmart), Mirko Saul (Schwarz Digital GmbH & Co.) and Ken Pilot (Ken Pilot Ventures, formerly GAP)
The overwhelming impact of AI on customer experiences is in search, followed by the checkout experience, relating closely to our first track on CX. All the brands that were on this panel stated that they were incorporating AI both via in-house developments as well by working with vendors and tech companies specializing in AI.
“Search is one of the most important things done by AI, to connect the customer with exactly what they are looking for, regardless of how they are describing it,” says Ken Pilot, making a clear statement that search is one of AI’s most useful benefits in retail
“Customers come with a mission, and an AI powered search can help to generate what they need,” says Yana Roushtein, Walmart. AI helps to plan things better with AI powered search: customers don’t necessarily come with one item to buy, but a list of items connected by their demographic and psychographics.
“AI integrates user testimonials into product descriptions to make them more compelling,’ says Yaron Jacobs of Triumph, showing how AI can help with marketing by incorporating the wealth of data that comes in customer reviews, in addition to using it in promotional materials and A/B testing of what promotions work.
The checkout phase is where AI is also used heavily, helping with improving the checkout verification, e.g. age verification for alcoholic purchases. Other areas in checkout that can be aided by AI are fraud prevention and customer service, almost replacing the classical chatbot which was limited in capabilities, or human representative who couldn’t be available 24/7.
AI naturally helps with warehousing and stocking, predicting what demand will be and working with supply accordingly. For example, Walmart is piloting an auto-replenishment system to help with the buying process.
“AI should be called Augmented Intelligence, not Artificial Intelligence; it helps in fixed rule environments: Self driving cars, medical diagnostics, etc. but its limitation comes in flexible rule environments, as AI cannot understand the meaning of what it does, and can only serve to support humans with sentient cognition under such circumstances” concludes Yaron Jacobs of Triumph.
Keynote: AI-powered Digital Avatars: The Next Frontier for Retail CX
Tomer Zuker (D-ID)
“45% of retailers will use Gen AI by the end of 2025”, says Tomer Zuker of D-ID.
Humans are wired to engage with humans, and over time technology was starting to replace this, which is now making a comeback. In terms of statistics, 75% of customers are more likely to watch vidoe text and images, 92% find their attention is driven with faces versus objects, products or images.
D-ID creates avatars that can engage with customers to replace the overall shortage of humans that are trained or qualified to engage with the mass of customers. The 150 million plus avatars can cater to customer experience to education to customer service and content creation, and everything in between.
Since the expressions of emotions in human like avatars can be perceived similar to human emotions, the effects are likely as positive as a human-to-human interaction. This is why agents are useful to help in all avenues of retail and commerce: personalized shopping assistants, virtual fitting rooms, customer service and support, managing feedback and reviews, and more common steps in the ecommerce landscape.
AI has been widely used in retail operational spaces as these can greatly impact profit margins in retail.
“AI is embedded in everything we do,” says Orr Hameiri, P&G, giving an example of using AI to help phase in-phase out process in operations when operations is needed to switch between products with multi-branded, multi-sku retailers and brands, resulting in direct cost savings. Another example is integrating ingredient lists into data for R&D to generative alternative formulas for categories like detergents, which makes innovation and development more efficient.
“AI is used in the back-end before coming to storefronts, and the efficiency has a direct improvement on the P&L” says Ohad Elzur, AS Watson Group. Most efficiency is driven from time savings, for example workers asking commonly asked questions more quickly, or giving recommendations to store managers to help them achieve their monthly targets.
There has been a lot of concern on displacing jobs due to AI, but overwhelmingly it doesn’t seem to be a displacement of jobs, but an opportunity for people to improve their skills. Plus, there will be a need to maintain AI platforms, creating more tech jobs, while soft skills like customer understanding, are irreplaceable.
.
Keynote: How Flipkart is Implementing AI Solutions Across the Customer Journey (+Opportunities for Startups)
Adi Lev (Flipkart)
AI is utilized in many aspects of Flipkart is a giant ecommerce marketplace from India, acquired by Walmart and operates in areas of online travel, fintech, healthcare, and retail. These include discovery (search and recommendation), user acquisition, user insights, customer experience, pricing and planning, and more.
