Retail is entering a phase where shoppers no longer search – they ask.
Instead of browsing websites or comparing options across tabs, consumers are turning to social platforms and recommendation engines to decide what to buy. A single prompt can now replace an entire shopping journey, whether that’s ‘best office desk under £200’ or ‘running shoes that can arrive before Friday.’
This shift is happening quickly. Adobe Analytics reported that traffic to retail sites from generative AI sources grew by over 1,200%, while Capgemini found that more than half of consumers are already using or willing to use AI assistants for shopping decisions.
Products are now filtered and prioritised before a customer reaches a dedicated site, with decisions often shaped in advance – meaning retailers need to move quickly to secure their place in that shortlist.
Discovery depends on data that systems can trust
In this new world, AI systems act as gatekeepers. They determine which products are shown based on how reliable and complete the available information about them is.
That includes clear signals such as stock availability, delivery speed and accurate product attributes. For example, a shopper asking for ‘children’s winter coats available in size 7–8 with delivery before next week’ will only see options that meet all those conditions with certainty. A retailer with delayed stock updates or unclear delivery messaging will not be included, even if they carry the right product.
In categories like home improvement, the same principle applies. A customer searching for ‘paint available for click-and-collect today’ depends on real-time store-level availability. If that data is not accurate, the seller simply will not appear as an option.
Gartner estimates that poor data quality costs organisations an average of $12.9 million per year. AI-led discovery environments amplify that impact – products that aren’t surfaced are effectively removed from the sale.
Customer expectations continue to rise
While discovery is becoming faster and more precise, expectations around delivery are rising just as quickly.
Customers assume that what they are shown reflects reality. If a product appears available for next-day delivery, they expect it to arrive on time – without updates, delays or last-minute changes. When that expectation is broken, the impact is immediate.
These are rarely low-stakes purchases. A parent ordering school supplies days before term starts is working to a fixed deadline. A homeowner replacing a broken kettle needs a solution that day, not later in the week. In these moments, convenience turns into dependency, and any disruption feels disproportionate.
Research from Narvar shows that nearly half of consumers will reduce spending on a brand after a late delivery, underlining how quickly trust can be lost when expectations are not met.
Operational alignment is now a front-line priority
Many retailers are still operating on infrastructure that was designed for a different pace of commerce.
Inventory data may sit in one system, supplier updates in another, and delivery timelines elsewhere. Each updates independently, often with delays or manual inputs, which creates gaps between what is shown and what can actually be fulfilled.
In an AI-driven environment, those gaps become critical. Systems making recommendations need a consistent, up-to-date view of availability.
For instance, a warehouse selling garden equipment may depend on third-party suppliers for larger items such as sheds, furniture or lawn machinery. If supplier lead times are not updated in real time, a product may appear available for delivery within a few days when, in reality, it will take much longer to arrive.
The same applies to store collection. If a customer is told an item can be picked up that afternoon, that decision depends on accurate, store-level stock data. Duplicate counts or a generalised national view can result in customers being offered options that are not actually available at their chosen location.
What’s needed is not just better data, but better coordination of that data. This is an orchestration challenge. Retailers don’t necessarily need to replace existing systems, but they do need infrastructure that connects them – allowing updates to flow across suppliers, warehouses and channels in near real time.
Without that foundation, AI systems are making decisions on incomplete information and, in most cases, those decisions will favour competitors with clearer, more reliable signals.
Retailers that sync discovery with delivery will lead the next phase
As AI becomes a primary way consumers discover products, the rules of competition are shifting. Price and range still matter, but what increasingly sets brands apart is how reliably they can meet expectations set before the purchase even begins.
Those that consistently provide clear and accurate information will be prioritised and trusted. Those that don’t risk being filtered out before the customer ever has a chance to even consider their products.
