Pratik Patel

Unified omnichannel commerce is the only way to be! Some research-backed hard numbers: - 22% lower TCO on average - 20% faster POS implementation - Operational efficiencies equivalent to 5% sales uplift - Revenue enhancements equivalent to 8.9% sales uplift 

5

1 Comment

The shift towards self-checkout in shopping and digital services is driven by a desire for efficiency. Millennials and Gen Z lead this trend, valuing the autonomy and speed of self-service. Older generations tend to stick to traditional checkout methods. Beyond convenience, some self-checkout enthusiasts are drawn to the benefits of privacy, reduced social interactions, and concerns about job displacement in retail. This shift extends beyond retail to banking, where customers are increasingly adopting digital services and self-guided transactions. The rise of self-service technologies reflects a larger cultural move towards automation in multiple industries. #selfcheckout #retailbanking #digital #digitalservices #payment #payments #fintechs #fintech #banking #safety #security #banks #customer

3

Saks+Neiman Marcus+Amazon+Salesforce+Banks TLDR Saks and Neiman Marcus are merging. Amazon and Salesforce are both investing in the transaction. A timely insight about how important the software that powers retail is for retailers. Are you AI ready?

19

3 Comments

🚀 Similar Product Matching in Retail: The Power of AI Almost every conversation I have had with pricing and category teams in the recent past has veered towards one question: "How do we improve the match coverage of similar and substitute products?" Product matching is a core challenge in retail, involving grouping similar products across inconsistent names, descriptions, and formats. From differing retailer terminology (e.g., “MacBook Pro 2023” vs. “Apple 2023 Laptop”) to unstructured data (free text, tables, images), traditional matching methods often fall short. Example: "Apple 2023 MacBook Pro Laptop M3 Pro chip with 11‑core CPU, 14‑core GPU: Built for Apple Intelligence, 14.2-inch Liquid Retina XDR Display, 18GB Unified Memory, 512GB SSD Storage; Space Black" might not get matched with"Apple MacBook Pro 2023 (14.2 inch, M3 Pro, 18GB, 512GB, macOS, Space Black", making exact matching itself hard. A classical scoring function like Fuzzy Similarity scores their similarity at 43 -- even when they’re identical! 🤯 The situation is only exacerbated when it comes to matching similar products due to the category specific nuances and subjectivity. Consider the images below. While they are not the same obviously, they are much closer to each other than a traditional hash based scoring function deems them to be. This can happen due to different backgrounds, lighting, orientations, quality, and a host of other reasons. 🔍 How LLMs + RLMs Transform Product Matching: - Semantic Understanding: Interprets meaning over keywords—“6.5-inch display” ≈ “6.5-inch screen.” - Attribute Extraction: Identifies core details like brand, size, color from noisy data. - Implicit Relations: Recognizes product families (e.g., "iPhone 12" and "iPhone 12 Pro"). - Synonym Context: Links terms like "Sneakers" = "Running Shoes." Plus, CLIP embeddings boost multimodal matching accuracy, handling images with varying orientations, backgrounds, or overlapping objects. If we go back to the earlier examples and use CLIP embeddings, we get similarity scores of 89 and 86 respectively -- much closer to reality. 🌟 With AI-driven product matching, we help retailers achieve smarter, more accurate groupings—bridging gaps and driving better user experiences. #RetailTech #AI #LLMs #ProductMatching

26

To say this has been a big week in the world of shopping would be a massive understatement. It’s not just #CyberWeek— Perplexity just launched its groundbreaking shopping capabilities, enabling users to “Buy with Pro” and “Buy with Prime.” Meanwhile, Sundar Pichai recently revealed, as captured in The Verge article, that Google Search will “change profoundly in 2025,” as well as exchanging amusing barbs with Satya Nadella. Sundar’s response to Satya saying Google "should" have been the “default winner” in the AI race challenge was “I would love to do a side-by-side comparison of Microsoft’s own models", pointing out that Microsoft is “using someone else’s models.” 🔥 😂 Of course, Jeff Bezos is getting in the mix too. As highlighted in a recent Medium article, "Bezos’s investment in Perplexity suggests a shift in shopping dynamics. The future might not be about traditional marketplaces, but instead about conversation-first interfaces that understand context and intent. Bezos is betting that the next generation of shoppers will prefer talking to AI about their needs rather than scrolling through endless product listings—and I tend to agree." The race to dominate product discovery is heating up, and we’re seeing a massive disruption in online shopping. AI-powered “Answer Engines” like Perplexity AI are challenging the longstanding dominance of Google and Bing’s “Search Engines.” Retailers must act now to ramp up their Generative Search Optimization (GEO) efforts and augment their existing SEO strategies. If you haven’t addressed the fundamentals of your product data yet, you’ll find it increasingly difficult to keep pace in this rapidly evolving landscape. The Lily AI team is here to help you navigate these challenges. Feel free to reach out! In the meantime, here are some valuable resources for inspiration and insight into what’s working for leading retailers like ASOS.com, SEPHORA, Tapestry, Bloomingdale's, and Lowe's Companies, Inc.: - How Retail AI is Moving the Needle for Leading Retailers (by our very own Purva Gupta) - 2024 Retail AI Index: Discover the Top 100 Trailblazers (https://lnkd.in/e24qgBzF) We’re just scratching the surface of what AI can do to transform shopping, and it’s exciting to see the shift unfold in real-time. We’re looking forward to navigating it together with our clients! https://lnkd.in/ekj_bk6s

