Retail has built an extraordinary machine for reaching shoppers. Advertising now follows people through marketplaces, retailer websites, connected television, social platforms, apps, and stores. Commerce media spending is expected to rise by another 12.1% in the United States this year, faster than the wider advertising market.
AI and retail media are making messages cheaper to produce and easier to place, while customer attention remains limited.
Shopping platforms are also taking a larger role in deciding which products receive that consideration. Walmart’s Sparky compares products, summarises reviews, and suggests options for particular occasions. Amazon says Rufus was used by more than 300 million customers in 2025. The company renamed the service Alexa for Shopping in May 2026.
A product can now disappear from consideration before the shopper reaches its page. Its information may be incomplete, its reviews may be weak, or a rival may simply be easier for the platform to understand and recommend.
Precision means investing where marketing is likely to change what the customer does, and holding back where it adds little or risks weakening the relationship.
Retailers now need better judgment over which demand to create, which customer moments to use, and which immediate returns will cost the brand more later.

Personalisation is becoming easier to produce and harder to distinguish
As customisation becomes cheaper, every new customer signal can become another reason to communicate. Yet 81% of shoppers in one 2025 study said they ignored irrelevant marketing, while Qualtrics found that only 33% trusted companies to use personal information responsibly.
The next advance in personalisation will come from suppression. Retailers need to know when a customer is likely to buy, when a service issue should end promotional activity, and when a prior return should remove a product from consideration.
Most campaign systems calculate the value of sending. Few calculate the value of not sending. That omission matters because an unnecessary message can consume attention that would have been more valuable later.
Precision should therefore be judged by the incremental value of the contact and the value preserved by withholding it. A brand that knows more about the customer should need fewer wasted interactions, not more opportunities to display what it knows.
Utility is earning influence before the shortlist is formed
Some of retail’s most important technology investments now help customers work out what they need before they begin comparing products.
The Home Depot’s Magic Apron answers project questions and summarises reviews. An expanded version being tested in selected stores connects its recommendations with local inventory and exact product locations. A customer can begin with a leaking faucet rather than a technical product term.
Instacart’s AI assistant can turn a meal idea or an uploaded list into a cart based on current stock at a nearby store. Amazon Lens allows shoppers to search with an image when they cannot describe what they have seen.
The value of these tools lies in helping people find the right product faster, avoid unnecessary returns, and complete purchases they might otherwise abandon.
The larger commercial opportunity comes from influencing how the need is defined. A retailer that helps diagnose a home repair can shape the full set of products required. A grocery platform that builds the meal can influence the entire basket. The business is no longer competing for a single item once the shortlist has been formed.
This can alter category economics. If they work as intended, these services could increase basket value, reduce poor product choices, and give customers a reason to return for similar purchases. They can also reveal recurring points of difficulty that justify changes to the assortment, content, or service.
The strongest measure is therefore not tool usage. It is whether the service improves the value and quality of the transaction: larger completed missions, fewer failed purchases, lower return costs, and greater repeat use of the retailer for similar needs.
A higher price can conceal a shrinking market
Ralph Lauren increased average unit retail across its direct-to-consumer business by 18% in the third quarter of fiscal 2026. Full-year revenue then rose 15% to $8.1 billion, supported by stronger full-price selling and fewer promotions.
Coach revenue grew 29% in constant currency in Tapestry’s fiscal third quarter of 2026, while the group added more than 2.4 million customers globally. More than 35% of those new customers were from Generation Z. Both brands sold more at higher prices while continuing to attract new customers.
The wider luxury market shows why price alone is a poor measure of strength. Bain estimates that personal luxury goods sales fell 2% in 2025, while the sector’s customer base declined from roughly 400 million in 2022 to about 340 million. Bain said customer disengagement reflected concerns about the value offered as prices rose, including whether product creativity and the wider luxury experience had kept pace.
Average selling price can rise while the market contracts. Leaders need to examine who is still entering the brand, how often customers are buying, whether units are falling, and how much growth depends on markdowns or the wealthiest buyers.
Retail media is facing a higher standard of proof
Retail media can connect ad exposure with a transaction, but sponsored search often claims purchases from customers who already intended to buy the brand.
In November 2025, IAB and IAB Europe issued joint guidelines for incremental commerce-media measurement. Walmart Connect has also introduced incremental sales measurement for Sponsored Search, using test and control groups to estimate the sales created by advertising.
This can produce very different investment decisions. A campaign with a high attributed return may add little revenue when it reaches established buyers. Another may appear less efficient while recruiting customers who are new to the brand or category.
