A customer with a smartphone and a sales consultant with a tablet in a premium perfume store in the evening

Oct 19, 2026

Customers Now Decide Before They Reach the Store

Agentic Commerce: a new power is emerging at commerce's invisible door.

AI

Future of Commerce

Strategy

Context

The machine must be able to understand the product. The human must continue to desire the brand.

Once, a brand's biggest struggle was getting the customer into the store.

Then we started pulling the customer to the website.

After that, we competed to appear on the first page of search engines.

Now we are at the beginning of a much bigger change.

Customers no longer have to visit brands one by one. They can hand a significant part of their decision-making process over to artificial intelligence.

This change is not just a new technology trend. It means redefining the relationship between the customer and the brand.

And in my view, one of the biggest commercial competitions of the coming years will be reaching the customer's decision before reaching the customer.

Analysis

1. The new interface of shopping: the customer's digital representative

Imagine a customer who wants to buy a new perfume.

In the past, they would go to a shopping mall, walk through several stores and try the products.

Then they started researching on Google.

Later, they began ordering online the products they saw on social media.

Now they can say to an AI assistant:

“Recommend a men's perfume under 3,000 TL that I can wear at the office — with character but not heavy, and long-lasting. Compare the user reviews.”

An important change is hidden inside this question.

The customer does not name a brand.

They describe a need.

AI turns that need into products, brands and purchase options.

This new approach to commerce is called Agentic Commerce — commerce carried out through AI agents.

Here, AI is no longer just an interface that answers questions. It can research products, evaluate options, prepare a shopping cart in compatible systems and move transactions forward with the user's authorisation.

Google's expansion of its Universal Commerce Protocol (UCP) work in 2026 and Shopify's development of commerce infrastructure for AI agents are among the concrete signs of this transformation.

But the real issue is not the name of the technology being used.

The real issue is who will hold influence over the commercial decision.

2. A brand's new rival may not be other brands

In the traditional marketing approach, competition was quite clear.

Brands competed for shelf space.

They competed for visibility in advertising media.

In the digital world, they fought over keywords, advertising budgets and search engine rankings.

In the new era, another arena of competition is being added:

Being a meaningful option in the evaluation and recommendation systems of algorithms.

For example, a brand may have a communication budget of millions of lira.

Its campaigns may be very successful.

Its stores may be extremely impressive.

But if product features are not clear enough, if stock and price information is inconsistent, if it is unclear which need the product serves, or if it is not adequately described in trusted digital sources, it may be at a disadvantage in AI-assisted discovery.

A new marketing problem emerges here.

Can the value the brand creates for the consumer be interpreted correctly by digital systems?

We used to ask marketing departments:

“How high is our brand awareness?”

Now we need a second question:

“How accurately is our brand understood by systems searching for answers to customer needs?”

The answers to these two questions will not always be the same.

Luxury perfume product page on a smartphone

Server room corridor with blue and green lighting

In Depth

3. A product's digital identity is becoming as important as its packaging

For many years in product management we talked about the importance of packaging, design and visibility on the shelf.

These still matter.

But now a product has a second identity: its digital information identity.

Think of a cosmetics product.

If the product's brand, category, purpose of use, volume, ingredient information, target audience, price, stock status, images and delivery terms appear in different forms across different systems, a serious inconsistency arises.

A human can sometimes tolerate this inconsistency.

Automated evaluation systems, however, need reliable and comparable information.

That is why product management is no longer just the job of purchasing, marketing or the category department.

Product information is becoming the shared responsibility of technology, e-commerce, operations and brand management.

This is even more critical in multi-category cosmetics, fashion, home and retail companies.

Because correctly defining hundreds or thousands of SKUs is not merely a technical catalogue-editing exercise.

This work determines how strongly products are represented in new commerce channels.

4. AI may increase the risk of brands becoming alike

I see a more important danger here.

AI systems compare products using measurable features.

Price.

Volume.

Material.

Purpose of use.

Delivery time.

User ratings.

But not everything that differentiates a strong brand fits into these fields.

A perfume is not made of essences alone.

A premium fashion brand does not sell only fabric quality.

A luxury store is not just a place where the product is handed over.

Brands produce meaning, culture, experience and desire.

McKinsey's May 2026 research on luxury retail also stresses that, in AI-mediated shopping, brands must protect their own meaning and not lose control of the customer relationship.

That is why brands face two separate tasks.

First, making their products understandable to algorithms.

Second, not reducing their brands to features that algorithms can easily compare.

In my view, the successful brands of the future will be those that can balance these two worlds.

The machine must be able to understand the product. The human must continue to desire the brand.

5. The organisational structure of marketing has to change

Preparing a brand for the world of Agentic Commerce is not a task to be handed only to the digital marketing team.

Because in an AI-driven shopping experience, the product being found is not enough on its own.

The recommended product needs to be in stock.

Price information needs to be correct.

The delivery promise needs to be kept.

Customer data needs to be managed securely.

The after-sales experience needs to be consistent with the quality the brand promises.

All of this requires the shared success of the commercial organisation.

At this point, I believe classic departmental boundaries need to be reconsidered.

For example, if the marketing department increases demand for a product while the supply chain cannot ensure its availability, the system has failed.

If the e-commerce team manages product data correctly while customer service gives wrong information, the system has failed.

If AI recommends the right product but the customer's payment or delivery experience is problematic, the system has failed again.

The new competition will not be between independent departments, but between commercial systems that work together.

Roadmap

6. What should brands do now?

I am not suggesting that companies change all their technology investments.

First, they need to understand how ready their current commercial structure is.

In the first stage, product data quality should be reviewed.

In the second stage, it should be determined which customer needs the products are associated with.

In the third stage, how the brand is represented in search engines, shopping platforms and AI tools should be investigated.

Then the consistency of price, stock and product data should be checked.

In the final stage, it should be decided which parts of the customer experience can be improved through automation and in which parts the human touch must be preserved.

Especially for premium brands, automating the entire process may not be the right strategy.

Because in some purchase decisions, what the customer expects is not speed, but attention.

AI can make research easier.

It can speed up comparison.

It can reduce repetitive tasks.

But how the emotional relationship the brand creates will be protected must be designed separately.

Conclusion

7. Who will win in the future of commerce?

In Deloitte's 2026 Global Retail Industry Outlook, around 68% of the executives surveyed expect to use Agentic AI in specific business processes within the next 12–24 months.

This shows that the technology has entered executives' strategic agenda.

But the spread of a technology does not create competitive advantage on its own.

Consider why two companies with the same tools produce different results.

In one, data is consistent, the organisation works in coordination, inventory management is strong and the customer experience is clear.

In the other, information is scattered, departments are disconnected and commercial decisions are delayed.

Both companies can use AI.

But they cannot be expected to achieve the same commercial result.

That is why the decisive factor in the retail of the future will not be merely owning AI technology.

It will be building the commercial architecture in which that technology can work.

Final word

For years, brands fought to capture the customer's attention.

Now they will fight to earn a place in the customer's decision-making mechanism.

Companies that see this change merely as a new digital marketing channel may miss an important part of the transformation.

Because what is changing is not the screen on which the ad is shown.

It is the place where the purchase decision is formed.

And in the history of commerce, when the place where the decision is made changes, the rules of competition change too.

Sources