Retailers often expect AI to deliver rapid improvements in customer service, but many find that initiatives lose momentum before delivering value.
With many retail companies wanting to use AI in their CX, they encounter gaps in core foundations rather than bad technology, leading to fragmented pilots, unclear impact, and wasted spend.
Most companies don’t know what the right foundation is supposed to be.
As a result, AI in retail CX fails to scale when layered on top of disorganized systems, meaning retailers that skip the readiness work continue to hold back so many implementations.
Hideki Hashimura, CRM/CX Strategist at redk, argues that retailers need to check whether the correct foundations are in place to make a pilot succeed today and scale tomorrow.
“Are we creating the right foundations? That’s the question,” he said.
“Are we creating the right foundations to make sure that the technology delivers on its objectives, but also that it will support expansion and over the midterm will deliver transformational impact?”
The Quick Win Trap: Why Retail AI Keeps Stalling
Retail operates under constant pressure, seeing demands spiking overnight, undersupply issues, or social media attention.
On top of this, customer expectations are rising across all channels and time zones, forcing service teams to prioritize speed and fix immediate problems over long-term stability.
A 2024 Gartner report reveals that customer expectations during service interactions continue to rise, resulting in many service leaders deploying quick-fix solutions in customer journey analytics, self-service, and AI-supported operations.
Whilst problems are temporarily resolved, they are often implemented in isolation, creating problems that deliver short-term wins but hinder long-term successes.
“The question we generally hear from brands is: ‘What is the best chatbot out there?’” Hashimura continued.
“That question shows that the approach is not right. The right question orgs should be asking is: ‘Is our foundational framework able to maximise the use of chatbots and LLM technology? ’”
Starting with the chatbot question creates the quick win trap, by falling into a repeatable cycle in which every problem triggers another short-term fix, these eventually stop being solutions and become dependencies.
The Three Foundations Retail Teams Underestimate
Retail AI readiness depends less on the tools used and more on the three core foundations that are often left underdeveloped:
- Knowledge management
- Data integration
- Omnichannel consistency
Knowledge management
This foundation is how an organization captures and maintains what it knows, serving as the information layer that systems rely on to respond accurately.
Being the least visible of the foundations, LLMs depend entirely on structured, accessible information to generate useful responses.
If this knowledge is inconsistent across teams, AI systems will likely produce unreliable answers or fall back to generic responses.
“Knowledge management is becoming a foundational piece to leverage the LLMs,” explained Hashimura.
“An LLM works on natural language and uses natural language processing to deliver its value. If you don’t have content, it serves no purpose.”
Strong knowledge management also requires ongoing ownership of a maintained system that reflects how the business operates.
Data integration
This refers to how well customer and operational data are connected across systems, such as purchase behavior, service interactions, and customer feedback.
An incomplete customer view means organizations are working with fragmented signals rather than a full picture.
When data lives in separate environments, teams cannot build a complete view of the journey, limiting personalization and making it difficult to understand the root service causes.
“Nine out of ten times, the impact is zero, or nine out of ten times, they can’t even measure the impact,” said Hashimura.
“You don’t even have a benchmark, or if you do, it’s very disconnected from the rest of the ecosystem.”
Without a connected data layer, AI outputs cannot be properly evaluated, meaning performance improvements are less often proven.
Omnichannel consistency
CX is unified across different touchpoints, defining how customers experience AI in practice.
Customers frequently move between channels and expect continuity at every step, meaning retailers require a cross-functional design that connects service, back office, finance, and product into a single resolution flow.
If an AI is deployed to only one channel, but the rest of the system operates differently, the experience feels fragmented rather than seamless.
A big part of maintaining omnichannel consistency has to do with having the mechanisms to orchestrate and align different moving parts in service operations, the right foundation includes having the technology that aligns AI-driven channels with processes in order to enable automation.
Consistency requires shared processes, aligned messaging, and coordinated workflows across teams, and depends on how the business is connected when resolving customer issues.
Building the Blueprint: Governance, Skills, and What Leaders Should Ask Now
The difference between AI pilots that stall and those that scale comes down to governance, structure, and whether there is a clear strategy guiding the introduction of AI into the business.
“If you don’t have a strategy, you’re going to be wasting a lot of time,” explained Hashimura.
“It’s like building a house without architectural plans. You have to have a vision from the beginning of where you want to get to.”
Practical governance defines a clear and achievable direction for how AI will be leveraged across the organization over time so teams can work with it on a day-to-day basis.
Retail brands need to give AI a defined target state, even if this definition evolves over time, with each initiative contributing to a continuous improvement of the operating model rather than existing as a standalone, isolated improvement.
Within this structure, AI’s role is to handle execution at scale, whilst end-to-end customer resolution orchestration still requires human oversight.
“The role of the human is changing in the service team, but they still play a very, very important role,” Hashimura continued.
This shift requires new ways of working inside service teams, where employees are more focused on optimizing across systems, and AI cannot be treated as a standalone technology investment.
The leadership needs to have a vision on how to upskill its service teams into AI service architects, roles who understand the nuances of the business so they can orchestrate AI activity.
As a result, impactful outcomes will depend on how well the business process is designed around the technology, meaning retailers must define clear objectives and performance measures from the start.
The value of AI lies in more structural change, shifting the focus from efficiency to redesigning how service is delivered across the customer journey.
“You don’t implement AI to get a 5% improvement. You can probably get that just by tweaking your existing team and your systems,” explained Hashimura.
“You want something transformational – not only speed in terms of KPIs, but an amazing customer experience that is fluid, streamlined, effective, accurate, and fast.”
The key readiness question is not whether AI is being used, but whether the organization has the structure to support it at scale.
