BFCM Was Busy, AI Was Busier: The AI Launches Reshaping Ecommerce
2026-03-10 23:37 Diff
https://www.klaviyo.com/blog/how-ai-changed-bfcm-in-2025

BFCM is finally behind you. The campaigns are shipped, the carts have converted, and your brain has likely moved on from monitoring dashboards to reclaiming sleep. But while you were keeping the revenue machine humming, the AI world… did not take a break.

In fact, the biggest week in ecommerce was also one of the biggest weeks ever for AI. New shopping interfaces dropped. Retailers launched entire app experiences inside ChatGPT. Payments moved inside LLMs. And agentic commerce—the idea that AI becomes an active shopping participant, not just a recommendation engine—leapt from theory to reality.

Why should you care? Because AI is no longer just a tool for marketers. It’s becoming a channel in its own right. A place where customers browse, compare, ask questions, and increasingly buy without ever leaving the interface.

As you start planning for 2026, AI belongs right next to email, SMS, and site experience on your channel strategy roadmap. It’s already shaping discovery, influencing purchase decisions, and rewriting expectations around personalization.

If 2025 was the first AI-powered peak season, 2026 will be the first where AI is a front door. The brands preparing now—with clean data, AI-friendly content, and real 1:1 personalization infrastructure—will be the ones customers can actually find, trust, and buy from in an AI-first world.

Here’s what happened, and why it matters now.

OpenAI launches ChatGPT Shopping Research, a new entry point for product discovery

On November 24, OpenAI released ChatGPT Shopping Research: a new experience directly in the conversational ChatGPT interface that “does the research for you to help you find the right products.”

“It’s everything you like about deep research,” explained OpenAI’s announcement on X, “but with an interactive interface to help you make smarter purchasing decisions.”

The new experience simplifies product research and marks a major improvement over the previous one: users simply select “Shopping research,” describe what they’re looking for, answer a series of clarifying questions, and receive a personalized shopping guide of potential products that fit the bill.

What B2C marketers should do about it

As agentic commerce continues to reshape what consumers expect from the brands they’re buying from, it’s more important than ever to adapt your website and online presence more generally for LLMs like ChatGPT. That means:

  • Auditing your structured product metadata to make sure it’s clear and crawlable, it loads fast, and it contains as much information as possible (e.g., accurate GTINs and SKUs)
  • Writing AI-friendly product descriptions that answer cross-platform questions, provide comparative context, and include specific use cases and benefits
  • Testing how your listings appear in AI shopping results (e.g., visibility, appearance rate, context, accuracy, gaps)
  • Posting AI-optimized educational content, like detailed and descriptive FAQs, long-form articles with question-based headers, and comparison guides that help LLMs understand where your products fit in the market
  • Investing in PR and content marketing partnerships to get high-quality mentions of your product and brand in reputable third-party publications and websites

Perplexity responds to OpenAI’s Instant Checkout by teaming up with PayPal to launch Instant Buy

Also on November 25, Perplexity responded directly to OpenAI’s Instant Checkout with its own in-chat purchasing experience: Instant Buy, built in partnership with PayPal. The timing alone is telling: this was a full-scale interface overhaul shipped two days before peak shopping, signaling just how fast the AI commerce arms race is moving.

Unlike its previous “Buy with Pro” model, Instant Buy is now free for all users, and Perplexity is positioning it as a more intuitive, less ad-influenced alternative to Google Shopping and ChatGPT’s list-based discovery. Perplexity leans hard into its differentiators:

  • Discovery that thinks the way you think: delivering fewer, more contextually relevant product cards instead of generic top-ten lists.
  • A shopping assistant with memory: it learns your style, preferences, and past behaviors to improve recommendations over time.
  • Anti–Google Shopping positioning: Perplexity explicitly frames its results as what you actually want, not what advertisers want to push first.
  • Don’t dig through digital racks: Perplexity narrows choices to a curated handful of high-signal products, eliminating endless scrolling.
  • Stay in the flow: PayPal’s Fastlane and wallet features turn recommendations into instant purchases without leaving the chat.

This strategic positioning matters because it represents the shift from search to AI curation, where engines don’t just index products, but they interpret your intent and make increasingly personal suggestions based on memory.

What B2C marketers should do about it

For right now, Instant Buy is only available for larger retailers, and it seems like Perplexity is handpicking who to include. In 2026, that will likely change as the LLM builds partnerships with commerce platforms and allows many more brands access to Instant Buy.

Still, payment surfaces haven’t expanded like this since Amazon launched one-click check-out. And there are things you can do right now to be prepared:

  • Your structured product data must be flawless. AI agents can only curate well if they understand your catalog. GTINs, SKUs, attributes, materials, measurements, and use cases—all of it must be clean, crawlable, and complete.
  • Prepare messaging for an environment where the AIm not the shopper, initiates product comparisons.
  • Invest in conversational surfaces (web chat, RCS, WhatsApp) that mirror how shoppers are already interacting with LLMs.
  • Train your AI customer agent on your full catalog and customer data so that when consumers shift from Perplexity/ChatGPT back to your owned channels, the experience feels equally intelligent and personal.

AI became the engine of BFCM, and Klaviyo data shows just how much it moved the needle

AI wasn’t just happening around BFCM this year. It was powering it.

