For years, brazilian retailers approached the final quarter with three dates in mind: Children’s Day, Black Friday and Christmas. They still matter. But while stores plan those campaigns, shoppers are getting invitations to buy on three other dates: 10.10, 11.11 and 12.12.

It is a change in the calendar that could also change when people decide to shop.

The 11.11 sale grew out of Singles’ Day in China. A celebration of being single became a shopping festival. The four ones made the date easy to recognize and remember. With investment and repetition, it became an occasion to buy. www.alibabagroup.com

In 2025, nearly 600 brands generated more than RMB 100 million in sales during the festival on Alibaba’s platforms. In Brazil, Shopee said it sold 20 million items in a single day on 11.11 that same year. These figures come from different companies and measure different things. They tell us nothing about sales in physical stores. But they do show how much attention the platforms can bring to the date. home.alibabagroup.com

And that is precisely the point.

There is no reliable nationwide figure proving that these double day promotions are already generating additional revenue for physical stores in Brazil. For many stores, they may not be. The movement gained momentum on platforms that invest in advertising, coupons and entertainment to bring shoppers back month after month.

For this year’s 9.9 campaign, Shopee announced R$20 million in coupons. That is an investment in making people notice the date. And this is not simply a contest with products shipped from abroad. In June 2026, the platform said local stores accounted for more than 85% of its sales in Brazil. shopee.com.br

A physical retailer does not need to copy a marketplace. But it should recognize that its customers are being invited to shop on new dates. If someone walks into a store on November 11 and finds exactly the same experience they would on any other day, all the attention built around 11.11 stays outside.

There are three particularly interesting opportunities still ahead this year.

October 10 falls on a Saturday, two days before Brazil’s Children’s Day. For a toy store, a children’s clothing shop or a gift retailer, it could be a chance to help undecided customers choose something they can take home immediately.

November 11 comes sixteen days before Black Friday. Instead of moving the entire sale forward, a store could select one category, test an offer and meet customers who might not yet know the business. It could also learn what they want before the main November campaign.

December 12 is another Saturday, thirteen days before Christmas. By then, many people simply want to get their gifts sorted. Seeing the product, getting a recommendation, putting together a gift set, having it wrapped and taking it home can be worth as much as a few extra percentage points off the price.

None of this guarantees additional sales. A promotion might merely bring forward a purchase that would have happened on Black Friday or before Christmas. It might lift revenue for one day while reducing the month’s margin. That is why any test needs to be measured carefully: how many people came in, how many bought, what was the margin and what happened in the following weeks?

I would start with just one date. Choose a category, a small selection of products and a clear reason to visit the store. Make the offer visible in the window, on WhatsApp and across social media. Then compare the results with a similar period.

The aim is not to add three more clearance sales to the year. It is to find out whether these dates can generate new visits, new customers and purchases that would not otherwise have happened.

Marketplaces have already made people notice double day promotions. Physical retailers now have to find out whether they can turn that attention into something of their own.

Shoppers have added new dates to their promotional calendar. The question is whether your store will have something relevant to offer when those days arrive.

NRF Europe did not bring new topics. It brought new consequences

What really changed in the conversations about artificial intelligence, customer experience, stores, and expansion

Customer experience, omnichannel retail, artificial intelligence, data, loyalty, and emotional connection are not new topics for Brazilian retail.

Nor is it new to say that physical stores will remain relevant, that customers move across channels, or that technology should make their lives easier. These discussions have appeared for years in events, studies, and articles, including many of the pieces I have published.

So, what was truly different at NRF Europe 2026?

It was not exactly the topics. It was their consequences.

In Paris, some conversations moved beyond intentions and began to change the structure of retail itself. How customers search, how brands are discovered, how stores are evaluated, how employees are incentivized, and even who controls the commercial relationship.

More than new trends, I found changes in scale and consequence.

1. Consumers are not simply searching differently. They are beginning to outsource the choice

Personalization has been discussed for years. So have recommendations. What is emerging now is different.

Consumers are beginning to hand their intentions over to an agent and allow it to navigate the journey on their behalf.

“I need a black coffee maker for less than €200, delivered by Friday.”

“I want a gift for my wife within this price range.”

In this model, consumers do not need to understand categories, filters, brands, or technical features. They explain the problem and receive a ready made selection.

Mango, L’Occitane, Kingfisher, MediaMarktSaturn, and Grupo Boticário presented experiences moving in this direction.

It is what I have been calling the Waze effect applied to consumption. Just as many people no longer know the routes because they have handed that responsibility to an app, we may stop knowing brands, products, and alternatives because we will hand the choice over to an agent.

What is new is not that artificial intelligence can recommend something. It is that consumers will choose less and less.

2. The competition is moving from the first page to the final shortlist

For years, brands and retailers competed for traffic. The goal was to appear in the results, earn the click, and guide consumers through their own environments.

Agents change this logic.

Customers may no longer receive twenty alternatives. They may receive three. And they may choose one without ever visiting the websites of the others.

According to Kingfisher, traffic coming directly from artificial intelligence models still represents less than 1% of its online business. But the conversion rate of that traffic can be two to four times higher.

