Conversational commerce: what Amazon's AI shopping shift means for your online store

The quiet shift in how people start a purchaseFor two decades, the online buying journey has followed a familiar pattern: a search box, a results page...

The quiet shift in how people start a purchase

For two decades, the online buying journey has followed a familiar pattern: a search box, a results page, a product page, a basket. That pattern is now being challenged from the inside by conversational assistants that sit between the shopper and the catalogue. Amazon has been the most visible mover here, folding its earlier shopping assistant into a broader AI companion that now handles far more than product search — from comparing items and tracking price history to managing recurring deliveries and updating delivery addresses.

The company has reported substantial growth in the number of people using the assistant and in the volume of interactions, and it notes that shoppers who use it tend to spend more per order. Those figures come from Amazon itself and should be read with care: the methodology behind the comparison has not been published, and no incremental revenue figure has been disclosed. Still, the direction of travel is clear. The question for any retailer selling online in Spain is not whether conversational commerce arrives, but whether their store is ready to be part of it.

Why this matters for independent stores in Catalonia

It is tempting to file this under "big marketplace news" and move on. That would be a mistake. When a marketplace of Amazon's scale trains shoppers to ask questions in natural language — "find me a gift under €40 that arrives before Friday" — expectations travel. Customers who buy from a Barcelona boutique or a Lleida specialist on Monday will expect the same fluidity when they visit that shop's own site on Tuesday.

At the same time, marketplaces are tightening their grip on the conversation itself. Amazon has restricted some external agents from its marketplace while building its own assistant as the primary interface. For brands, that is a reminder that owning the direct relationship with the buyer — through your own ecommerce site — remains strategically valuable. An online store built on PrestaShop or WooCommerce is not just a sales channel; it is the place where you keep the customer data, the product knowledge and the margin.

What a conversational-ready store actually needs

The technical foundations are less exotic than the headlines suggest. Most of the work is about making your catalogue and your checkout legible to both humans and machines:

Structured product data. Clear attributes, consistent naming, accurate stock levels and rich descriptions. An assistant can only recommend what it can understand, and so can a hurried shopper on a phone.
Fast, stable infrastructure. Conversational traffic is spiky. A store that stalls under load loses the sale before the conversation begins.
Frictionless checkout. Fewer steps, transparent shipping costs in euros, and payment methods your customers actually use — cards, Bizum, bank transfer where relevant.
GDPR-compliant data handling. Any personalisation or assistant integration must respect EU data protection rules, with clear consent and a lawful basis for processing.
Search that forgives. Synonyms, typos, regional terms in Catalan or Spanish. If a customer types "samarreta" and your catalogue only knows "camiseta", you have a gap.

Conversion optimisation is the real battleground

Assistants raise the ceiling on discovery, but the sale still happens — or fails — on your product pages and at checkout. This is where a tailored online store earns its keep. A generic template will get you live; a store tuned to your catalogue, your margins and your customers will get you profitable.

In practice, that means measuring where people drop off, testing product page layouts, simplifying forms, surfacing delivery times and returns policies early, and making sure the mobile experience is genuinely pleasant rather than merely functional. Small, disciplined improvements compound. A store that converts a little better funds the next round of improvements.

A pragmatic roadmap for the next twelve months

You do not need to build your own AI assistant to benefit from this shift. A sensible sequence looks like this:

Audit your data. Are product attributes complete? Are prices and stock accurate across channels? Fix the basics first.
Choose a platform that fits. PrestaShop and WooCommerce both offer mature ecosystems, strong European hosting options and the flexibility to integrate new tools as they mature.
Instrument everything. Analytics, consent-aware tracking and a clear view of the funnel from landing page to order confirmation.
Test and iterate. Run structured experiments on search, product pages and checkout rather than redesigning on instinct.
Keep security and maintenance on a schedule. Updates, backups and monitoring are unglamorous but non-negotiable for a store that takes payments.

The takeaway

Amazon's numbers are a signal, not a blueprint. The lesson for retailers in Lleida, Tarragona, Girona and Barcelona is that the interface of commerce is changing, and the stores that thrive will be those whose foundations — data, speed, checkout, compliance — are solid enough to plug into whatever comes next. Conversational assistants will amplify good stores and expose weak ones. Building a store that genuinely sells, rather than one that merely exists, is the most durable response.

Put these ideas into practice

Talk to ALMC about a solution for your business. Explore your options or contact our team.

Results without leaving the page.