Personalization 2.0: How AI is Making Affordable, Hyper-Personalized Commerce a Reality

Key Takeaways

• Consumers across all economic demographics are increasingly price sensitive and looking for affordable products. 
• At the same time, consumers are seeking hyper-personalized commerce experiences. 
• However, hyper-personalization has historically been reserved for premium experiences and been prohibitively expensive due to brands’ operational constraints. 
• AI is transforming how brands optimize Consumer Discovery, Supply Chains, and Product Development, resulting in hyper-personalized products and experiences at significantly reduced costs.

‍

Consumers’ budgets are tightening

After recent accelerated consumer spending, the signs are clear that consumers are tightening their discretionary budgets. Consumers built up excess savings while confined at home during the pandemic, supported by government stimuli. Total savings reached a peak of $2.1T in August 2021—more than $8k per U.S. adult. These savings were steadily depleted over the subsequent three years as consumer spending rose by 9.2% and 5.9% in 2022 and 2023, respectively. These savings were fully exhausted by March of this year yet consumers are now only saving 2.9% of their income, a low not seen since 2007. 

‍

Meanwhile, recent high inflation and a cooling labor market have further constrained discretionary spending, and the effects are being felt across the consumer ecosystem. Restaurant chains like McDonald’s, Burger King, and Domino’s and retailers like Target and Nike are responding to this increasing price sensitivity by holding off on price increases, dropping prices, or offering discounts. Additionally, Walgreens, Craft, and furniture stores like Michael’s and Ikea have also dropped prices on popular items. From fashion to hospitality to personal care, Americans are increasingly seeking budget-friendly alternatives.

‍

Demand for affordable products is shaping new spending habits

Within retail and fashion, Temu and Shein have emerged as leading shopping destinations in the U.S., topping app stores and creating a frenzy with consumers, underscoring a profound shift in consumer shopping behavior. Both known for low prices, Shein initially created a cult following for its fast-fashion apparel and has since branched out into other offerings, such as home goods, while Temu runs a marketplace for virtually everything from home goods to apparel to electronics. These platforms keep costs low by shipping items directly to consumers from manufacturers in China, and leveraging China’s low-cost manufacturing and inexpensive labor (more on this later). Shein’s novel approach to R&D, small batch testing, and leveraging real-time consumer insights from social platforms like TikTok and Instagram, helps them to identify emerging trends and compresses product development cycles.¹

‍

Temu and Shein by the numbers:

• Since its launch in September 2022, Temu has reached $15B in total GMV and >250M downloads in 2023. The U.S. accounted for approximately 40% of total downloads, making it the most downloaded app in the U.S. that year. Temu made a significant splash with six advertisements during the Super Bowl this year (each costing around $7M for a 30-second spot), and also offered $10M in giveaways. 
• In 2023, Shein generated $33B in revenue and amassed 238M downloads, with 20% of its active shoppers based in the U.S. By the end of 2022, Shein accounted for 50% of U.S. fast-fashion sales, compared to 12% in January 2022. 

‍

The data is clear: Temu and Shein have answered U.S. shoppers’ call for affordable products and penetrated cultural consciousness at a scale and speed not seen since platforms like TikTok. Both have bet that consumers are willing to sacrifice immediate shipping and comprehensive customer service for affordability. This approach inverts the accepted wisdom of the past 15 years where the ‘Amazon Prime-ification’ of consumer has placed the utmost importance on features like next-day delivery and seamless returns. Notably, despite their reputation for bargain prices, a substantial portion of their customer base includes high-income earners and Temu’s popularity is growing the quickest with people making $130,000+ per year, signaling a broad appeal for low-priced items across economic demographics.

‍

Temu and Shein are now challenging the dominance of major retailers like Amazon which is now developing its own discount service and H&M which is speeding up production times to compete. Additionally, Shein’s move to sell its technology to other brands further underscores this ‘Sheinification’ of the retail industry. 

‍

Temu and Shein’s popularity is also intertwined with a growing "dupe culture," (short for “duplicate”) where consumers seek alternatives that replicate luxury products at lower prices (in other words, knock-offs). Websites like Dupe.com, which let you instantly search any product for cheaper dupes, have seen a sharp increase in traffic, particularly from the U.S., as young consumers strive to maintain appearances of luxury while grappling with financial pressures. 

