Executive Summary

The 2025 holiday shopping season marks a definitive inflection point in digital commerce: the transition from algorithmic recommendation to Agentic Commerce. This report provides a comprehensive analysis of how Large Language Models (LLMs) and autonomous AI agents are reshaping the retail landscape, with specific focus on the critical “Cyber Five” period (Thanksgiving through Cyber Monday).

Synthesizing data from primary market intelligence sources—Adobe Analytics, Salesforce, Shopify, Mastercard SpendingPulse, and Google—the research reveals a fundamental restructuring of the consumer purchase journey. Consumers are no longer passive recipients of targeted ads; they are increasingly employing AI agents to execute research, price comparison, and transactions on their behalf.

Key Findings:

Traffic Surge: AI-driven traffic to retail sites is projected to rise by 520% year-over-year, peaking in the ten days leading up to Thanksgiving.
Economic Impact: AI agents are expected to influence approximately $73 billion in global sales during Cyber Week alone, with retailers deploying agentic tools seeing 7x higher sales growth compared to non-adopters.
Conversion Efficiency: Traffic originating from generative AI sources is converting at a rate 16% higher than non-AI traffic, reversing historical trends where exploratory AI traffic lagged in conversion.
The Agentic Checkout Reality: The 2025 season marks the commercial debut of scalable “Buy for me” functionality, where platforms allow agents to bypass traditional merchant front-ends to finalize purchases securely.
Behavioral Shift: A “Value and Vibes” consumer mindset is driving AI adoption not for novelty, but for extreme efficiency in deal-hunting, with 64% of shoppers utilizing AI to stretch inflation-impacted budgets.

This report details the macroeconomic backdrop, the technological mechanisms of agentic commerce, specific metrics for monitoring Black Friday 2025, and strategic implications for the retail ecosystem.

Part I: The Macro-Economic Crucible of Holiday 2025

To understand the explosive adoption of Agentic AI in 2025, one must first analyze the economic pressure cooker in which this technology is being deployed. The consumer behaviors driving LLM usage are deeply rooted in the financial realities of the post-inflationary economy.

1.1 The “Value and Vibes” Economy

The overarching theme for the 2025 holiday season, as identified by the Mastercard Economics Institute, is “Value and Vibes.” This dichotomy defines a consumer who is hyper-sensitive to price elasticity on commodity goods yet willing to spend disproportionately on goods that offer high emotional resonance or cultural capital.

The Value Component: Continued inflationary pressure, though stabilizing, has permanently elevated the baseline cost of living. Consumers are entering the holiday season with depleted savings relative to the stimulus-era peaks. This financial constraint makes the “deal-hunting” capability of AI not a luxury, but a necessity. The labor of finding the absolute lowest price—previously a manual task requiring hours of cross-referencing—is now offloaded to LLMs.

The Vibes Component: Despite economic headwinds, spending is not crashing; it is concentrating. Shoppers are prioritizing “experiences” and “viral products” (often discovered via TikTok or Instagram). LLMs play a crucial role here by acting as cultural interpreters, translating vague requests (e.g., “gift for a brat summer enthusiast”) into specific product SKUs.

1.2 “Black November” and the Dilution of Event Days

The traditional “Cyber Five” (Thanksgiving, Black Friday, Small Business Saturday, Sunday, Cyber Monday) remains the peak, but its gravitational pull has weakened in favor of a month-long promotional cycle dubbed “Black November.”

The October Pull-Forward: Retailers, anticipating a competitive environment, began aggressive discounting as early as October 1st. Adobe Analytics data confirms that $88.7 billion was spent online in October 2025 alone, an 8.2% YoY increase. This surge was partly driven by competing events like Amazon’s Prime Big Deal Days.

The AI Implication: The elongation of the shopping season increases the utility of AI price-tracking agents. A human shopper cannot monitor price fluctuations across 50 items for 60 days. An AI agent can. Consequently, consumers are using LLMs to “watch” products starting in October, with instructions to execute purchases only when historical lows are breached during Black Friday week.