The volume is still small. The quality of the intent is already significantly greater.

This transfers power to whoever controls the recommendation. The agent may become a new kind of intermediary, capable of deciding which brands enter the consideration set and which simply cease to exist within that purchase.

The battle will not be only for attention. It will be for the right to become part of the choice.

3. Being found by people is no longer enough. The company must be understood by machines

The conversation about SEO is gaining a new layer.

Products, prices, inventory, reviews, delivery times, payment methods, and fulfillment conditions must be organized so that an agent can interpret, compare, and recommend them.

KPMG and FEVAD described this as being Agent Ready.

This is not only about producing optimized content that appears in answers. Companies need consistent data, connected systems, updated information, and the ability to allow an agent not only to consult the catalog, but eventually to build a cart and complete a purchase.

Kingfisher was very direct in saying that retailers will have to communicate with two audiences: people and machines.

This may be one of the newest points for the Brazilian market.

For years, we organized retail so that it could be understood by consumers. Now, we will have to organize the business so that it can be correctly interpreted by their digital representatives.

4. The retailer may receive the purchase without knowing the journey

There is another consequence that has received little attention.

Today, when customers navigate a website, retailers can observe searches, products viewed, comparisons, abandoned carts, and other signals of intent.

In a journey conducted by agents, consumers may arrive with the cart almost complete.

The sale happens, but part of the knowledge about how that decision was made remains with the platform that intermediated the journey.

The retailer may gain conversion while losing understanding.

This change affects data, media, personalization, and customer relationships. It may also create a new dependency, similar to the one retailers already have on search engines, marketplaces, and social platforms.

The concern, therefore, should not be limited to selling through agents. Retailers will also need to discuss which data comes back, who controls the customer’s identity, and how to preserve a direct relationship after the transaction.

5. The age of proofs of concept is beginning to give way to industrialization

Artificial intelligence continues to produce many interesting projects. But the most mature conversation at NRF Europe was no longer about what can be tested.

It was about what can be scaled.

MediaMarktSaturn summarized this with an uncomfortable but necessary statement:

A pilot without a path to scale is just theater.

EuroCommerce and McKinsey argued that AI should be connected to a clear value map. L’Occitane began by examining its P&L to identify expensive, slow, and inefficient activities, such as translation and product feed management.

Żabka presented a transformation built over several years, beginning with data, assortment, pricing, and location. The company did not wait for generative AI to organize its foundations.

What is new is not the use of artificial intelligence. It is turning an isolated experiment into a new way of working.

This requires budget, adoption, process change, and clear accountability. Technology can be implemented within a few months. Companies rarely change at the same speed.

6. The next advance will not be a better agent. It will be the orchestration of multiple agents

Another important difference appeared in MediaMarktSaturn’s presentation and in the EuroCommerce and McKinsey study.

The first phase was marked by isolated solutions. One agent for customer service, another for content, another for analysis, and another for inventory.

The next phase will be about making these intelligences work together.

One agent identifies a problem. Another consults the data. A third recommends an action. Another executes it within the boundaries established by the company.

The advantage will not come only from the quality of each tool, but from the ability to coordinate the entire system.

This reinforces something I have believed for a long time: AI should not be treated as a product. It is a process. It is M(AI)ndset.

It is not about adding a new tool to the old way of working. It is about rebuilding the work around what can now be analyzed, decided, and executed differently.

7. Technology is leaving the stage so that people can return to it

For a long time, the more screens a store had, the more modern it appeared.

At NRF Europe, the most interesting vision was almost the opposite.

Arket, Aroma Zone, IKEA, Lagardère, Żabka, and MC Sonae discussed technologies that operate behind the scenes or in the hands of employees. Systems that organize tasks, identify out of stocks, interpret information, and reduce administrative work.

The goal is not to make customers admire the technology. It is to give them a simpler store and an employee with more time to serve them.

Even Hive’s humanoid robot case followed this logic. The robot is being tested to replenish shelves overnight, initially through remote operation and progressive learning. The proposed business model is to charge by the hour, at a cost close to that of an employee.

The most interesting aspect is not its humanoid form. It is the evolution model.

First, the robot is operated by someone. Then it learns parts of the task. Gradually, it gains autonomy without requiring the store to be rebuilt around it.

Automation stops being a distant promise and becomes a service that can be contracted.

8. Stores are beginning to be measured as media, acquisition, and trust

Saying that physical stores matter is nothing new.

What has changed is the attempt to measure their influence beyond the cash register.

Aroma Zone does not look only at sales completed inside the store. It analyzes how the location increases brand awareness, attracts new customers, and influences the digital relationship.

Westwing monitors the growth of online sales in the areas surrounding its stores. Granado uses its international locations to educate consumers about a brand that does not yet have the same recognition abroad that it enjoys in Brazil.

This changes the store’s P&L.

A location may not appear extraordinary when measured only by the sales made at that address. But it may be extremely valuable when its effect on acquisition, repeat purchases, and digital sales is considered.

The challenge becomes measuring this influence more effectively, rather than requiring each channel to prove its profitability in isolation.

