Artificial intelligence is changing how brands understand and engage with customers. Instead of relying only on broad audience segments, marketers can increasingly use individual behaviour, real-time data and AI-driven recommendations to decide what each customer sees. However, 2026 research reveals a widening gap between what businesses consider personalised and what consumers actually find useful, raising questions about whether AI-powered individualisation can deliver better experiences without becoming intrusive.
Imagine two customers opening the same shopping application.
Both are 30 years old, live in Mumbai and have previously purchased fitness products. Under a conventional marketing strategy, they might receive similar recommendations because they belong to the same customer segment.
But one customer is searching for affordable running shoes, while the other is looking for a premium pair after returning an unsuitable size last week.
Should they receive the same advertisement?
For years, marketers have used demographics, browsing history and purchasing behaviour to divide consumers into groups. A customer might be classified as a frequent traveller, a beauty enthusiast or a high-value shopper, with campaigns designed around those categories.
Artificial intelligence is allowing brands to move beyond those classifications.
Instead of asking which group a customer belongs to, AI-supported systems can assess what that particular person might need during an interaction. Recommendations can change according to recent purchases, browsing activity, location, preferences or previous responses.
The industry increasingly describes this shift as moving from personalisation to individualisation.
The distinction lies in how decisions are made. Personalisation often adapts experiences for groups of similar customers, while individualisation aims to make decisions using the circumstances of one customer at a particular moment.
However, individual-level recommendations are not entirely new. Retailers and streaming platforms have used recommendation algorithms for years. What generative AI and emerging agents add is the ability to combine customer information with content creation, conversational interactions and automated decisions across channels.
Salesforce’s 2026 State of Marketing report, based on 4,450 marketers across 26 countries, found that 78% needed more personalised content than they could produce, while 84% acknowledged that their organisations continued to run generic campaigns.
The findings suggest that although marketers are investing in AI, delivering genuinely relevant experiences remains difficult.
1. From Customer Segments to Individual Decisions
Consider a customer who regularly buys expensive skincare products.
A traditional recommendation system might continue promoting premium products because previous purchases suggest an interest in that category.
However, the customer could currently be looking for an inexpensive gift.
An AI-supported system using more recent information could potentially recognise that change and adjust its recommendations.
The difference is not simply showing a different advertisement. It is understanding that a customer’s immediate requirements may differ from their established purchasing patterns.
Individualisation can also involve deciding when not to communicate.
A customer who has already purchased a refrigerator may not need repeated advertisements for the same model. More useful communication could involve installation information, warranty registration or after-sales support.
Such decisions require more than generative AI.
Customer data platforms can help connect information across applications and channels. Predictive models can estimate likely customer actions, while generative systems can produce relevant messages. AI agents may eventually coordinate and execute some of those decisions within defined limits.
However, the technology only becomes useful when companies can connect reliable information with an appropriate response.
Deloitte Digital’s 2026 research, based on surveys of 1,000 US consumers and 480 senior business-to-consumer brand leaders, highlighted a significant disconnect.
While 87% of surveyed brand leaders rated their personalisation capabilities as good or excellent, only 29% of consumers agreed.
The findings suggest that companies may be overestimating the value customers receive from existing approaches.
Kimberly Storin, Head of Global Field Marketing, GTM and ABM at Zoom, described the ambition behind more advanced customer experiences as “answering their questions before they even know that it’s a question that they have.”
For marketers, achieving that level of relevance involves understanding not only previous behaviour but also the context of a current interaction.
A customer searching for flights during a school holiday may have different requirements from someone travelling alone for work. Similarly, an online shopper browsing winter clothing may be purchasing for another person.
The challenge is interpreting these circumstances accurately without making unnecessary assumptions.
An individualised recommendation is useful only when it reflects what the customer actually needs.
2. How Retail, Travel and Banking Are Moving Towards One-to-One Experiences
The shift is becoming visible across industries where companies manage large amounts of customer information.
In retail, AI can support product recommendations based on recently viewed items, previous purchases, spending preferences and availability.
A fashion retailer, for example, could avoid repeatedly recommending clothing sizes a customer has returned. Instead, it could highlight alternatives that better match the person’s previous choices.
Streaming platforms are also exploring more detailed recommendations.
HBO Max has experimented with using generative AI to help recommendation systems interpret additional emotional and social context within programmes, moving beyond conventional genre classifications.
This could eventually help platforms distinguish between someone who generally enjoys comedy and someone looking for something comforting to watch on a particular evening.
Travel companies are pursuing similar capabilities.
In July 2026, Radisson Hotel Group collaborated with Accenture to launch an AI-powered hotel discovery experience within ChatGPT.
The service allows travellers to search, compare and explore accommodation through conversational requests. Deeper personalisation, loyalty recognition and booking-related capabilities have been identified as future extensions.
Gianni Di Fede, Chief Commercial Officer at Radisson Hotel Group, said, “AI is radically transforming how people search for and book hotel stays.”
The development illustrates how a customer’s immediate request could become more important than a predetermined marketing segment.
Someone searching for a hotel may specify a budget, accessibility requirements, preferred location and travelling companions in a single conversation.
However, conversational hotel discovery should not be confused with a fully individualised customer journey. Some advanced capabilities remain planned rather than universally available.
Banking presents another application, particularly in India.
A financial institution may have information about an existing customer’s savings, recurring transactions and financial products, subject to applicable privacy requirements.
Instead of categorising people only by income or age, AI could help identify which services may be relevant to different financial circumstances.