1.3 Inflation, Tariffs, and Spending Forecasts

Despite the economic pressures, the hard numbers project a season of moderate but resilient growth:

Table 1: Holiday 2025 Spending Forecasts

Forecast Entity	Metric	Projection	Context
Adobe Analytics	U.S. Online Spending	$250 Billion (+5.3% YoY)	Driven by mobile & AI efficiency
Salesforce	Global Cyber Week Sales	$334 Billion	Record-breaking volume
Mastercard	U.S. Retail Sales (Ex-Auto)	+3.6% YoY	Includes offline & online
Shopify	Planned Spend per Shopper	$192 (up from $155 in 2024)	Indicates higher intent, likely inflation-adjusted

The looming threat of potential tariffs has also created a “buy now” urgency in certain categories, particularly electronics and imported goods, further incentivizing the use of AI to find inventory before price hikes materialize.

Part II: The Agentic Revolution—Technology & Mechanisms

The defining shift of 2025 is the evolution of AI from Conversational (chatbots) to Agentic (doers). In 2023-2024, LLMs were search engines that could talk. In 2025, they are shoppers that can buy.

2.1 Defining Agentic Commerce

Agentic commerce involves users authorizing AI agents to shop on their behalf. Users provide high-level intent, and the agents execute the low-level logic of discovery, comparison, negotiation, and transaction.

The Three Tiers of Agentic Capability:

Informational Agents: “Find me the best TV under $500.” (Output: A list/summary)
Monitoring Agents: “Track the price of this TV and alert me if it drops.” (Output: Notification)
Transactional Agents: “Buy this TV if it drops below $450 at a retailer with 2-day shipping.” (Output: A receipt)

2.2 Google’s Agentic Checkout & “Buy for Me”

Google has aggressively integrated agentic capabilities directly into the Shopping Graph. This is arguably the most significant disruption for Black Friday 2025.

Mechanism: When a user finds a product via Google Search (or “AI Mode”), they are presented with a “Buy for me” option for eligible merchants (including Wayfair, Chewy, and select Shopify stores).

Process:

Authorization: The user authenticates with Google Pay and sets parameters (price, size, color).
Execution: The Google Agent navigates the merchant’s checkout flow in the background. It is not an API call in the traditional sense; the agent simulates a user session but at machine speed, inputting shipping and payment details securely.
Verification: To prevent fraud, the system uses “verifiable digital credentials” to prove to the merchant that the agent is acting on behalf of a specific, authenticated human.
Inventory Checks: Google’s agents also utilize Duplex technology to physically call local stores to verify in-stock inventory for “near me” searches, solving the “ghost inventory” problem that plagues holiday shopping.

2.3 Salesforce Agentforce & The Retailer’s Defense

While Google empowers the consumer’s agent, Salesforce empowers the retailer’s agent. “Agentforce” represents a suite of autonomous agents deployed by brands to handle the influx of AI and human queries.

Impact: Retailers utilizing Agentforce for customer service and sales experienced 7x higher sales growth (13%) in the weeks leading up to Cyber Week compared to those who did not (2%).

Function: These agents do not just answer “Where is my order?” They can actively upsell (“That jacket goes well with these boots”), negotiate bundles, and resolve complex shipping queries without human intervention. They are the “defensive line” handling the “offensive” consumer agents.

2.4 Mastercard Agent Pay & Tokenization

A critical enabler of this ecosystem is trust. Consumers are hesitant to give an AI their credit card number. Mastercard’s “Agent Pay” solves this via tokenization.

The Protocol: The user’s payment credential is replaced by a unique token restricted to a specific agent, merchant, or timeframe.

Integration: This allows agents (like those in Microsoft Copilot or specialized shopping bots) to execute transactions across the open internet without ever exposing the underlying funding source. This infrastructure is essential for the “Zero-Click” commerce vision.

2.5 OpenAI & “Instant Checkout”

OpenAI has rolled out “Instant Checkout” in the US, allowing ChatGPT Plus users to purchase third-party products directly within the chat interface. This moves the transaction point away from the retailer’s website and into the LLM, fundamentally altering the attribution model for web traffic.

Part III: Quantitative Analysis—Traffic, Conversion, & Attribution

The most significant data regarding the 2025 holiday season comes from the metrics surrounding AI traffic and conversion. The narrative that AI is “just a research tool” is empirically dead; the data shows it is a closing tool.

3.1 The Traffic Surge

Adobe Analytics forecasts a 520% year-over-year increase in traffic referred by AI sources during the 2025 holiday season.

