Ecommerce AI Is Everywhere. Scaled AI Is Still Rare
Insights / Ecommerce AI Is Everywhere. Scaled AI Is Still Rare
Table of Contents
Ecommerce has moved well beyond asking whether AI belongs in the business. According to McKinsey and Stord, 89% of retailers have adopted AI in some form, but only 7% have actually scaled it across their operations.
That gap is the more interesting story. AI is already being used across personalization, customer service, fraud detection, pricing, and demand forecasting. But adopting individual AI tools is very different from making AI part of how the business operates.
So what is actually working in ecommerce AI and why are so few retailers getting the full value from it?
Where AI Actually Shows Up in Ecommerce
AI is already being applied across some of the most important parts of the ecommerce journey.
| Use Case | What It Actually Does |
|---|---|
| Personalization & recommendations | Surfaces the specific product or offer a shopper is likely to want, based on real behavior |
| Customer service | Resolves order, delivery, and return queries directly, at any hour |
| Fraud and risk detection | Distinguishes a genuinely risky transaction from a legitimate one a rule would have blocked |
| Dynamic pricing | Adjusts pricing based on demand, inventory, and competitive signals |
| Demand forecasting | Predicts inventory needs before a stockout or overstock happens |
These aren’t experimental use cases anymore. The question is increasingly how deeply retailers are integrating them into their operations.
Personalization: The Most Proven Area, and Still the Most Underused
Shoppers increasingly expect ecommerce experiences to reflect what they actually want. 71% of consumers expect retailers to recognize their individual needs and preferences, while 76% get frustrated when they don’t, according to McKinsey research.
There is also a clear commercial case. McKinsey found that faster-growing companies generate 40% more of their revenue from personalization than slower-growing companies. And the scale is already significant.
Salesforce measured $60 billion in online sales influenced by AI during Cyber Week 2024, alongside roughly 60 billion AI-powered product recommendations, up 21% year over year.
There’s another change happening in how shoppers discover products. Generative AI referral traffic to retail sites grew 693% year over year during the 2025 holiday season, according to Adobe Analytics, with more than 1 trillion visits tracked. Shoppers are increasingly arriving at ecommerce sites because an AI assistant recommended the store or product — rather than through a traditional search engine.
For retailers, that creates a new discovery channel to understand and optimize for.
Customer Service: Where Response Time Directly Affects Revenue
Order status, delivery delays, and returns make up a large share of ecommerce support volume. They’re also exactly the kind of structured, repetitive queries AI can handle well.
An AI system can resolve these questions immediately, at any hour, without making a shopper wait for an agent — including at the point where they are deciding whether to complete a purchase. But the biggest opportunity isn’t simply replacing a human response with an automated one.
The retailers getting more value from AI customer service are connecting support to the same customer and order data used by marketing and sales. That matters because a support conversation and a personalized offer shouldn’t be working from two completely different views of the same shopper.
Fraud Detection: A Bigger Problem Than the One Retailers Usually Solve For
Retailers naturally focus on preventing fraudulent transactions. But there’s another expensive problem: false declines. This happens when a legitimate transaction is incorrectly identified as fraudulent and blocked.
Industry estimates suggest false declines can cost several times more than actual fraud losses. A rules-based system designed to be extremely cautious can therefore create a different kind of revenue problem: turning good customers away. AI can help by looking at more signals and distinguishing genuinely risky behavior from legitimate transactions that simply look unusual.
Dynamic Pricing: Adopted Slowly, Despite a Clear Case
Dynamic pricing is another area where the commercial case is fairly straightforward. Yet fewer than 15% of retailers currently use AI-driven pricing, despite potential margin improvements of 5–10% and commonly cited payback periods of 6–12 months, according to McKinsey and industry analysis.
The challenge isn’t necessarily the technology. Pricing decisions affect merchandising, finance and customer trust at the same time, making them harder to implement than a more contained backend automation.
This is one reason why adoption doesn’t always follow the strength of the business case.
Why is Scaling AI So Difficult?
The 89% versus 7% gap starts to make more sense when you look at how AI is actually deployed.
1. AI is often deployed one function at a time
Marketing gets one AI system. Customer service gets another. Fraud gets another. Without a shared customer data foundation, each system ends up working from a partial picture.
2. Customer experiences become disconnected
A recommendation engine may know what a shopper is browsing. Customer service may know about their recent complaint. Neither may know what the other system knows. The result can be a personalized offer that makes little sense in the context of the customer’s most recent support interaction.
