Nobody buys a TV or a washing machine on impulse. Naldo's shoppers come in, compare, leave and come back. Synerise gave Naldo an end-to-end solution for that whole journey. One behavioral AI recognizes each shopper, from the first anonymous visit to every return, and decides what they see at every step. Today, 55% of Naldo's revenue follows a Synerise interaction in the previous 24 hours.
Subtitles available in English.
The client
Naldo is an Argentine retail chain selling home appliances, technology and household goods, in its stores and online at naldo.com.ar.
The goal
Naldo wanted to stop selling products and start answering needs. "It's no longer just about offering products to sell them, but about understanding what the customer needs and the right moment to offer it," says Natalia Garetto, E-commerce Manager at Naldo.
The category makes that hard. "Home appliances are not a rushed or impulse purchase," explains Santiago Bozzetti, E-commerce Marketing Lead. "Users usually come in, look, compare, come back, leave and return again." Generic messages sent to everyone in the same way could not keep up with decisions that long. Every visit had to build on the one before.
The solution: one engine, end to end
Synerise's behavioral AI infrastructure gives every user, identified or not, an ID and a history. The AI records and centralizes every interaction, recognizes the customer on each return and decides what to show them. Synerise delivered it end to end. The same engine runs every step of the journey: from the first visit, through the home page, listings, product pages and checkout, to the messages that bring shoppers back.
Step 1. First visit: reachable before identified
A first-time, unidentified visitor sees on-site pop-ups: a newsletter sign-up, a web push consent request or a promotional offer. Web push builds an audience Naldo can reach before it has any hard data, such as an email or phone number.
Step 2. Home page: different for every visitor
"What one person sees is not what another person sees," says Bozzetti. Recommendations follow each shopper's history across the site, and 14% of Naldo's sales now come from transactions with recommended products.
Step 3. Listings: installments as a ranking criterion
Shoppers can now sort and filter listings by number of installments: 12, 9, 6 or fewer. Before, installments were not even a sorting option. For higher-value products, installments, financing or a discount often decide the purchase.
Step 4. Product page: similar products and cross-sell
Each product page carries two personalized carousels. The first shows similar products: a shopper looking at a TV sees other TVs. The second cross-sells: the same shopper sees a wall mount or a soundbar.
Step 5. Checkout: the last chance to complete the basket
The checkout carousel draws few clicks, because not every shopper reaches the last step. Those who do add products they missed earlier in the journey. Products per ticket in the cart grew by 108%.
Step 6. After the visit: bringing shoppers back
When a shopper leaves a cart behind, the journey continues. Naldo's abandoned-cart automation sends an email first and a WhatsApp message second. Recovered purchases carry a ticket 30% higher than Naldo's average. "An abandoned cart stopped being a lost opportunity and became an opportunity for recovery," says Garetto. When shoppers return, the AI recognizes them with their full history and decides what to show them next.
What the AI changed for the team
The team stopped curating by hand. "We stopped thinking about which products we wanted to show each week," says Garetto. The AI now selects the products with the highest interest for each customer, automatically.
The team also changed how it measures success. "If we had made decisions based on click-through rate alone, we would probably have switched off components that seemed to underperform at first glance," says Garetto. "But the click is only one part of the journey. The real challenge is understanding what happened after it." With one engine across every step, Naldo sees what each component contributes to the whole journey, not only to its own clicks.
What's next
Naldo plans personalized listings: product lists ordered by each user's interactions and history, beyond alphabetical, price or installment sorting.
Results
- 55% of revenue had a Synerise interaction in the previous 24 hours
- 14% of sales come from transactions with recommended products
- +108% products per ticket from the checkout carousel
- Recovered carts carry a 30% higher ticket than average
Summary
Naldo's shoppers take their time, and Synerise uses every step of it. One behavioral AI recognizes each shopper from the first visit to every return and decides what appears at every point in between. Garetto's advice to other e-commerce managers: "It's time to get to know our customers better. AI should no longer be treated as a project separate from e-commerce, but as the engine for understanding our customers. And that clearly has an impact on business conversion."
Thank you to the Naldo team, especially Natalia Garetto and Santiago Bozzetti, for sharing their story.
