From Segments to Individual Memory: Why AI Is Replacing How DTC Brands Think About Customers

The era of broad customer segments is ending. What's replacing it is something more like a relationship, and the brands that adapt first are already seeing results.

For the past fifteen years, "personalization" in DTC has meant one thing. Segmentation. You take your customer list, slice it into buckets like VIPs, lapsed buyers, or dog owners in the Northeast, and you send each bucket a slightly different version of the same email. The bucket gets the message. The individual doesn't.

That model is breaking down. It's not that marketers have lost faith in it. It's that customers themselves don’t always behave like their specific segment. They behave like individuals with context. And they now expect brands to remember that context, respond to it, and act on it in real time.

The conversation has moved past "AI for email writing" into something more fundamental. Customer segments are being replaced by persistent, individual memory. Here's what that means, why it's happening now, and what changes for DTC brands as a result.

What's Actually Wrong With Segmentation

Segmentation made sense in an era of constrained data and constrained channels. When the only way to reach a customer was a weekly email blast, slicing your list into eight or ten cohorts was a real improvement over sending everyone the same message.

The problem is that the customer has moved on. A "VIP loyalty segment" treats someone who bought twice last quarter the same as someone who clicked four product pages this morning, opened a return last week, and asked your chatbot about sizing yesterday. Those are different people in any meaningful sense. To most martech stacks, they look identical.

McKinsey's research on personalization is blunt about where this ends up. Eventually, each customer becomes a segment of one. McKinsey partner Eli Stein has described traditional personalization as "Mad Libs." You fill in the blanks at scale, but in a way that has almost depersonalized the experience. The bar customers now apply is much higher. Over 75% of consumers say they're turned off by content that doesn't feel relevant to them, and that threshold keeps rising.

Segments can't clear that bar. Individual memory can. The shift is happening because the technology to operate at the level of the individual finally exists, and that's where the next customer picks up.

Why This Shift is Happening Now

Two things are happening together. The first is a change in how consumers shop. The second is a change in what platforms can actually do.

On the consumer side, the most useful data comes from Klaviyo's 2026 AI Persona Research, which surveyed nearly 8,000 consumers across eight countries. Three findings tell the story:

• 60% of consumers now use AI at least weekly, and more than one in five start with AI tools when making decisions, learning something new, or solving a problem.
• 30% of AI searches contain eight or more words, and 78% of consumers include emotional or personal context at least some of the time when searching with AI.
• 41% of global shoppers have purchased an AI-recommended product in the past six months. Klaviyo's own platform data shows traffic from AI-referred sources like ChatGPT and Gemini surged 1,936% year-over-year.

Consider a consumer who types "I live in Chicago and want a red winter coat that's waterproof and has a fur lining." Klaviyo's research includes that as a real example. You can't serve that person with a segmented email about "outerwear shoppers." They've already told you who they are, where they are, what they need, and why. The job is to remember that and act on it. Not to drop them into a bucket and send a generic "winter is here" campaign two weeks later.

On the platform side, AI agents are what makes individual memory operationally possible. Gartner forecasts that 40% of enterprise applications will be integrated with task-specific AI agents by the end of 2026, up from less than 5% in 2025. By 2029, agentic AI will autonomously resolve 80% of common customer service issues without human intervention. The infrastructure is no longer theoretical. It's here.

What Individual Memory Looks Like in Practice

Traditional CRMs store customer data as static attributes that sit still until a marketer reaches for them. A CRM built around individual memory works differently. Every interaction flows into a single profile in real time. A browse, a click, a return, a chat conversation, a loyalty redemption. An AI agent sitting on that profile can act instantly, without a marketer building a segment first.

Klaviyo's product architecture is one of the more explicit attempts to build for this model. The company positions itself as an AI-first B2C CRM that unifies data, marketing, service, and analytics in one place. Two main pieces matter:

• Customer Agent is the consumer-facing layer. It's a 24/7 AI assistant that resolves issues, answers product questions, recommends products, and escalates to a human with full context.
• Marketing Agent is the brand-facing layer. It autonomously plans campaigns, creates on-brand content, personalizes every send, and learns continuously.

Both run on the same underlying data layer. The conversation a shopper had with Customer Agent at 11pm informs the campaign Marketing Agent suggests at 9am. There's no handoff, no sync, no nightly batch job. That's the difference between AI bolted onto a CRM and a CRM built around AI.

The proof points back it up. Folk Clothing, a London menswear brand, saw 53% of support conversations resolved by Customer Agent within 90 days, alongside a 75% reduction in average ticket resolution time in 30 days using Klaviyo Helpdesk. 

Ministry of Supply, the MIT-founded apparel brand, saw Customer Agent resolve 84% of product recommendation queries and 75% of "where is my order" queries in 60 days. In both cases the agent isn't just deflecting tickets. It's converting service interactions into sales, because it has the full customer context to do so.

[inline-cta title="Try Klaviyo for free—no commitment required" link="https://klaviyo.sjv.io/ZVBbz0"]

What This Changes for DTC Teams

A few practical shifts worth thinking through.

The segment isn't dead, but it's downstream now 

Segments still matter for cohort analysis, LTV modeling, and paid spend planning. But they're no longer the unit of execution. The unit of execution is the individual customer profile, updated in real time, accessible to every channel and agent.

Service becomes a revenue function

When the same agent that answers a sizing question can surface a complementary product based on the customer's purchase history, the line between "support" and "selling" disappears. Gartner found that 77% of service and support leaders feel pressure from senior executives to deploy AI, and 75% report increased AI budgets. The people writing those checks aren't doing it to clean up the support inbox. They're doing it because the service conversation is now where a meaningful share of revenue is being influenced.

The data infrastructure question becomes existential

If individual memory is the new substrate of marketing, then having customer data fragmented across an email platform, a helpdesk, a loyalty app, and a Shopify backend isn't just inconvenient. It's strategically disqualifying. The fix is to consolidate onto a system where every customer interaction writes to the same profile in real time, so that every agent, channel, and team is reading from the same source of truth.

Trust calibration is the new segmentation skill

Only 13% of consumers completely trust AI, and 40% of AI Enthusiasts say they notice low-quality or generic "AI slop" from brands multiple times per week. The brands leaning hardest into AI are also the ones whose customers are most likely to punish them for using it badly. 

The skill marketers used to apply to segmentation, namely figuring out who wants what message, now applies to AI itself. The answer is to calibrate how aggressively you deploy agent-driven experiences based on what each customer actually wants, which is exactly what a profile built on individual memory makes possible.

The Honest Truth of What’s Changing

Brand still matters. Creative still matters. The fundamentals of what makes a product worth buying doesn’t go away because AI now sits between the brand and the shopper. If anything, they matter more. A customer interacting with your agent is interacting with the most distilled version of your brand voice you've ever shipped.

What's changing is the operational model. The unit of marketing is shifting from segment to individual. The data layer is shifting from siloed to unified. The interface is shifting from one-way broadcast to two-way conversation. The platform requirement is shifting from a tool that sends messages to a system that remembers, learns, and acts.

For DTC operators evaluating where to invest over the next 12 to 24 months, the question isn't whether to add AI features to your current stack. It's whether your current stack can actually support the model that's coming. One where every customer has a persistent, accessible memory that every channel can act on. Most stacks can't. Klaviyo is one of the small handful of platforms architected around that bet, and for operators thinking about what the next era of customer relationships looks like, it's worth understanding.