How AI helps small e-commerce sellers decide what to produce

How AI helps small online sellers decide what to produce

MIT Technology Review has shown how small online sellers use artificial intelligence capabilities to decide what products to produce next — a question that has remained one of the most expensive and risky for small brands over the years. As a starting point, the article tells the story of Mike McClary (Mike McClary), who for years sold through his small outdoor brand a Guardian LTE Flashlight — a sturdy model with a black body, designed for brightness and durability; over time, this product became one of the most popular in his assortment.

The story of one flashlight

According to the publication, McClary discontinued selling the Guardian LTE Flashlight around 2017, but this did not stop interest in the product: customers continued to send him letters asking where they could now buy this flashlight. Such a situation is a classic example of a problem familiar to almost any small goods producer: demand for a specific product often becomes obvious after the fact, when it has already been removed from production, rather than at the moment of making the decision about what exactly to launch for sale.

Key facts from the story:

• Product — Guardian LTE Flashlight, a heavy, durable model with a black body.
• Sold through a small outdoor brand of Mike McClary.
• Became one of the most popular products in the brand's assortment.
• Sales discontinued around 2017.
• Customers continued to write with questions about purchasing the product even after it was removed from sale.

How AI helps decide what to produce

For small online sellers, the main limitation has traditionally been money and time: full-scale market research, analysis of competitor reviews, testing prototypes and audience surveys — a luxury available mainly to large companies with separate marketing and product teams. AI-based tools are changing this balance: they make it possible to analyze massive arrays of customer letters, reviews, social media discussions and search queries, identifying demand patterns that previously required months of manual analyst work. Such signals — from repeated customer questions about a discontinued product to trends in related categories — become the basis for more informed decisions about what exactly version of the product to launch next.

Why this matters for small business

For the owner of a small brand like the one described in the flashlight story, each decision to launch a new product is a real financial risk: bulk orders, production, warehouse, marketing. A mistake in demand forecast could mean money frozen in unsold inventory, while a missed signal about real customer interest — lost revenue that would otherwise go to competitors. Reducing the cost of quality market analysis thanks to AI levels the playing field between retail giants with staff of analysts and owners of small niche brands who historically made such decisions more intuitively than on the basis of data.

The Guardian LTE flashlight story is also indicative in that customer letters themselves have always been an accessible source of information about demand — the problem was not a lack of data, but the fact that a small team simply did not have the time and resources to systematically analyze them and turn them into specific product decisions. This is exactly the routine but labor-intensive work — reading and categorizing thousands of letters, reviews and mentions on the network, identifying recurring requests in the noise — that modern AI-based tools are capable of taking on, leaving the business owner with the final decision but saving him from the need to manually review each message to notice a pattern that could otherwise remain unnoticed until it was too late.

A separate layer of capabilities opens up where demand analysis connects with generative design tools: instead of just recording that customers want a returning model of goods, a small brand can more quickly test hypotheses about product variations — color, material, configuration — even before investing in production of a trial batch. For companies the level of Mike McClary and his outdoor brand, this fundamentally changes the speed of iteration between idea and launch decision, reducing the gap that previously was available only to large players with full-fledged market research departments.