69

3 Comments

For a company that has thrived for over 218 years, Colgate-Palmolive continues to innovate with its supply chain technology. Forbes recently spoke with Colgate’s CSCO, Luciano Sieber, about how the company approaches AI with a focus on practicality and scale. Here are 3 key takeaways: 1️⃣ AI is essential: 89% of CSCOs surveyed by Zero100 believe AI is mandatory for supply chain transformation. But Colgate emphasizes a balanced approach, avoiding "science fair" experiments and focusing on AI's measurable impact on growth and efficiency. 2️⃣ Successful AI applications: Colgate's use of traditional AI in manufacturing, predictive maintenance, and replenishment planning, and generative AI for enhancing e-commerce experiences, is setting new standards in the industry. 3️⃣ Not just surviving but thriving: Colgate's AI strategies have improved productivity and contributed to a 29.7% increase in its share price. Colgate's steady performance and innovative mindset show how AI can unlock untapped opportunities across the supply chain, transforming processes and boosting business value.  #SupplyChain #IndustrialAI #DigitalTransformation

6

1 Comment

This is an awesome new feature! Basically it allows each store or multi-store the ability to have brick and mortar locations for customers to pick up items OR be limited to ship products to. Very cool and very useful for #B2BCommerce use cases! #B2Becommerce #B2B #ecommerce #featurerelease

7

1 Comment

This is the first of a series of weekly articles exploring retail convergence and the dramatic changes likely over the next decade.  Each will explore retail transformation; reviewing indicators which will drive these changes. 𝗙𝘂𝘁𝘂𝗿𝗲 𝗼𝗳 𝗥𝗲𝘁𝗮𝗶𝗹 Over the last 2+ decades, e-commerce has mimicked traditional brick & mortar retail.  Consumers shop with a product in mind, purchase it and then arrange getting it home. While retail digitization has moved rapidly and encompasses nearly every product traditionally purchased in person, it hasn’t changed the flow of a transaction. The natural question is “why would it?” or even “why should it?” Simple answer: e-commerce isn’t fully addressing consumer needs.  While an old shopping paradigm's now faster and efficient, consumer buying isn't much better or pushing the boundaries of what's possible.  Narrowing margins have slowed adoption of new tech and automation but the definition of retail is the real limitation. 𝗘-𝗰𝗼𝗺𝗺𝗲𝗿𝗰𝗲 𝗳𝗼𝗰𝘂𝘀𝗲𝘀 𝗼𝗻 𝗶𝗻𝘀𝘁𝗮𝗻𝘁 𝗴𝗿𝗮𝘁𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻. Speed is a primary objective for nearly every retail transaction. Consumers tend to favor product acquisition time over nearly all factors aside from cost and retailers have done little to change this mindset. There’s been little consideration for the actual consumer’s need or the consequences of ‘fast’ (beyond $). 𝗘-𝗰𝗼𝗺𝗺𝗲𝗿𝗰𝗲 𝗶𝘀 𝗼𝗽𝘁𝗶𝗺𝗶𝘇𝗲𝗱 𝗮𝘀 𝗮 𝘁𝗿𝗮𝗻𝘀𝗮𝗰𝘁𝗶𝗼𝗻. This is the most impactful limiter of retail today. Some companies consolidate shipments for closely placed orders and at thru fulfillment and delivery flow, but most companies have not. Even patches over a broken process haven't stemmed a flood of small requests and the consequences. 𝗘-𝗰𝗼𝗺𝗺𝗲𝗿𝗰𝗲 𝗵𝗮𝘀 𝗶𝗻𝘁𝗿𝗼𝗱𝘂𝗰𝗲𝗱 𝗼𝗿𝗱𝗲𝗿 𝗰𝗼𝗻𝘀𝘁𝗿𝗮𝗶𝗻𝘁𝘀. Consumers have few options that decide the way they buy. For the most part, a consumer can choose a direct seller or marketplace.  This decision encapsulations the unit which will be operated on and limits improvements beyond the entity handling the request. 𝗘-𝗰𝗼𝗺𝗺𝗲𝗿𝗰𝗲 𝗵𝗮𝘀 𝗰𝗿𝗲𝗮𝘁𝗲𝗱 𝗲𝘅𝗲𝗰𝘂𝘁𝗶𝗼𝗻 𝗽𝗶𝗽𝗲𝗹𝗶𝗻𝗲𝘀. E-commerce has resulted in fulfillment and delivery silos. Retailer decisions have decided vendors and created bespoke pipes. Vendors optimize for individual shipments and profitability. These decisions prevent focus on broader goals like large scale optimization, environmental improvements and better consumer experience. 𝗘-𝗰𝗼𝗺𝗺𝗲𝗿𝗰𝗲 𝗵𝗮𝘀 𝗶𝗺𝗽𝗿𝗼𝘃𝗲𝗱 𝗱𝗲𝗹𝗶𝘃𝗲𝗿𝘆 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆 𝗯𝘂𝘁 𝗶𝗴𝗻𝗼𝗿𝗲𝗱 𝗯𝗲𝘆𝗼𝗻𝗱 𝘁𝗵𝗲 𝗱𝗼𝗼𝗿𝘀𝘁𝗲𝗽. Some companies claim CX obsession but what has really changed? Packages arrive and the consumer must deal with the consequences. 𝗨𝗽 𝗻𝗲𝘅𝘁: we will explore changes breaking existing constants and dig into the evolution of consumer behavior, product definition and changes consumer demand fulfillment.  #retail #ecommerce #supplychain #technology