Timing: This traffic is expected to peak in the 10 days leading up to Thanksgiving, indicating that consumers are using AI primarily for the “preparation phase”—building wish lists, comparing specs, and setting price alerts.

Category Affinity: AI services are most heavily utilized for:

Toys: Finding the “hot” toy before it sells out
Electronics: Spec comparison
Jewelry: Visual search/inspiration
Personal Care: Routine replenishment

3.2 The Conversion Inversion

Historically, “high-intent” traffic came from search engines (Google), while “low-intent” traffic came from social media or browsing. AI was initially viewed as “low-intent”—people experimenting with chatbots.

October 2025 Data Reversal:

Metric	AI-Referred Traffic vs. Non-AI
Conversion Rate	+16% higher
Revenue Per Session	+8% higher
Time on Site	+44% longer
Pages Viewed	+12% more

Interpretation: This data suggests that the LLM acts as a “pre-qualifier.” By the time the agent sends the user to the retailer, the user has already asked the questions, compared the alternatives, and validated the price. The click is not “exploratory”; it is “transactional.”

3.3 The Agentic Premium

Salesforce data corroborates the Adobe findings but from the retailer’s perspective.

AI-Influenced Sales: Salesforce projects $73 billion of global sales will be influenced by AI during Cyber Week, up 22% from 2024.

Effectiveness: The 7x growth gap between retailers with agents and those without suggests that Agentic Commerce is a winner-take-all dynamic. Retailers that expose their data to agents (via APIs or structured data) win the sale; those that don’t remain invisible.

Part IV: Consumer Behavior & The “SantaGPT” Phenomenon

Who is using these tools, and how? The 2025 shopper is not monolithic, but distinct patterns have emerged across demographics.

4.1 Generational Divides

Adoption of AI shopping tools is heavily skewed by age, creating a digital divide in deal access.

Generation	AI Shopping Tool Adoption
Gen Z (18-24)	84% plan to use AI tools
General Population	64% plan to use AI
Primary Use Cases	Finding deals (36%), Product research (53%)

For Gen Z, the AI is the primary search interface, replacing the Google search bar. The “Black Friday” tradition of camping out in lines is effectively dead, replaced by online browsing and agent delegation.

4.2 “SantaGPT” and Specialized Bots

A new class of specialized AI agents, colloquially termed “SantaGPTs,” has emerged. These are not general-purpose LLMs but fine-tuned models specifically designed for gift recommendation and logistics tracking.

Use Case: “Find a gift for a 35-year-old male who likes cooking and sci-fi, under $50, that can arrive by Dec 20.”

Psychology: Surveys indicate that 64% of shoppers are using these tools to reduce “holiday stress” and decision fatigue. The AI provides a curated shortlist, removing the “paradox of choice.”

4.3 Mobile Dominance

The smartphone remains the remote control of the shopping experience, but the nature of mobile shopping has changed.

Volume: Mobile is expected to drive 51.4% of online spend ($45.6 billion in October alone) and up to 70% of purchases on peak days like Black Friday.

The App-less Shift: Consumers are downloading fewer retailer apps. Instead, they interact via “Super Apps” (social media with embedded commerce) or mobile browser agents. Shopify reports that mobile shopping continues to dominate, with 57% of sales on Cyber Monday coming from smartphones.

Part V: The Cyber Five 2025 Forecast & Deep Dive

The “Cyber Five” period (Nov 27 - Dec 1) remains the Super Bowl of retail. Here is the day-by-day analysis based on aggregated data.

5.1 Thanksgiving Day (November 27)

Status: “Couch Commerce” Peak

Behavior: Consumers browse on mobile devices after dinner. AI agents are heavily active in “monitoring mode,” waiting for midnight deal drops.

Projected Dynamics: High mobile traffic, moderate conversion.

5.2 Black Friday (November 28)

Forecast: $10.8 billion in U.S. online sales

The Shift: Black Friday is now a “hybrid” event. While 65% of sales may still have a brick-and-mortar component (BOPIS - Buy Online, Pick Up In Store), the discovery is almost entirely digital.

Agentic Impact: This is the stress test for “Agentic Checkout.” Millions of “Buy for me” triggers set by users in previous weeks will fire simultaneously as prices hit their floor. Retailers with poor bot mitigation may crash; those with agent-ready infrastructure will dynamically allocate inventory.

