Fisher & Paykel treats AI discovery as a marketing problem
Fisher & Paykel is connecting AI discovery, service and commerce as customers increasingly research brands through conversational interfaces.
Fisher & Paykel is starting to treat AI as more than a customer-service efficiency tool.
The appliance company says changes in how people research products and discover brands are increasingly shaping where it invests. Chief digital officer Rudi Khoury said customers are already using GPT-style tools as research interfaces, which pushes AI into a broader set of questions around marketing, commerce and growth.
“Customer behaviour is not going to change, it has changed,” Khoury said.
The important shift is not that another enterprise has deployed an AI agent. It is that the boundary between discovery, service and selling is becoming harder to maintain.
Table of contents
Jump to each section:
- Why AI discovery changes the marketing problem
- Why service agents are becoming customer-journey infrastructure
- The strategic tension: efficiency versus customer behavior
- Why brand accuracy becomes an operating issue
- What marketers should know about AI-mediated discovery
Why AI discovery changes the marketing problem
Marketers have traditionally treated discovery, service and conversion as separate stages with separate systems.
AI interfaces weaken that separation. A customer can ask a product question, compare alternatives, clarify a specification and move toward a purchase inside the same conversational environment. The marketing implication is that visibility is no longer only about earning a click.
Fisher & Paykel’s response is useful because Khoury frames the issue around customer behavior rather than around a specific AI tool. The company is using AI across its business, but he is particularly focused on what happens when customers increasingly discover brands through GPT-style systems.
That changes the job of brand information. Content is no longer only persuasive material for people. It also becomes source material that machines may interpret before a customer reaches the brand’s own site.
Adobe has described a similar shift in AI search behavior, where generative systems can become an early touchpoint in the customer journey. The common thread is not a new SEO tactic. It is the growing importance of making brand information both discoverable and reliably interpretable.
The new visibility problem is partly a knowledge problem.
If an AI system cannot find accurate product information, the marketing team may lose influence before media, creative or conversion optimization even enters the picture.
Why service agents are becoming customer-journey infrastructure
Fisher & Paykel’s most visible AI deployment today sits in customer service, where an Agentforce-based system helps customers resolve enquiries.
The company does not describe the agent as something that can simply be switched on and left alone. Employees monitor its performance, identify gaps in its knowledge and test how it responds to customer questions.
That operating model matters because the same infrastructure can extend beyond support. Salesforce Agentforce Commerce connects guided shopping, merchandising and commerce workflows around the same broader agentic model.
The more interesting question is what happens when a service system becomes part of product discovery.
A customer who asks an agent about dimensions, installation, compatibility or product differences is not necessarily seeking “support” in the traditional sense. The interaction may also be research, consideration and conversion.
The channel may stay conversational while the commercial intent changes underneath it.
That creates an organizational challenge. Marketing, ecommerce, customer service and product teams may all have a stake in the same answer, even if the interface belongs to only one department.
The strategic tension: efficiency versus customer behavior
The common assumption is that enterprise AI adoption starts with efficiency.
That is understandable. Customer service offers clear repetitive tasks, large volumes and measurable operational costs. But Fisher & Paykel’s roadmap points to a contrasting reality: the bigger strategic pressure may come from customers changing faster than internal processes.
Khoury argues that businesses should not let the availability of AI tools dictate the roadmap. The starting point should be the experience customers are moving toward.
This is a different investment logic.
Instead of asking where AI can remove the most labor, teams ask where customer behavior is creating a new interface between the brand and the market. That can lead to different priorities because the highest-value use case may not be the one with the clearest immediate cost saving.
Salesforce executive Frank Fillmann described another part of the same problem: enterprises can build interesting pilots that never become meaningful operating systems.
The strategic implication is that AI maturity should not be measured by the number of pilots or agents deployed. A more useful question is whether those systems are connected to an actual customer or business process.
Automation without a customer-journey thesis is still just automation.
Why brand accuracy becomes an operating issue
When discovery moves into AI systems, marketers lose some control over the interface while becoming more accountable for the information feeding it.
Fisher & Paykel’s own framing is direct. Khoury said brands need to make sure they are discoverable and that the information customers encounter is accurate.
That makes content governance more operational.
A product specification, service policy or feature description can influence discovery, support and conversion at the same time. If different teams maintain conflicting versions, an AI layer can expose that inconsistency more quickly because it is assembling answers across sources and systems.
BCG’s work on research-led consumer journeys similarly argues that AI and LLM research are becoming meaningful touchpoints in purchase decisions. The practical lesson is not that every brand needs a separate “AI marketing” strategy.
It is that existing information architecture now has another audience.
Machines are becoming intermediaries between brand knowledge and customer decisions. That raises the value of clear product data, consistent claims and ownership over what information is authoritative.
What marketers should know about AI-mediated discovery
Fisher & Paykel’s roadmap is useful because it treats AI as a customer-behavior shift rather than a software category.
Discovery and service are converging. A conversational interface can move between research, support and purchase intent without changing channels, which makes departmental ownership less obvious.
Brand visibility is becoming operational. Teams need to think about whether product information can be found, interpreted and kept accurate across the systems that AI assistants use.
Human oversight is not disappearing. Fisher & Paykel is monitoring agent performance and knowledge gaps, showing that autonomy still creates management work rather than eliminating it.
Tool selection should follow the journey. The company’s approach challenges the idea that AI strategy should begin with whatever capability vendors make available next. Customer behavior provides a more durable organizing principle.
For marketers, this widens the meaning of digital presence.
A website can still be the canonical destination, but it may no longer be the first place where a customer forms an understanding of the brand. AI assistants can summarize, compare and answer questions before the customer reaches an owned channel.
The brands that adapt well will need more than stronger prompts or another discovery dashboard. They will need marketing, service, commerce and product information to agree with one another closely enough that an AI intermediary can represent the business without creating a new layer of confusion.
