Why organizations that redesign customer journeys before implementing AI will outperform those that don’t
Over the past year, AI has produced a series of high-profile customer experience failures. Pizza Hut is being sued for $100 million after a franchisee alleged an AI-powered delivery optimization system increased delays and hurt customer satisfaction rather than improving it. Air Canada was held responsible after its automated AI agent provided customers with incorrect information about bereavement fares. In both cases, the technology performed as it had been deployed, but there were critical failures in how AI was designed into the customer experience.
For decades, organizations have chased technology platforms as an answer to growth and efficiency. The pattern is familiar: a compelling capability emerges, leaders move fast to adopt it, and the technology gets layered onto journeys and workflows that were never redesigned to support it. AI magnifies this problem. Unlike many technologies, AI agents don’t sit inside one system or one team, they influence the entire journey, from how customers discover and buy to how they receive service and remain loyal. Every AI interaction is now a customer experience decision, and every customer experience decision has commercial consequences.
The critical question all business leaders must ask is where AI belongs and where AI doesn’t belong. The quality of those design decisions will determine whether AI becomes a growth engine, an expensive operational burden, or the catalyst for enterprise and brand value destruction.
Don’t Create Worse Customer Experiences Using AI
There is extremely high pressure to implement AI agents, in fact, more than 90% of servicing leaders report executive urgency to implement AI. At the same time, the margin for error has never been lower – it’s never been easier to switch brands, and 80% of customers said they have switched brands because of poor customer experiences. This equation should be scary for business leaders. Most leaders are responding by moving fast, piloting agents across customer touchpoints, and hoping efficiency gains will follow. This is the same pattern organizations experienced chasing technology to answer growth and efficiency challenges.
When AI is introduced into journeys that were already fragmented, it accelerates the fragmentation. Customers encounter inconsistent answers across channels, context disappears between digital and human interactions, employees hesitate to act on recommendations that conflict with their experience, and escalations increase even as automation expands.
The business impact can quietly compound, resolution times slowing due to rework, cost-to-serve increasing as people compensate for poorly designed automation with more outreach, or dramatically increase, if customer trust is eroded and customers decide to leave en masse. Ultimately, customers will leave if their experiences are poor.
The risk lives in the deployment of technology. How, where, and why AI gets deployed are the key questions leaders need to answer, and this is only answered with a thoughtful and pragmatic AI design strategy.
Four Design Decisions Leaders Need to Make
The organizations positioned to win with AI are moving deliberately. Before deploying agents, they are doing something that sounds simple but requires real discipline: quantifying which experiences justify AI investment and then designing how AI, people, data, and workflows should work together to deliver experiences customers will trust.
This deliberate action is what we call Agentic by Design. It is not a maturity model or a deployment checklist, Agentic By Design is a strategic operating lens built around four components that must be addressed together before AI agents are integrated into customer experiences: decision design, workflow orchestration, human and machine roles, and value realization.
- Decision design. Which decisions shape the customer experience and who should make them? Not every decision should be automated. The organizations creating the most value with AI deliberately define where AI acts, where people lead, and where human judgment remains essential.
- Workflow orchestration. How does work move across people, systems, and AI without breaking the experience? It’s an old customer experience adage – customers don’t experience functions; they experience journeys. The biggest failures in agentic AI aren’t the models themselves, but the handoffs where context, accountability, and continuity are lost.
- Human and machine roles. When should AI act autonomously, when should it assist, and when should people step in? Clear role definition builds trust, improves adoption, and ensures AI augments employees rather than creating new operational friction.
- Value realization. Which experience improvements justify the investment? AI should be evaluated like any strategic investment with a clear connection between customer experience, operational performance, and business outcomes. Without that connection, organizations struggle to prioritize investments, prove ROI, or scale what works.
These four components are interdependent. When addressed together, they create the conditions for AI to accelerate value; when addressed in isolation, or not at all, they are where AI investments stall or cause public-facing brand and customer value crises.
The Best Time To Design Is Before You Build
Not every organization starts in the same place, and not every customer journey needs autonomy. Some organizations need to reduce operational friction before introducing AI, others are ready to orchestrate work across people, systems, and AI. A smaller group is prepared to deliver proactive, highly personalized, highly autonomous experiences. The objective isn’t to climb a maturity ladder, but rather, to start where customer experience, trust, and economics intersect most sharply.
Virgin Voyages took this approach. Rather than deploying a chatbot to reduce call volume, the team first identified the service interactions creating the highest customer effort and cost-to-serve. They discovered AI was creating work rather than reducing it. By redesigning AI-to-human handoffs, clarifying decision ownership, and defining where human judgment remained essential, the organization reduced call center escalations by more than 20% at launch.
Customer-facing AI is fundamentally a design challenge. Most organizations can deploy technology, but far fewer can determine where AI belongs, how work should change around it, and how value will be created because of it.
Agentic By Design is centrally focused on intent. Intent, not the model, the budget, or the number of agents in production, is the differentiator. Organizations that design deliberately before they deploy will create compounding value. Those that don’t will spend years unwinding the cost of moving too quickly.