5.3 Small Business Saturday & Sunday (November 29-30)

Focus: Local & Niche

Google’s Role: The “Local Inventory” agentic calling feature will see peak usage here. Users will ask Gemini, “Call stores near me to see if they have [specific artisanal product] in stock.”

5.4 Cyber Monday (December 1)

Forecast: $13.3 billion (+7.3% YoY) — The biggest day of the year

Dynamic: The “Desktop” day. However, in 2025, cross-device continuity allows a cart built by an agent on a phone to be executed on a desktop.

Discount Depth: Discounts are expected to peak at 30% off list price for electronics, driven by algorithmic pricing wars between Amazon and Walmart, monitored in real-time by consumer agents.

Part VI: Merchant Strategy—Surviving the Agentic Era

For retailers, the rise of LLMs is an existential challenge. The strategies that worked for SEO (Search Engine Optimization) do not work for LLMO (Large Language Model Optimization).

6.1 The “Zero-Click” Threat

The “Zero-Click” phenomenon—where users get their answer from Google/AI and never visit the website—is expanding to commerce.

The Risk: If an agent buys the product via “Agentic Checkout” or “Instant Checkout,” the retailer gets the sale but loses the customer relationship (email signup, cross-sell browsing, brand immersion).

The Counter-Strategy: Retailers are focusing on Post-Purchase Experience (unboxing, support) to capture the customer after the agent has done the deal.

6.2 Inventory & Bot Management

With agents polling prices every few seconds, server loads are immense.

“Good Bots” vs. “Bad Bots”: Retailers must distinguish between a Google Agent (buying for a legitimate user) and a Scalper Bot (buying for resale).

Real-Time Inventory: Retailers are bleeding ad spend on Google Shopping by advertising out-of-stock items. In a study of 500 retailers, 97% paid for clicks on unavailable products. In an agentic world, this is fatal; an agent that encounters an “Out of Stock” error will instantly pivot to a competitor and down-rank the retailer in future queries.

6.3 The Rise of “Hybrid” Shopping

Shopify reports that 43% of shoppers plan to discover products in-store but buy online, or vice versa.

Unified Commerce: Retailers using conversational AI are bridging this gap. “Scan this QR code to chat with our AI stylist” allows the in-store shopper to continue the conversation online.

Part VII: Monitoring Dashboard—Where to Track LLM Impact

For analysts, investors, and retailers, monitoring the impact of LLMs on Black Friday 2025 requires specific, real-time data sources. This section provides the definitive guide to “Mission Control” for Cyber Week.

7.1 Primary Data Sources

Table 2: Holiday 2025 Monitoring Resources

Platform	Resource URL	Primary Utility	2025 Specific Feature
Adobe Analytics	news.adobe.com	US Spend & AI Traffic	”Generative AI” traffic segment breakdown
Salesforce	salesforce.com/shopping-index	Global & Agent Impact	”Agent-Influenced Revenue” metric
Shopify	bfcm.shopify.com	Real-Time SMB Viz	Live Globe & Mobile/Cross-border stats
Mastercard	mastercardservices.com	Total (Online+Offline)	“Value and Vibes” economic analysis
Google Trends	trends.google.com	Intent & Interest	Search volume for “AI shopping” terms

7.2 Adobe Analytics (The Gold Standard)

Adobe offers the most granular view of transactional data.

Where to Monitor: news.adobe.com and business.adobe.com

Key Reports:

“Cyber Week Spend Forecast” (Baseline)
“Daily Holiday Recap” (Released each morning of Cyber Week)

Specific Metrics: Look for the “Traffic by Source” breakdown. Adobe explicitly separates “Generative AI” from “Search” and “Direct.” Tracking the YoY growth of this specific slice is the primary indicator of LLM impact.

7.3 Salesforce Shopping Index (The Agentic Monitor)

Salesforce is the go-to for global data and specifically agent performance.

Where to Monitor: salesforce.com/shopping-index (“Shopping Insights HQ”)

Key Metrics:

AI-Influenced Revenue: The percentage of total sales where AI played a role
Agent-Driven Sales: New for 2025, distinguishing between passive recommendation and active agentic service

Real-Time: Salesforce often provides a “flash” dashboard with near real-time order volume and cart abandonment rates.