‍

At the same time, consumers expect hyper-personalized commerce experiences

In addition to affordability, American consumers increasingly expect personalized shopping experiences: 80% of frequent shoppers only shop with brands that personalize the experience and 81% of consumers expect to see “just for me” options over the next five years. As the variety and complexity of choices in ecommerce have expanded, the conventional product discovery process is being rethought. Instead of sifting through hundreds of SKUs, consumers want a curated selection of products they are likely to enjoy. Today, search pages are inundated with ads and are only getting more crowded with the proliferation of AI-generated content.  Images of yet-to-be manufactured products and content that mimics reviews or influencer testimonials are flooding the internet, all created by AI. More than ever, consumers need tools that cut through the noise and provide personalized experiences. 

‍

Hyper-personalization has historically been reserved for premium experiences

Starting in the early 2010s, many DTC  brands were founded on the premise that they could use in-house human expertise to make personalized products and recommendations. We saw Blue Apron and Hello Fresh deliver personalized meals; Curology for skincare; Function of Beauty for hair care; Trunk Club and Stitch Fix for personalized clothing and accessories. This first wave of brands offered personalization by assuming the basic human roles of a personal shopper, dietician, dermatologist, etc. Let’s call this ‘Personalization 1.0’.

‍

However, most of these Personalization 1.0 brands were very operationally intensive with three main inefficiencies:

‍

1. Consumer discovery reliant on human teams: Large teams of experts were required to consult with users and make recommendations (by 2019, 64% of Stitch Fix’s 8k employees were stylists, representing ~10-15% of total SG&A). Recommendations these experts make are often not personalized enough, leading to high return rates. Stitch Fix has turned to offering incentives to reduce returns, offering a 25% discount if customers keep all items in their box.
2. Complex, costly supply chains: Manufacturing and distributing personalized products requires a more complex and expensive supply chain than mass producing products, which benefits from repeatability and economies of scale. For example, Curology’s monthly formula starting cost is $29.95 vs comparable non-clinical products which are bought off the shelf for $5-10 because each customer’s skincare formula is formulated and packaged individually. Nike By You custom shoes cost $20 more than non-custom shoes because of the additional time it requires to create a custom design.
3. Inefficient product development: ~70% of fashion brands still rely heavily on creative designers to forecast trends and then mass manufacture products based on their recommendations. This often leads to a lag of 6-12 months between trend identification and product availability. If the designers overestimate demand this leads to overstock which requires discounting or liquidation which harms margins, and if they underestimate demand, products sell out quickly, leaving revenue on the table.

‍

Stitch Fix has never turned a profit and currently has a market cap of <$600M, down from a peak of $10B+ in January 2021. Similarly, Nordstrom bought Trunk Club in 2014 but could never make it profitable, so they shut it down in 2022.

‍

Optimizing consumer discovery, supply chains, and product development will be critical to deliver hyper-personalized products at significantly reduced costs. Let’s call this ‘Personalization 2.0’.

‍

Introducing Personalization 2.0

‍

Consumer Discovery: Curated and Individualized

With advancements in AI, shoppers are on the verge of accessing hyper-personalized search and discovery at significantly reduced costs. We are already seeing a wave of startups vying for a slice of the market:

‍

• AI-powered multimodal search engines are reshaping how consumers discover products online by taking users’ colloquial language or images and turning them into tailored product recommendations. Companies like Daydream, Plush, Deft, and Objective are leading the charge in this category. 
• Personalized landing pages and personalized CX APIs are emerging as a necessary category to personalize ecommerce discovery experiences. By leveraging intent and context signals like traffic channel, location, local weather, shopper browse/purchase data, shopper demographics, etc these platforms put the right products in front of the right visitors as opposed to every visitor seeing the same static landing page. For instance, companies like Fermat, Spangle, Malachyte, and Psykhe are already doing this. While most of the innovation in this category has been focused on personalized websites, we’re also excited about AI’s potential to create adaptive app interfaces (e.g. imagine each user having a tailored version of an app depending on their frequency of interaction, tech-proficiency, location, spending behavior, etc). 
• Size-and-fit personalization tools offer accurate body measurements, size recommendations, and virtual try-on to address the leading cause of apparel returns—poor fit. Companies like AIMIRR and True-Fit are key players in this area. 
• Search-by-image apps are making it easier to discover products by enabling users to find similar products based on images or screenshots. Companies like Cherry are enabling this.
• Prompt-to-product marketplaces allow users to use AI tools to design one of one custom products. Torch portfolio company Arcade is enabling this by connecting consumers with independent makers who then bring their crafts to life. Platforms like Arcade bypass the traditional search process and put personalized search in the hands of the consumer by allowing them to seamlessly design their dream products themselves. 

