Developer: AI is transforming online store product search beyond recognition, and online retailers must respond fast

The rise of AI chatbots has rapidly changed the way people look for information: instead of googling keywords, they describe a problem to ChatGPT and let the AI come up with a solution in seconds.

According to Margus Eha, CEO of the software development company Opus, online shops need to keep pace, because consumers increasingly expect the search engines of web stores to offer the same intuitive experience.

“While Google used to be the main helper in online searches, there is now a clear trend of people preferring ChatGPT or other AI solutions instead,” said Eha. In his words, getting results with AI works a little differently: instead of keywords, people describe a problem or need in full, and the AI responds with solutions that take the context into account. Eha says this direction means changes are coming for the search engines of online shops too, where consumers increasingly expect product results to be given in answer to a problem, rather than a product category or a single keyword.

“AI makes it possible to understand customers’ needs better and to present products in a way that matches their expectations, and that brings better sales results for the merchant as well,” he explained. It won’t be long before a customer can tell a pharmacy’s online shop that their shoulder hurts, for example – and the AI will ask how the pain started and then recommend the right painkiller based on the answer. “If online shops ignore this trend, it may soon happen that customers can’t find the right product and abandon the purchase. This is especially true of the younger generation, who are more used to using AI,” said Eha.

AI works out for itself what is in the picture

Eha says that behind the scenes of an online shop, AI works on a different logic from traditional search systems. Where a shop would usually display search results based on keywords and categories that have to be entered manually for every product, a product catalogue enriched with AI works on context. “That means AI can analyse images itself and understand what they show, even if the product description doesn’t say so explicitly. If a customer is looking in a furniture web store for a lamp with a wooden frame, the AI can deliver the right results even when the product only has a photo and a lamp model attached,” he explained.

Another benefit of using AI is more relevant product recommendations for customers based on their previous purchases or searches. If a customer doesn’t find a suitable product at first attempt, Eha says the AI can react quickly and start offering more precise results, taking the user’s earlier searches and preferences into account. “It is also untrue that integrating AI is overly difficult. A good development partner can do it with fairly little effort, and right now is exactly the time to prepare for it. The first step is therefore to enrich the product catalogue with AI information, so that it understands the product’s properties and context better,” he explained, adding that this means feeding data to the AI. Once that is done, the next step is to enrich the search with AI as well, making the whole purchasing process much smoother and more effective.

According to Eha, the main reason to use AI is simple – customers find the product they need faster, and the chances that the consumer actually goes on to buy are much higher. “If search doesn’t deliver quick results, that is a lost sale for the shop. That’s why it’s extremely important that search works intuitively,” he said.