7.4 Shopify Live Globe (The Visual Pulse)

For a visceral view of the SMB (Small and Medium Business) economy.

Where to Monitor: bfcm.shopify.com (The “Live Globe”)

What it Shows: Real-time transactions lighting up a 3D globe.

Interpretation: High density in specific regions correlates with “Shop Pay” usage. Since Shopify is a key partner for Google’s “Agentic Checkout,” high volume here is a proxy for agentic success.

7.5 Mastercard SpendingPulse (The Total Economy)

Where to Monitor: mastercardservices.com

Key Insight: “Value vs. Volume.” Mastercard data distinguishes between sales growth due to inflation (prices went up) and volume (more units sold). This is critical for validating the “efficiency” claim of AI.

7.6 Google Trends & Gemini Usage

Where to Monitor: trends.google.com

Metric: Monitor searches for “buy for me,” “AI shopping assistant,” or “agent checkout.” While Google doesn’t release public real-time checkout data, search volume for these intent keywords is a leading indicator.

Part VIII: Strategic Implications & Recommendations

8.1 For Retailers

Machine-Readability is the New SEO: Expose structured data via APIs and well-formatted feeds. Retailers invisible to agents are invisible to a growing segment of consumers.
Real-Time Inventory Accuracy: The agent penalty for out-of-stock is severe—immediate pivot to competitor and long-term down-ranking.
Post-Purchase Experience Investment: If you can’t own discovery, own the unboxing. Make delivery, packaging, and support so compelling that customers return directly.
Agent Infrastructure Readiness: Prepare for simultaneous “Buy for me” trigger events. Server capacity and bot management become mission-critical.

8.2 For Brands

The New Metrics That Matter:
- Inclusion rate across target queries
- Rank share within those inclusions
- Copy integrity (is the agent representing your product accurately?)
- Price position versus peers
Audit Agent Outputs: Monitor how LLMs describe and recommend your products. Inaccurate representation by agents requires correction at the data source level.
Observability Infrastructure: Build exposure logs and receipt telemetry. Traditional analytics don’t capture agent-influenced purchases.

8.3 For Investors & Analysts

Watch the 7x Gap: The growth differential between agent-adopters and non-adopters is not subtle. This metric will stratify retail performance.
Attribution Model Disruption: Companies with legacy attribution systems will undercount AI influence. Discount their reported metrics accordingly.
The “Zero-Click” Commerce Shift: Monitor which platforms capture the transaction point. The move from retailer checkout to LLM checkout redistributes value in the ecosystem.

Conclusion

The 2025 holiday season represents the end of the “Search Era” and the beginning of the “Syntax Era.”

Consumers are no longer searching for keywords; they are prompting with syntax. They are describing complex problems (“I need a warm coat that looks good in Paris in December and costs less than $200”) and expecting the machine to handle the logistics of discovery and transaction.

The data from Adobe (520% AI traffic growth) and Salesforce ($73B impact) proves that this is not a niche behavior for the tech-elite; it is mass-market adoption driven by the economic imperative of “Value and Vibes.”

The winners of Holiday 2025 will not necessarily be the retailers with the best doorbusters, but the retailers with the most machine-readable inventories.

As the “Cyber Five” approaches, the industry must look beyond the traditional metrics of foot traffic and page views. The real story of 2025 is happening in the invisible, high-speed exchanges between consumer agents and retailer agents—a digital negotiation occurring millions of times per second, reshaping the global economy in its wake.

The storefront was the interface. Now the agent is the interface. The storefront is just infrastructure.

That is the Agentic Shift.

Appendix: Data Sources & Methodology

This analysis synthesizes forecasts and data from the following primary sources:

Adobe Analytics: Holiday shopping forecasts, traffic attribution, conversion metrics
Salesforce Commerce Cloud: Global shopping index, agent-influenced revenue projections
Shopify: BFCM live data, mobile commerce statistics, Shop Pay adoption
Mastercard SpendingPulse: Economic analysis, value vs. volume breakdowns
Google: Shopping Graph data, Agentic Checkout announcements, Trends data

Projections are based on pre-season estimates published in November 2025. Actual results will be tracked through the monitoring resources detailed in Part VII.

Signal Dispatch Research | November 2025