How does Saks use AI agents in customer service and personal shopping?
Saks set out a plan in 2024 to let AI agents handle routine service requests and to give advisors AI recommendations from a unified customer profile. Here is what was announced, what is known and what is not.
KEY TAKEAWAYS Summary by the editors
- In September 2024 Saks expanded its Salesforce partnership to add Data Cloud and Agentforce, with an autonomous service agent planned for simple tasks such as updating shipping addresses.
- Salesforce's Saks customer story says AI agents are intended to handle routine requests such as order status and returns so that human service staff can focus on personalised service.
- Saks planned for AI agents to use a unified customer profile to suggest recommendations and actions to customer representatives and, eventually, to style advisors.
- Saks Global executives framed the agents as augmenting staff; then CEO Marc Metrick said Agentforce would free up people to work differently with clients.
- Saks Global emerged from chapter 11 as Exemplar Luxury Group on 26 June 2026, and neither company has published measured results for its AI agents.
Saks has positioned AI agents as a way to take routine work off its people, not to replace its personal shopping model. Its 2024 plan, built with Salesforce, was for an autonomous agent to handle simple service tasks such as updating a shipping address or checking an order, while a unified customer profile would let AI suggest recommendations to service representatives and, later, to style advisors. Public detail on outcomes is limited, and the group has since been through a major restructuring.
What did Saks announce about AI agents?
On 18 September 2024, Saks Fifth Avenue expanded its partnership with Salesforce, according to Retail Insight Network. The retailer was deploying the Salesforce Platform including Customer 360 applications, with Data Cloud and Agentforce, Salesforce's AI agent platform, to follow. Commerce Cloud would support sales tracking, order processing, inventory management and fulfilment, and Slack AI was already being used to summarise threads and answer internal questions.
Two agent use cases were described:
- Customer service: customers would be able to use an autonomous Agentforce Service Agent for simple tasks such as updating shipping addresses, freeing human agents for more complex inquiries.
- Advisor support: Data Cloud would combine data from different systems into a unified customer profile, and agents would use it to offer AI-driven recommendations and actions to customer representatives and, eventually, to style advisors, based on the customer's history across sales, service and commerce.
Salesforce's own customer story on Saks, last updated in August 2025, adds order status and returns as examples of routine requests the agents are meant to handle, and describes the use of first-party data and AI to personalise shopping across digital, mobile and in-store channels. It describes these as plans and does not report measured outcomes.
How does Saks describe the role of human advisors?
Saks Global's executives have been consistent that agents support staff. Then CEO Marc Metrick said: "Agentforce will free up our people to work in a different way with our clients," adding that it would augment rather than replace them. Chief technology officer Mike Hite said: "Agentforce is really allowing our humans to have a deeper, more impactful interaction with our customers," and, in an interview with Retail Insider, that it would "improve our effectiveness across customer touchpoints, empowering our employees".
This framing reflects how luxury department stores make money. Personal shoppers and style advisors build relationships with high-spending clients over years. An agent that answers "where is my order" at any hour protects the advisor's time; an agent that tried to replace the advisor would put the relationship at risk.
| Task | Who handles it in Saks's stated plan | Data required |
|---|---|---|
| Order status, returns, address changes | Autonomous service agent | Order, shipping and returns systems |
| Complex service inquiries | Human service agents | Full case and order history |
| Product and outreach suggestions | Agent suggests, representative or advisor decides | Unified customer profile across sales, service and commerce |
| Client relationships and styling | Style advisors | Purchase history, preferences, appointments |
What data do AI agents need for personal shopping?
The Saks plan makes the dependency explicit: agent recommendations rely on a unified profile combining sales, service and commerce history. In practice that means resolving one client's identity across online accounts, store purchases and service contacts; holding consent for how that data is used; and keeping product and inventory data current so that suggestions are available to buy. Without this, an advisor-facing agent produces generic suggestions that experienced advisors will ignore.
There are also operational risks that any retailer copying this model should plan for. A service agent that misstates a returns policy or a delivery date creates a complaint rather than resolving one, so answers must come from live systems, not from a model's general knowledge. Escalation to a person needs to be fast and to carry the conversation history with it, so the client does not have to repeat themselves. And in luxury, tone matters: an agent's replies should be reviewed against the house's service standards before launch and monitored afterwards, because a curt or generic answer can undo years of relationship building by an advisor.
How did the restructuring of Saks Global affect the picture?
The AI plans were set out before a period of severe financial pressure. Saks Global, which brings together Saks Fifth Avenue, Neiman Marcus and Bergdorf Goodman, went through chapter 11 proceedings in the first half of 2026. On 26 June 2026 it announced that it had emerged under new ownership as Exemplar Luxury Group, citing a reduction in debt of nearly 75%, with Geoffroy van Raemdonck as chief executive. The emergence release speaks of "harnessing rich customer insights to curate distinct product assortments and deliver personalized experiences" and refers to remote selling services, but does not mention AI or specific platforms.
For observers, this is a reminder that technology announcements and operating reality can diverge. Neither Saks nor Salesforce has published figures on how many interactions the agents handle, resolution rates or effects on sales, and the leadership that announced the programme has changed.
What can retailers learn from the Saks approach?
- Start agents on bounded tasks such as order status, returns and address changes, where answers come from systems of record.
- Keep a clear handover to human staff for complex or high-value interactions.
- Build the unified profile first; advisor recommendations are only as good as the joined-up data behind them.
- Frame agents for staff as tools that protect their time, and involve advisors in testing suggestions.
- Measure from day one: containment, resolution, satisfaction and advisor uptake, so the programme can survive budget scrutiny.
Frequently asked questions
Does Saks use AI for customer service?
Saks announced in September 2024 that it would use Salesforce's Agentforce Service Agent for simple tasks such as updating shipping addresses, with order status and returns cited as further examples. Neither company has published performance results.
Will AI replace personal shoppers at Saks?
Saks executives have said no. Then CEO Marc Metrick said Agentforce would augment rather than replace staff, and the plan was for agents to suggest recommendations to representatives and style advisors, who keep the client relationship.
What technology does Saks use for AI agents?
Saks expanded its partnership with Salesforce in 2024, covering Commerce Cloud, Data Cloud, Agentforce, Service Cloud and Slack AI, according to Retail Insight Network and Salesforce.
What is Exemplar Luxury Group?
Exemplar Luxury Group is the name under which Saks Global emerged from chapter 11 on 26 June 2026. It comprises Neiman Marcus, Saks Fifth Avenue and Bergdorf Goodman.
One edition every weekday morning. Read in five minutes. Free for industry professionals.