47

2 Comments

I love when our customers ask us for case studies. It’s what inspired Lily to go beyond our own customer case studies to analyze the entire state of retail adoption of #AI in our inaugural 2024 Retail AI Index. Of course, the Retail AI Index is chock full of impactful AI applications, many of which are powered by Lily’s own customer-centric product attributes. Our award-winning attributes provide a critical base layer for personalized product discovery and smarter search. But it’s peak, and if you’re in retail you don’t have time to read a 33+page report! So sharing the latest AI case study roundup from The Robin Report which highlights some of our favorite use cases from the Index, including how retail AI is moving the needle for ASOS, Sephora, Lowes, and others. Lily AI is helping many of those retailers drive the kind of sales impact that actually moves the needle. It's why Lily was selected as a finalist in the Modern Retail Awards as a "Best Use of AI" and why we recently made the Deloitte Technology Fast 500. It works! 🚀🚀🚀

52

Inventory management has always been a challenging task for retailers. However, Instacart has innovated in a way that removes the need to maintain store inventory separately for online orders. By allowing Instacart shoppers to come in store and fulfill orders themselves, retailers can save time and resources. But what about in South Korea? Are there any innovators disrupting this space? Shoot me a message if so. I am ready to buy in. #inventorymanagement #retail #innovation #SouthKorea #instacart #emart #retailinnovation #storeinnovation #inventory #inv #ai #ax #digitaltransformation #dt #dx #artificialintelligence #delivery #onlinedelivery #ecommerce #openai #gpt4 #genai #startups #entrepreneurship #innovation #EmartTech

1

Are in-store purchases more sustainable and incur less carbon emissons than eCommerce purchases? The answer may surprise you. Online shopping can often be the greener option to stores, provided there's "Willingness to Wait" as I elaborated on my post: https://lnkd.in/eEPq64hR. Thanks for the messages and comments. To read the full thesis published by MIT CTL researcher, Dimitri Weideli, go to: https://lnkd.in/eFgZ4vEq In another MIT research on inbound logistics (https://lnkd.in/eusGtJW7), it was found that fast shipping produces significantly higher CO2 emissions as it may prevent cargo consolidation. To decarbonize meaningfully, track your carbon footprint effectively - visit www.finboot.com to read more.