EY India’s 2026 customer-experience research surveyed 2,030 banking customers across urban, semi-urban, semi-rural and rural markets.
It found that 49% of respondents classified as Rising Professionals and 51% of entrepreneurs prioritised smart digital capabilities, including AI-powered financial advice, automated savings and budgeting support.
However, half of the Golden Transitioner segment preferred human interaction over AI.
Pratik Shah, National Financial Services Leader at EY India, said banks must “balance automation with human touch to maintain empathy, overcome legacy system constraints, scale personalization, and uphold stringent data privacy standards.”
The findings highlight why individualisation cannot mean offering every customer a more automated experience.
For some, convenience may involve a quick AI-generated response. Others may prefer speaking directly with an employee, particularly when dealing with sensitive financial decisions.
3. AI Can Generate Thousands of Messages, But Can It Make Them Relevant?
One of the biggest changes AI brings to marketing is the ability to produce content at scale.
Previously, a marketing team might create separate campaigns for a handful of customer segments because producing numerous versions required considerable time and resources.
Generative AI can now help prepare variations of advertisements, emails, product descriptions and promotional messages much faster.
But producing more variations does not automatically create individualisation.
A customer receiving five different versions of an irrelevant advertisement is still being targeted with irrelevant advertising.
Bobby Jania, Salesforce’s Agentforce Marketing CMO, described the problem: “We are using the most powerful technology in history to send more one-way spam, faster.”
The limitation is especially important as marketing teams experiment with AI agents capable of taking action across multiple systems.
Individualisation requires decisions about which messages should be delivered, what information should inform them and whether communication is necessary.
Infosys Knowledge Institute’s CMO Radar 2026 surveyed 1,503 senior marketing leaders at companies generating more than $1 billion annually across nine countries and 14 industries.
The research found that 17% of personalisation initiatives had been cancelled or implemented without generating value, while only 15% were reported to have delivered significant value.
Among surveyed leaders, 14% were using agentic AI, and just 5% had embedded it throughout workflows.
These findings suggest that many organisations remain in the early stages of coordinating AI-driven decision-making.
A company may have customer information, content-generation tools and recommendation software without connecting them effectively.
Salesforce’s research also found that marketers with satisfactorily unified customer data were 60% more likely to use AI agents.
However, not every organisation had complete access to customer information across departments.
That can create problems when a customer interacts with the same brand through different channels.
For example, someone who has already raised a service complaint could receive an unrelated promotional offer because the marketing system does not have access to the latest service information.
The message may be personalised using purchase history, but inappropriate given the customer’s circumstances.
Individualisation therefore depends on coordination between marketing, sales, commerce and customer service, not simply generating more creative material.
4. When Does Individualisation Become Too Personal?
Greater accuracy introduces another question: how much does a brand need to know about someone?
Consumers may appreciate recommendations based on products they have deliberately browsed. They may react differently when advertising appears to anticipate private decisions or infer personal circumstances.
Canva and The Harris Poll’s 2026 research surveyed 1,415 marketing leaders and 3,547 consumers across seven countries, including India.
It found that 52% of consumers considered advertising invasive when it appeared to know what they were about to purchase before they had searched for it.
Another 58% did not want brands using AI to predict their needs, while 80% wanted greater control over how personal advertising could become.
The findings show that relevance and acceptance are not necessarily the same.
A recommendation can be accurate yet still make someone uncomfortable about how their information was collected or interpreted.
The concern is especially significant in sectors involving financial, health-related or other sensitive information.
An AI system may identify patterns suggesting a customer could benefit from a particular service. However, whether the organisation should use those signals for marketing is a separate question involving consent, privacy and responsible data practices.
Deloitte’s research also found that 58% of consumers valued clear and respectful data use, while 38% wanted greater direct control over how companies personalised their experiences.
For businesses, this adds requirements beyond improving prediction accuracy.
Customers may need understandable explanations of how information is used, meaningful privacy choices and the ability to correct preferences that no longer represent them.
Data quality remains equally important.
Outdated purchases, shared devices and incorrectly linked profiles can produce recommendations based on assumptions that are no longer relevant.
AI systems can also reproduce biases from historical data or generate inappropriate responses when information is incomplete.
These limitations suggest that human oversight, clear operational boundaries and accurate customer information remain important even as automated decision-making advances.
The movement from personalisation to individualisation is therefore unlikely to be measured simply by how many advertisements a company can customise.
A more meaningful measure will be whether customers receive better information, fewer irrelevant messages and support suited to their circumstances.
For marketers, AI offers the possibility of treating individual customers as more than members of a demographic category.
But the research suggests that businesses still face substantial challenges in connecting data, understanding customer preferences and ensuring that increasingly personal experiences remain welcome.
AI is making it easier for brands to decide what an individual customer might need next. Whether that becomes better customer engagement or more intrusive marketing will depend on how accurately those decisions are made and how much control consumers retain over their information.
The article uses five 2026 datasets from Salesforce, Deloitte Digital, EY India, Infosys Knowledge Institute and Canva/The Harris Poll, alongside four attributed expert quotations. It distinguishes deployed applications from proposed capabilities and keeps the research limitations in context.
For the headline, I recommend retaining the data-led version above. The comparison is based specifically on Deloitte’s surveyed US brand leaders and consumers, rather than a global consensus, and that distinction is made clear in the article.
Disclaimer: This article may include information derived from interviews, press releases, public statements, research, company communications and other publicly available or third-party sources. Such material may be summarised, paraphrased or contextualised for journalistic and editorial purposes. All rights in third-party content remain with their respective owners.
Related Articles