3. Success gets measured by deployment, not business impact
It’s easy to report that an AI tool has been launched. It’s harder and much more useful to show that it improved:
- Conversion
- Retention
- Customer value
- Margins
- Revenue
That’s where many AI initiatives struggle to move from experimentation to scale.
How Worktual Supports AI in Ecommerce
This is where Worktual‘s approach is different. Worktual’s Cognitive CDP brings browsing, purchase and support history into one continuously updated customer profile. Customer Value Management (CVM) then builds on that shared profile to score signals such as purchase intent and churn risk and turn those signals into actions such as a personalized recommendation, cart-recovery offer or retention outreach.
The timing matters too. Instead of relying only on fixed campaign schedules, the action can be triggered based on what the customer is doing and what the system understands about them at that moment.
The same principle extends to customer service. Worktual’s AI Contact Centre, powered by Lola, handles order status, delivery and return queries across chat and voice, using the same unified customer profile.
So if a shopper receives a personalized offer and later contacts support about another order, the support interaction doesn’t have to start from zero.
Campaign Management also works from the same shared data, allowing promotions to reflect what the platform actually knows about the shopper rather than treating them as just another generic segment.
Conclusion - The Goal Isn't More AI; It's Better-Connected AI
The retailers capturing the most value from ecommerce AI aren’t necessarily the ones with the largest number of AI tools. They’re the ones where those capabilities work from a shared, current understanding of the customer.
Personalization, customer service, fraud detection and other AI capabilities can each solve a specific problem. But when they operate on disconnected data, they can still produce disconnected customer experiences.
The bigger opportunity is to connect them. That is what turns AI from a collection of individual tools into part of the way an ecommerce business actually operates. The gap between the 89% of retailers that have adopted AI and the 7% that have scaled it isn’t just a technology problem. It’s an architecture problem.
Frequently Asked Questions
1. How widely has AI been adopted in ecommerce?
According to McKinsey and Stord, 89% of retailers have adopted AI in some form, but only 7% have scaled it across their operations.
2. What are the main uses of AI in ecommerce?
Common applications include personalization and product recommendations, customer service, fraud detection, dynamic pricing and demand forecasting.
3. What is the most proven AI use case in ecommerce?
Personalization and product recommendations have some of the strongest evidence of commercial impact. McKinsey found that faster-growing companies generate 40% more revenue from personalization than slower-growing peers, while Salesforce measured $60 billion in AI-influenced online sales during Cyber Week 2024.
4. How can AI improve ecommerce customer service?
AI can resolve high-volume queries such as order status, delivery and returns immediately and at any hour. The value increases when customer service uses the same customer data as marketing and personalization rather than operating as a separate system.
5. Why are retailers struggling to scale AI?
Many retailers deploy AI separately across functions without a shared customer data foundation. This can leave personalization, customer service, fraud and other systems working from different views of the same customer.
6. How does Worktual use AI in ecommerce?
Worktual’s Cognitive CDP creates a unified customer profile from browsing, purchase and support data. CVM uses that profile to identify intent and churn risk and trigger actions, while the AI Contact Centre, powered by Lola, uses the same customer information to handle support interactions.
Related Posts
AI-Native vs AI-Powered SaaS: What Is the Difference?
Insurance organizations operate in a complex, regulation-driven, margin-sensitive environment where growth depends on efficient onboarding, strong retention, and effective cross-sell across product lines. AI in insurance is central to this shift, as customers expect seamless digital onboarding journeys, real-time responses, and personalized communication across channels. Whether purchasing a policy, renewing coverage, or managing claims, expectations are shaped by digital-first experiences.
Why the Contact Center Is Becoming a Revenue Engine, Not a Cost Center
Healthcare providers are managing an increasingly complex environment where operational efficiency, patient access, and revenue performance are tightly interconnected. Rising patient expectations, increasing administrative burdens, and workforce shortages continue to place pressure on both clinical and non-clinical teams. Patients now expect immediate responses, seamless scheduling, and consistent communication across channels, yet many healthcare systems still rely on fragmented processes and delayed engagement. These gaps result in missed appointments, underutilized capacity, and inconsistent patient care experiences that directly impact both outcomes and financial performance.
Enterprises Don’t Need a Bigger AI Model. They Need Better Customer Context.
Insurance organizations operate in a complex, regulation-driven, margin-sensitive environment where growth depends on efficient onboarding, strong retention, and effective cross-sell across product lines. AI in insurance is central to this shift, as customers expect seamless digital onboarding journeys, real-time responses, and personalized communication across channels. Whether purchasing a policy, renewing coverage, or managing claims, expectations are shaped by digital-first experiences.