16

1 Comment

We all know the struggle of sifting through lengthy descriptions and insufficient details when making buying decisions. That's why our team is leveraging GenAI to develop new item summary and review capabilities. These new tools capture the essence of customer sentiment for the most spoken of product aspects within the thousands of textual reviews allowing customers to quickly understand the overall sentiment and make more informed decisions. But wait, there's more! Customers also have the power to choose any aspect they want to read more about by clicking on their preferred aspect to access a second-level summary with more specific details and sentiments. This is another testament to how we’re putting our customers first and enabling better and more efficient solutions making shopping easier and more enjoyable for all. #WalmartProduct #ProductManagement #CustomerExperience #Innovation #GenAI

212

11 Comments

2024 is already faster than 2023 ⏩ On yesterdays Amazon's earnings call Andy Jassy commented again on how they are making Prime faster. "In this past Q1, we delivered to Prime members at our fastest speeds ever. In March, across our top 60 largest US metro areas, nearly 60% of Prime members orders arrived the same or next day. Faster delivery times have another important effect. As we get items to customers this fast, customers choose Amazon to fulfill their shopping needs more frequently, and we can see the results in various areas, including how fast our everyday essentials business is growing and the continued increase in Prime member purchase frequency and total spend with us." This is why consumers abandon cart when they see a 4-7 business day delivery SLA when shopping off Amazon. Why wait over a week for something I can buy for the same price on Amazon and get it today or tomorrow? Doug Herrington (Amazon Stores CEO) talks more about their obsession with faster delivery in this video 📽

5

This is a wake-up call to all retailers and brands CEO/CDOs/CMOs, you are going in the WRONG direction. Stop competing with commodity commerce sites like Amazon and Walmart and start innovating. You all “say” the right things at trade shows: “better customer experiences” “personalization” “real connections and brand engagement” Yet your actions are the OPPOSITE You don’t encourage your teams to take any risks or come up with creative ideas. You have teams of paper pushers. Everyone is looking to press a button or implement the newest Acme Company technology (yes, a Roadrunner cartoon reference). You are looking for the latest AI bot to do the trick Newsflash, consumers hate bots…no matter how smart they are You are cutting programs left and right, reducing customer service hours, getting rid of free returns You are being lazy You're playing defense not offense You need to do some real work, some elbow grease type stuff Case in point: When Domino's pledged to get you your pizza in under 30mins – that took effort and it worked. It was all about the customer experience. When Enterprise said “we’ll pick you up” – that wasn’t a push the button effort, it was a program and executed well. When my friends at Sweetwater promise you access to a real musician expert 1:1 to help buy the right products, that was a real effort (and it worked, driving their DTC to $1.7B) The biggest risk is the risk you don’t take. NEED Help? Here’s an example of an idea to inspire creative thinking – it may be a terrible idea but I want to get the point across on how you NEED to be dissecting the customer experience and coming up with creative solutions I call it Pizza Party - that will both drive conversion and reduce returns for apparel brands Hire fit models of varying sizes and body types near you warehouse For every new drop, have them try on the clothes and record their fit feedback Share that feedback on every PDP page Have fun with it, make your fit models ambassadors and show their personalities. 6 fit models $500/a day x 3 days of work = $7,500 1 Producer at $500/day x 3 days - $1,500 8 times a year = $72,000 Web manager to project manage this whole thing - $100,000 For $172K a year An apparel brand doing $50M GMV a year with 30% returns – that’s $15M in lost revenue + probably another $1M in return logistics A 2% drop in returns would recover approx. $1M in revenues Add another 2% conversion lift and that’s another $1M in revenues Net net - $2M gained for $172K spent Sounds like a pretty freaking good ROI to me any my CVR and reduction in returns #s are conservative  Plus you’ll get some amazing content, brand ambassador, etc etc That’s old school innovation stuff You get my point – start innovating.

54

9 Comments

Modern online shoppers won’t trust a slow eCommerce site. It’s really that simple. Backend metrics are often overlooked within technical teams because user-facing IT delivery metrics are easier to track and seem more impactful. But, keeping backend metrics in the safe zone is crucial for scaling and maintaining a high-performing site. When it comes to eCommerce, the devil is usually in the backend. I created a cheat sheet to help your site stay in the safe zone. 👇 But here are 3 ways AI can enhance tracking of website backend metrics: 1. Anomaly detection through AI algorithms that analyze patterns in backend metrics like server response times, database query performance, and resource utilization. By learning normal behavior, AI can quickly flag anomalies or unusual trends that may indicate issues, allowing for proactive troubleshooting. 2. Predictive maintenance by analyzing historical data and current trends. AI can predict when systems will likely experience problems or reach capacity limits. This allows for preventive actions like scaling resources or optimizing code before issues impact users. 3. Automated root cause analysis: When problems occur, AI can correlate data from various backend systems and logs to identify the root cause more quickly than manual analysis. This can significantly reduce the time needed to resolve incidents. Keen on diving deeper into the subject? Send me a DM, and we can talk! ✉️ #management #ai #retai #ecommerce #digitaltransformation #RemoteNative Andreas Anding - Technical Leader with 20+ years of experience in large-scale digital transformation projects.