Across Klaviyo brands, AI became one of the strongest performance levers of the entire weekend, reshaping how marketers targeted audiences, personalized onsite experiences, and supported customers in real time. The numbers were clear:

  • 45% more brands used AI-driven product recommendations in email
  • 68% more messages were powered by AI recommendations
  • These messages generated a 71% increase in revenue YoY

This wasn’t just message-level impact. AI changed onsite behavior:

  • Shoppers who engaged with AI-powered personalized pages via Customer Hubviewed 2.4 more pages per session
  • Unified data + AI created significantly deeper engagement, stronger intent signals, and more efficient conversions

AI even showed up as a service channel: Brands, especially in apparel, swimwear, and fitness, used K:AI Customer Agent to answer fit questions and guide purchases, reducing friction in moments that would have otherwise caused cart abandonment.

All told, BFCM 2025 was the first peak season where:

  • AI didn’t just personalize a few messages. It personalized the entire shopping journey at scale.
  • AI didn’t just assist shoppers. It materially increased revenue.
  • AI didn’t just optimize. It shaped the behavior of both new and repeat buyers.

This is a great looking glass into 2026, a year where AI becomes the connective tissue between marketing, service, and commerce.

What B2C marketers should do about it

AI is now influencing discovery, browsing, trust, and conversion across the entire shopper journey. As a marketer and service leader, that should change how you approach your 2026 plans. Here are our recommendations.

  1. Unify your customer data because AI can’t personalize what it can’t see. Brands that centralized browsing, purchase, loyalty, and service data saw the strongest incremental lift. Unified profiles gave AI the context needed to deliver accurate recommendations, better segmentation, and on-brand answers through Customer Agent.
  2. Scale AI recommendations across the full lifecycle, not just peak season. BFCM results showed AI recommendations convert in browsing moments, replenishment flows, post-purchase nudges, and website personalization. AI isn’t just a tool for BFCM season. It’s a year-round revenue driver.
  3. Build AI-powered website personalization into your standard shopping experience. Customer Hub’s performance proves shoppers want guidance: fit, sizing, loyalty status, shipping timelines, bundle suggestions, and contextual recommendations. Replicate that “help me shop” experience everywhere your customers land.
  4. Treat AI service as a conversion tool, not a cost-saving tool. The highest-performing brands used Customer Agent in carts and at checkout to answer fit, shipping, or product detail questions that normally create hesitation. When AI eliminates last-mile uncertainty, it directly unlocks revenue.
  5. Use AI-led segmentation before promotions to improve margin and efficiency. By identifying high-intent shoppers early, brands warmed key segments before discounting. This was one reason revenue increased while discount rates fell for BFCM 2025.
  6. Invest now in 1:1 personalization infrastructure for 2026. The data is unequivocal: the brands that entered BFCM with real AI foundations—unified profiles, predictive intelligence, preference data, and AI-powered messaging—outperformed those relying on batching and blasting.

Target becomes the first major retailer to launch an official ChatGPT app

On November 25, Target became one of the earliest major retailers to launch an integrated ChatGPT app—a strategically meaningful move, even if the execution is still early. In its current state, the app only installs reliably for some Pro users, and the experience is occasionally buggy while the teams work through permissions and load issues.

Still, the vision is clear: Target is building a future where ChatGPT becomes the interface, and Target’s catalog and account experience power the fulfillment behind the scenes.

“We wanted to make it easier to shop by taking the pressure off needing to know exactly what you want or having to sift through an endless number of search results,” said Purvi Shah, vice president, UX design, research and accessibility at Target, in the company’s press release.

“A guest can ask a broad conversational question, and Target will respond with those curated, relevant, on-point ideas,” Shah added. “We might even suggest things that you didn’t even know that you needed.”

The app acts as a personal shopping buddy. Users can simply type “Target,” ask a question, and receive curated suggestions across the retailer’s full assortment. They can browse, add to cart, and check out, all without leaving the chat.

The significance isn’t the glitches. It’s the direction of travel:

  • This is the first major retailer signaling that LLM-based shopping is not an experiment. It’s a channel.
  • It reinforces the shift from search → AI curation → in-chat purchasing.
  • And it demonstrates how essential clean structured data, SKU-level clarity, and product attributes will be for discoverability.

What B2C marketers should do about it

Keep a close eye on how big-box retailers like Target and Walmart are integrating AI into their shopping experiences. In the same way a mobile app doesn’t make sense for every brand, not every brand will need or even want a ChatGPT app. But every brand does need:

  • Structured product data optimized for LLMs: attributes, GTINs, materials, descriptions.
  • AI-friendly product storytelling that clearly communicates who the product is for, when it’s used, and how it compares.
  • A plan for AI agents to represent your brand with accuracy and consistency across channels.

Note that ChatGPT’s Instant Checkout feature, coming soon for Shopify and Etsy merchants, and Perplexity’s Instant Buy feature will level the playing field for smaller brands soon—but an app gives retailers more control over the experience. As this ecosystem matures in 2026, brands that are LLM-ready will capture a disproportionate share.

AI should enhance connection, not replace it

The past week reinforces something we’ve believed for years: AI should make brands more human, not more generic.

As discovery shifts off-site and into AI environments, retailers need to show up consistently with accurate data on every customer and a recognizable voice when reaching out to them. That means:

  • Real-time customer profiles that reflect browsing, buying, and service interactions
  • AI-powered 1:1 personalization that adapts offers, timing, and content for every individual
  • AI marketing and customer service agents trained on brand storefronts and customer data with humans in the loop, so outputs stay on-brand and grounded in reality
  • Connected marketing and service infrastructure, built on the same customer context

Peak season ends. Customer expectations don’t. As you look back on AI news over BFCM, ask: Is my brand ready for discovery that starts outside my site? Or is it only optimized for journeys that begin on it?

If you’re preparing for the first, you’re setting the pace.