9. Loyalty begins before the purchase and goes beyond discounts

The Snipes case brought a different perspective to loyalty programs.

Its app combines benefits, early access, payments, and a wallet where customers can save money before making a purchase.

Instead of simply buying now and paying later, customers can set aside small amounts until they can afford the sneakers they want.

The brand becomes part of the purchase planning before the transaction takes place.

Another interesting point was employee involvement. Each store employee received an individual code to invite customers into the program. With tracking and incentives, registrations almost doubled overnight.

Employees were no longer simply responsible for explaining the program. They became measurable microinfluencers for the brand.

Loyalty begins to combine payments, access, recognition, cultural experiences, and personal relationships.

It is no longer just a points program. It is an infrastructure for connection.

10. In international expansion, the greatest risk may be adapting too much

In the panel I moderated with Granado, Joe & The Juice, and L’Osteria, one of the strongest conclusions was that entering a country should not be confused with opening a store.

The company needs to know whether it can sustain the operation, supply it, communicate effectively, and achieve scale in that market.

A single location will rarely justify all the complexity of entering a new country.

But another point emerged strongly: the risk of excessive adaptation.

Dengo, Bauducco, Granado, and Joe & The Juice need to adjust products, formats, and communication. However, if they change the proposition too much, they may eliminate precisely what made the new market interested in the first place.

Dengo does not need to compete merely as another good chocolate brand in France. It needs to present what makes it different, its Brazilian origin, its ingredients, and its relationship with cocoa.

Products can be adapted. Operations can be adapted. The identity must remain recognizable.

What was truly different in Paris

NRF Europe did not present a list of completely new topics.

It discussed customer experience, data, physical stores, loyalty, artificial intelligence, and expansion. All of these are already part of the Brazilian retail agenda.

What changed was the depth of the consequences.

Artificial intelligence does not merely improve search. It can remove part of the choice from consumers.

Agents do not merely recommend products. They may become the new owners of intent and of the customer journey.

Stores do not merely complement digital channels. They are beginning to be measured as media, acquisition, and trust.

Employees do not merely execute tasks. They can become consultants, influencers, and operators of increasingly invisible technologies.

Loyalty does not begin after the purchase. It can begin when customers decide to save money for something they want to buy in the future.

Expansion is not a sequence of store openings. It is the ability to sustain scale without losing identity.

Perhaps this was the main message from Paris.

The conversations still sound familiar. But their consequences have become much greater.

When technology begins to choose the paths, being available is no longer enough.

Brands must be understood by machines, recommended by agents, and still desired by people.

Caio Camargo

In Edmond Rostand’s play Cyrano de Bergerac, Christian was young, handsome, and in love with Roxane. But he did not have the words to win her heart.

Cyrano did.

So Cyrano began writing the letters that Christian signed. Roxane fell in love with the intelligence and sensitivity behind those words, believing they belonged to the man delivering them.

More than a century later, artificial intelligence is recreating this story.

Cyrano is AI, hidden behind the scenes and finding the perfect words.

Christian is the user, who publishes, signs the work, and receives the applause.

Roxane is the audience, captivated without knowing exactly whose intelligence it is admiring.

This is what I call the Cyrano Effect.

It happens when someone uses artificial intelligence to win over an audience with thoughts, arguments, and words they would not be able to produce or defend on their own.

I see nothing wrong with using AI to write. Editors and proofreaders have always helped people organize their ideas. Technology has simply made this support faster and more accessible.

However, there is a significant difference between improving an idea and manufacturing intelligence.

When the experience, knowledge, and opinion belong to the author, AI acts as an editor. It organizes, questions, and finds a better way to communicate the person’s thinking.

In this case, Cyrano helps Christian express what he truly feels.

But when the tool also has to create the thought, the argument, and the opinion, Christian begins playing a character written by Cyrano.

The writing starts promising someone the conversation cannot deliver.

This debate gained an interesting new chapter when Anthropic announced that Claude would begin embedding an invisible watermark in the text it produces. By creating statistical patterns through word choices, the system can estimate whether Claude was involved in producing a piece of content. In programming code, the watermark is more limited and may appear only in comments and other sections that do not affect how the software works. The initiative is connected to the transparency requirements of the European AI Act. Anthropic

The announcement has angered some users. Some fear being identified even when they only use AI to review or translate their own writing. Others question whether the watermark could affect the quality of the final text.

It is important to clarify that the system cannot definitively prove that content was created by AI. It indicates a probability, works better with longer texts, and does not identify the user.

Even so, there is something almost poetic about it.

Cyrano himself will begin leaving clues in the letters he writes for Christian.

But perhaps some people are worried about the wrong detector.

The best way to find out whether someone truly understands what they published has never been to search for a hidden mark in the text. You only need to ask a question that was not in the script.

An article can be created in seconds. Knowledge, experience, and the ability to defend an argument still take years to build.

Before publishing something created with the help of artificial intelligence, perhaps the question should not simply be, “Is the text good?”

The more important question is:

Could I personally defend everything I am signing?

If the answer is yes, Cyrano merely helped you find better words.

If the answer is no, be careful.

Sooner or later, Roxane will ask a question.