AI's Retail Frontier: Jeanel Alvarado of RETAILBOSS

Jeanel Alvarado spent over a decade moving through fashion buying and marketing leadership before founding RETAILBOSS, a publication for retail executives focused on technology and e-commerce strategy. She's collaborated with Victoria's Secret and a roster of major retail brands, and her consulting work sits at the intersection of fashion and emerging technology. She has a practitioner's view of what AI in retail actually looks like beyond the hype.

AI's impact on Gen Z and millennial shopping

Alvarado's starting point is the shopping behaviour of younger consumer cohorts, who have grown up with search engines and social media as primary discovery channels. For them, the expectation is that finding the right product should be effortless — and AI is increasingly making that expectation achievable.

Visual search is her lead example. Rather than typing a description and hoping the search engine surfaces the right product, a customer can upload a photo — from social media, a street photo, a screenshot — and find matching or similar items across a retailer's catalogue. This reduces the vocabulary gap between what customers want and what search queries can articulate. It also significantly reduces return rates, because customers who find exactly what they're looking for are less likely to send it back.

"When you're buying cosmetics online, how do you find your shade? I'm sure you have probably maybe gone to a couple of our favourite beauty brands and now they have a find your shade feature. That's all driven by AI. So that helps you analyse a photo of yourself. It then is able to figure out based on your skin tone, what may be the best makeup or skincare product for you."

The beauty industry example is instructive because it's a category with high sensory purchase uncertainty — if you can't test the product, you can't know if the shade matches. AI addresses this by doing the sensory work computationally, removing a significant conversion barrier.

Beyond customer experience, Alvarado notes that AI-powered image analysis tools are helping retailers combat counterfeiting and retail theft — another dimension of brand protection that's particularly relevant as marketplaces and social commerce channels proliferate.

The data challenge: why open-source isn't enough

Alvarado is direct about the biggest barrier to effective AI in retail: data quality. Machine learning models are only as good as the data they're trained on, and for most consumer brands, first-party data is limited — especially for newer brands with a smaller transaction history.

The temptation is to supplement with open-source or third-party data. But Alvarado flags two problems with this approach: accuracy and legal complexity. Open-source data may be outdated, poorly labelled, or drawn from demographics that don't match your customer base. And depending on how it was collected, using it may expose brands to data privacy liabilities.

Her preferred approach: invest upfront in first-party data collection through surveys and quizzes. Ask customers about their preferences, skin type, body measurements, fit preferences, lifestyle — data that a purchase transaction alone can't capture. This creates the training foundation for personalisation models that actually reflect your specific customer base.

The innovation she wants most: seamless size translation

When asked about the AI application that would have the highest impact on her own shopping behaviour, Alvarado doesn't hesitate. The problem she describes is universal for online apparel shoppers: knowing your size in one brand tells you almost nothing about your size in another.

"Ideally, when I'm certain about my size — say, a 34 seam in a particular denim brand and style, it can vary depending on the materials, wash and treatments — it would be fantastic to have an AI-enabled function with the knowledge to translate a size seamlessly to other brands. Imagine knowing exactly which size on their chart would be a perfect fit for me! For example it could say go up a size for this, or go down a size. A company can get the data and insights from past customers who purchased and based on return/exchange patterns by previous buyers."

This is a tractable AI problem — the inputs are purchase and returns data (which retailers already have), and the output is a mapping function between brand-specific sizing systems. Several companies are working on this, but it hasn't yet become a standard feature across major retailers. The brand that makes it genuinely reliable first has a meaningful conversion advantage.

The next five years in retail AI

Alvarado's five-year outlook centres on a shift in consumer attitude toward AI and data. Right now, there's residual consumer apprehension — privacy concerns, scepticism about algorithmic recommendations, general distrust of how personal data is used.

She draws a parallel with sustainability: a decade ago, it was a niche concern. Now it's a mainstream purchase consideration. She predicts data ethics will follow a similar trajectory — moving from the margins to the centre of brand trust conversations. The retailers who build transparency around how customer data is used will be differentiated by consumer preference, not just regulatory compliance.

Her bottom line: AI's integration into retail is irreversible, and its pace is only accelerating. The question for retailers is not whether to engage with it, but which applications to prioritise. Her advice is consistent: "Keep up with the latest news and trends, there's so much happening." The window of competitive advantage from early adoption is short.