39

1 Comment

When the CEO of Target (Brian Cornell) says it's all about listening to consumers, other retailers should take note. Using #voiceofthecustomer #data combined with #ai and #predictiveanalytics helps retailers to make smarter decisions using their own consumers or their aspirational customer! First Insight, Inc. #retailtransformation

23

1 Comment

A huge shout-out to the exceptional engineering team working with me and my 'partner-in-crime' Piyush Garg for bringing this technically challenging project to life at JFL in just two years. Building a complete POS ecosystem that seamlessly integrates with our cloud-powered platforms when connected to the internet—while also functioning as a truly distributed system without a central coordinator or database during internet outages—was a monumental task that could have easily overwhelmed any other team. We faced all the challenges typical of distributed systems, such as ensuring fault-tolerance and reliability within a network of replicated in-app embedded databases, maintaining system safety and liveness under network asynchrony and partial failures, eliminating issues with distributed transactions, and ensuring consistency and availability at all times. Additionally, we had to prevent split-brain scenarios between cloud data-stores and in-app embedded databases and make the system fully observable down to the minutest details. To make things even more complex, we had to solve these challenges not just at the backend cloud layers—where these problems have been addressed for decades—but at the mobile app level, where constrained resources and the lack of ready-made solutions compounded the difficulty. Despite these obstacles, our team not only survived but thrived, delivering an amazing product that adds immense value to the JFL ecosystem. This achievement aligns with the vision of our CTO Pawan Kumar to transform JFL into a food-tech powerhouse. We will be soon posting a series of technical deep-dives here to explain how we accomplished this feat..keep tuned ! Credits : Manish Verma - for exceptional work leading the Android Team GOURAV THAPA Shashi Mehra Sumair Syed - for the wonderful work on ReactNative, ReactJS apps and the websocket gateway Vaibhav Tuteja,Deepak Sharma,Harsh kumar - as awesome backend engineers and tech leads Not to forget Ashwani Kuntal , Animesh Vibhuti and Susheela Bora for laying out some of the cloud-platforms earlier that have been leveraged in this POS journey A big thank you to Vidit Garg , Arjun Swarup , Yatish Upadhyaya, AYUSH AGARWAL guiding the product and ensuring Quality as always. Special mention for our infra and devops team under the professional guidance of Sandeep Bajpai for creating the huge infra powering this ecosytem. The exceptional UI of our in-store apps like OrderTaking, Kitchen Display, Lobby would not have been possible without Manu Pratap Singh and Darpan Bajaj Special thanks to all our junior engineers and freshers(GETs) who are the real backbone of this project :-) Jayant Sharma Hardik Tuteja Atharrva Hande T Kumar Vaibhav Akshi Chauhan and so many more If I were to name one person without whom this project would have definitely failed it would be without doubt for me my dearest friend Piyush Garg, you were the real-pillar for this project throughout its ups and downs, respect !

71

10 Comments

The Future of Retail: Embedded Finance Report Review Official Link: https://lnkd.in/gYsBassS This insightful report delves into how integrating embedded finance solutions—like Buy Now, Pay Later (BNPL), and loyalty rewards—can transform retail platforms. With 99% of surveyed marketplaces recognizing the importance of these offerings, it is clear that embedded #finance is becoming essential for enhancing customer experiences and driving innovation. Key Highlights: - Long-Term Vision: Most retail platforms anticipate enhancing embedded finance features over the next 2-5 years, focusing on loyalty programs and digital wallets. - Consumer Expectations: As digital wallet #payment options become standard, retailers must adapt to meet customers demands for seamless transactions. - Strategic Importance: The integration of embedded finance isn't just about payments; it’s about deepening customer relationships and attracting new #users. This report is a vital resource for businesses looking to stay competitive in an evolving marketplace. #embeddedfinance #retailinnovation #digitalwallets #BNPL #customerexperience

11

My last 20 years of experience with the retail industry taught me that ‘apparently simple’ problem of ‘demand forecasting’ has never been solved. Typical problems of ‘out of stock’ and ‘over stock’ are still prevailing in our high tech retail tech world of AI. My recent discussion with couple of my CIO friends reconfirmed my hypothesis that ‘demand forecasting’ unless supported by ‘supply chain’ and ‘last mile logistics’, it is bound to fail. The challenge does come from technology infrastructure. It really comes from understanding of data, interpretation and using multivariate analysis to a decision path and outcome. McKinsey’s article clearly emphasizes the lessons we learn in this regard. #retail #retailtech #AI #logistics #supplychain #demandforecasting #overstock #outofstock #deeplearning #Mckinsey #epam

31