For example, AI helps in pricing by taking inputs from cost, inventory, competition, traffic, brand, and more, computes the best offers, goes live with those across multiple timelines and preventing disruption. Then, AI systems measure performance relying on A/B tests and take these to further optimize the inputs, making it a virtuous self-learning cycle that is dynamic with customer demand and interactions.
The opportunity for startups Flipkart Innovation Network is looking to run an exclusive cohort to identify startups for a tech partnership, which includes both early stage startups and more mature ones looking to scale further.
.
Roundtable Discussions
The goal of the roundtable discussions was to delve into other states of retail tech and how they impact everyone from decision makers to shoppers in retail.
.
The Impact of AI on the Future of Retail – the long-term vision for AI in retail & its potential, and coming up with an AI implementation plan
Moderated by: Ohad Greenshpan (Wandz.AI)
There are a lot of wild theories of AI taking over all jobs and greatly impacting the way we interact, but to get there will takes many, many years, and in its current state, we can for sure take steps to improve AI in retail first.
AI is currently used to improve offline efficiency, as discussed during the panels and to provide real time action to quickly respond to customer needs and behaviors. Much of this is driven by cost and time available, but also based on the capacity of risk that a company is willing to take.
Predictive AI is also useful to predict not just operations in terms of supply and demand, but down to a customer’s next step or decision, based on their past decisions.
As AI grows, it also varies indifferent countries based on a country or culture’s adaptability for risk or AI, as well as how they expect retail and customer service. Some countries expect a very high level of customer service in terms of speed and efficiency, whereas others are used to a different style of communication. AI will thus have to adapt differently, especially for global companies.
Overall, the challenge of building an AI system is not in the actual setup, but in the appropriate analysis of data, given that there is just so much data in retail. This comes down to defining the inputs and outputs, identifying the key metrics, and measuring accurately to take action.
Retail Marketing with Generative AI – the use of GenAI for personalized ad & content creation, product recommendations, customer journey optimization and analyzing the impact in customer satisfaction
Moderated by: Sourabh Sharma (Histoires de Parfums)
While AI has been increasingly talked about in the operations and customer service landscape where it can help harness data for analytics and prediction, it is also very useful in retail marketing.
Personalization delivered by AI is all about data, however this brings up the element of privacy; so the key is to preserve privacy while still providing a personalized experience. Some companies as such don’t collect data, but analyze it and only report trends, thereby only delivering the insights.
AI can also be used for more than prediction, however if the inputs are skewed or biased in a certain way, as data often can be, the prediction can also be skewed. This is why AI can be better used to deliver informed choices, not just options and predictions, used for example in planning trips and travels.
AI can be used to more accurately build customer personas based on the wealth of data, to understand better the various customer segments that inevitably exist for ever brand. This then helps in marketing more accurately to these customer segments .
AI can help ameliorate the trend of product recommendations, many of which already exist, but make them smarter, relevant and more tailored than they have been.
Unlocking Product Innovation with Generative Design – how GenAI can help design new products, predict trends, and optimize existing offerings, including the material selection, packaging design, and personalized customization options
Moderated by Brandon Rael (Kyndryl)
Operations and R&D, as well as innovation, is a major playing field for AI. What are the challenges and ethical considerations in such process?
GenAI capabilities are maturing and providing value for both retailers and apparel manufacturers
GenAI capabilities are being leveraged for virtual try on, and right sizing solutions for custom clothing, something of great consideration given that the number one reason for customer dissatisfaction and returns in retail is inappropriate sizing.
Personalization is becoming extremely important in the retail landscape, as consumers want options and results that are tailored to them. Amidst the wealth of data, GenAI is a key enabler.
Actionable and clean data sources are critical for growth and to enable GenAI to drive value.
In Conclusion
Overall, the data and AI track highlighted the transformative impact of AI on the retail and ecommerce sectors, emphasizing its practical applications across various aspects of operations and customer experiences. AI’s evolution from a futuristic concept to an essential tool has enabled brands to enhance search functionalities, optimize checkout processes, improve inventory management, and personalize marketing efforts. Our speakers underscored AI’s ability to drive efficiency and innovation while addressing ethical considerations and potential job displacement concerns. They also showcased actionable insights and real-world examples, illustrating AI’s pivotal role in shaping the future of retail.
Yael is the CEO at Re:Tech, Israel's biggest community of retail related tech companies and industry leaders. She is also the CMO of buywith, the leading Livestream shopping platform, and an advisor to ReturnGo and Belle AI.