‍

In this future, LLMs can replace the costly human involvement required for Personalization 1.0. In doing so, they can deliver even more precise and tailored recommendations. The combination will reduce COGS and increase conversion while reducing return rates. 

‍

Supply Chain: Fast and Flexible

Lessons can be taken from the success of Shein and Temu’s models and applied in new and innovative ways across the commerce landscape. Shein's low-batch manufacturing model leverages a network of manufacturers to produce and ship goods directly to consumers, bypassing traditional middlemen in the supply chain (e.g intermediary warehousing, fulfillment centers, ocean freight, etc). This approach enables rapid adaptation to market trends, improves inventory management, and increases operational efficiency, resulting in higher profit margins and lower prices for consumers. The success of this model is inspiring more startups to adopt similar strategies.

‍

• Direct-to-consumer shipping & small batch testing - Several early-stage companies are emerging in the direct shipment and small batch manufacturing space. Platforms like Portless and Trendsi facilitate low-batch manufacturing and direct delivery solutions for brands looking to implement the “Shein” fulfillment model. This model reduces inventory lead times and improves merchant shipping costs, cashflow, and inventory risk.
• On-demand production engines - We’re starting to see companies use AI to alter the cost structure of brands’ supply chains, enabling much faster and cheaper product development. For example, platforms like Vveave, Tapstitch, and Cala streamline the design process, automate the creation of technical packs (a historically painful and time-consuming process for suppliers), and leverage a decentralized network of suppliers to reduce manufacturing costs and production times.
• Dupes as a service - These are platforms that enable business’ to browse designs, select a factory, sample products, and produce affordable alternatives to luxury items. For example, platforms like Gembah enable brands and manufacturers to efficiently create dupe products. 

‍

Personalization 2.0 brands can adopt elements of the Shein model to compress product development cycles and increase margins. 

‍

Product Development: Data-Driven and Dynamic

Brands can harness real-time data to enhance product development. The best fast fashion retailers optimize production by using first-party data and consumer insights to identify trends and align products with consumer preferences rather than relying solely on creative designers. This shift marks a move away from traditional mass production to a more agile and responsive model, presenting significant opportunities for the personalization enablement ecosystem:

‍

• Algorithms to identify emerging retail trends - We’ve seen a number of companies trying to sell consumer trend data to retailers to increase personalization of their products. Trendalytics provides insights into shifting consumer trends across beauty, skincare, clothing/fashion, and more; Datacy analyzes search word intent, attribute signals, and content signals to help retailers extract deep online shopping behaviors among their buyer base; Heuritech uses AI to analyze 2,000 fashion attributes across more than 3 million photos per day to predict which fashion trends will experience revenue growth in the future.
• Inventory planning and demand forecasting software - Several companies are making inventory planning more predictive. Syrup and Toolio determine how much and which kind of inventory your brand needs at any given time based on transactions, marketing and inventory, social media trends, the weather, and other factors. This allows brands to make more informed inventory decisions and reduce stockouts and overstock. 
• Product research and testing platforms - These companies help brands validate product ideas and generate targeted consumer insights. Companies like Highlight are disrupting in-home user testing (IHUT) by recruiting target consumers and delivering physical products to their homes. Brands get feedback from these users within a few days, and they use those insights to make more informed product decisions. 

‍

Conclusion

Brands are now rising to meet consumers' dual demands for affordability and personalization—once a costly and complex combination. With AI powering streamlined Consumer Discovery, flexible Supply Chains, and data-driven Product Development, ‘Personalization 2.0’ makes tailored experiences accessible without the premium price. This shift allows consumers to find products that feel custom-made, while brands achieve lower costs and faster turnarounds. The result is a retail landscape where price-sensitive shoppers don’t have to compromise on experience, marking a new era of commerce that’s both dynamic and distinctly personal.

‍

Footnotes

1. The Biden administration is currently trying to close a tax loophole called the de minimis rule that Temu and Shein have benefited from, which exempts shipments under $800 from import tax. Although this rule change would have some effect on Temu and Shein, we think the impact will be minimal as the majority of the cost savings they realize are a result of cutting out unnecessary middlemen in the shipping supply chain. Additionally, Temu and Shein have begun to expand and diversify their supply chains to limit their exposure to regulatory shifts, including by expanding their U.S. operations. Therefore, Temu and Shein should still be able to maintain their affordable prices, even if the de minimis rule is reformed by Congress.