Retail media partners should explain how long they claim credit for a sale, how they separate new sales from purchases that would have occurred anyway, and how they avoid crediting multiple ads for the same purchase. Incremental revenue should sit beside attributed revenue. Otherwise, brands risk paying platforms to document demand they already own.
Some demand should never be created
Campaigns are usually rewarded for generating a response, even when the business cannot profitably fulfil that demand. A campaign can raise sales while increasing the cost of fulfilling and servicing them.
Target has said it is connecting information across marketing, digital, merchandising, and supply chain systems while using AI to identify low-stock items earlier. Walmart says it is using AI to forecast demand and redirect inventory across several international markets.
Marketing should promote only what the business can supply and serve profitably. Stock levels, likely returns, fulfilment costs, and pressure on customer service should be checked before a campaign is approved. Campaign results should include the cost of fulfilling the demand they create.
Algorithmic pricing is now a brand decision
The Federal Trade Commission reported in January 2025 that pricing intermediaries could use precise location data, browsing history, and purchase behaviour to support individualised pricing and promotions. The capability exists even where companies choose not to use it.
Instacart’s price-testing controversy showed how quickly the issue can escape the pricing team. Research found that shoppers could see different prices for identical groceries at the same store and at the same time. Instacart said the tests were randomised by product and store location, rather than based on customers’ personal characteristics, and ended them after public criticism. New York’s attorney general later sought information about the company’s pricing practices and disclosures.
Leaders should know what customer information affects the price, how far prices may vary, and which products are too sensitive for this kind of testing. Once customers can compare and share price differences, those rules become part of how the brand is judged.
Customer access needs an accountable owner
Optimove found that 47% of consumers were already tired of holiday marketing by October 2025. By the end of November, the proportion had reached 63%.
Technology can impose frequency caps, but it cannot decide which message deserves the available space. That requires an executive with authority over total customer contact, visibility into every planned and triggered interaction, and the power to stop a promotion when a service problem, delivery update, or more important message needs priority.
Retailers should also keep a long-term control group that receives fewer messages. Comparing its spending, retention, and opt-out rates with those of more heavily contacted customers can show whether extra communication is creating value.

Efficiency can rise while the market narrows
Growth systems naturally favour recent purchasers, loyalty members, branded searchers, and customers already close to conversion.
Kantar found that brands people are already inclined to choose achieve nine times the volume share, command twice the price, and are four times more likely to grow value share. Brand investment creates the familiarity that allows precision to work later, often before the customer has taken any action the retailer can identify.
Nike shows how difficult it can be to rebalance growth after a sharp decline in its direct and digital business. Fiscal 2025 revenue fell 10%, and Nike Brand Digital declined 20%. In fiscal 2026, wholesale revenue rose 6%, while Nike Direct and digital remained under pressure.
Return on ad spend should be considered alongside new-buyer growth, branded search, customer age mix, and dependence on a small number of products.
Marketing is moving closer to the product decision
Target’s appointment of Cara Sylvester as chief merchandising officer in February 2026 is instructive. Her previous leadership roles spanned customer experience, marketing, digital operations, loyalty, and merchandising. The appointment suggests that Target is bringing customer insight closer to decisions about what it sells.
The value of that insight depends on when it enters the process. Customer behaviour can reveal which products attract new buyers, where interest fails to convert, and which parts of the range rely too heavily on promotion. Used early enough, that evidence can change the assortment. Used later, it can only change the campaign.
Competitors can often reach similar audiences with similar levels of precision. A distinctive product range is much harder to copy.
Marketing leadership, therefore, needs a formal role in assortment and innovation reviews. The CMO should be able to identify where the brand is losing distinctiveness, where a prominent product is failing to draw customers into the wider range, and where campaign performance is concealing weak product appeal.
Marketing’s job begins before the media plan, when the company decides whether the offer deserves demand in the first place.
Attention will become a test of judgment
Similar technology will be available across the market. The harder advantage will be deciding what the brand refuses to optimise.
Automation can generate offers, adjust media, recommend products, and vary prices faster than most organisations can consider the consequences. It will amplify the commercial logic already inside the company, including its weakest assumptions.
Senior leaders will need to reject activity that delivers a quick return while making customers less willing to choose the brand without another incentive. That may mean refusing an unfair price difference, a promotion that trains shoppers to wait for discounts, media aimed at people already certain to buy, or a product whose campaign results hide weak demand.
The strongest brands will preserve the ability to grow without chasing every purchase with another discount, ad, or reminder.
Need clearer insight into what earns attention, builds loyalty, and protects pricing power? Kadence can help you uncover the customer behaviours and decision points shaping retail growth.