AI in customer experience (CX) has now reached an inflection point. For years, the conversation has been dominated by bold promises; fully automated contact centres, human‑free service models, and instant cost savings.
However, most CX leaders know that buying AI for the sake of it rarely delivers any meaningful results. The organisations seeing real value from AI aren’t starting with technology. They’re starting with outcomes.
Rather than asking ‘Where can we add AI?’, they’re asking:
- Where is time being wasted?
- Where are agents overloaded with work customers never see?
- Where does friction exist that doesn’t improve the experience at all?
This outcomes‑first mindset is becoming essential. According to Gartner, 80% of customer service organisations are expected to use generative AI to improve CX and agent productivity by the end of 2025. Yet only those that align AI to clear operational goals are likely to see sustained ROI.
The real problem in CX today
CX teams are under intense pressure from multiple directions at once. They need to contend with:
- Rising customer expectations for speed, accuracy, and availability.
- Increasing operational costs, with labour accounting for 60–70% of total contact centre expenditure.
- Unpredictable spikes in demand, caused by system issues, external events, or internal mistakes.
At the same time, a significant portion of agent effort is spent on work that adds little or no direct value to customers.
Industry data shows that agents spend between 6% and 12% of their working time on after‑call work (ACW). That’s the work that involves logging notes, completing forms, or updating systems after an interaction ends.
Across large contact centres, this quickly adds up to the equivalent of multiple full‑time roles doing administrative work customers never see. Again, no value added.
This is what AWS often refers to as ‘undifferentiated heavy lifting’. It’s necessary work that doesn’t differentiate your brand or improve the overall experience.
A smarter way to apply AI: Start with outcomes
Many AI initiatives fail because they start with a tool rather than a problem. Organisations sign multi‑year contracts with platforms that promise transformation, only to find themselves locked into solutions that don’t evolve at the pace of the technology. The draw of shiny new tools can cloud judgement.
A more sustainable approach flips the model:
You shouldn’t be asking ‘What can this AI tool do?’. Better, high-performing CX teams will be the ones asking:
- What outcome are we trying to achieve?
- Where is the most time or value being lost today?
- How can AI remove friction without removing critical human judgement?
By adopting this approach, teams can avoid vendor lock-in. AI remains aligned to real operational needs as opposed to novelty.
Why the human still matters
Let’s break it down. AI is best at:
- Speed.
- Pattern recognition.
- Consistency.
Humans excel at:
- Empathy.
- Context.
- Accountability.
When it comes to customer service, 79% of consumers are in favour of interacting with humans rather than AI. Fast resolution is one thing, but nothing beats the human touch.
The winning model is human‑in‑the‑loop AI. Technology removes friction, humans deliver trust. That’s the goal.
CX as a differentiator (not just a cost centre)
CX is increasingly defining brand value. Customers are willing to pay up to 16% more for a superior experience, and 86% of buyers will spend more for better CX.
At scale, AI enables:
- Always‑on service.
- Multilingual support.
- Personalised engagement.
When relating this to multilingual support, automation becomes critical. Research shows that 76% of consumers will opt to buy from brands that provide information in their native language. If that language option isn’t available, then 40% won’t do it.
What does this mean for CX leaders?
When used correctly, AI will reshape how agents spend their time. If AI is replacing entire CX teams, then leaders are creating problems rather than solving them.
What are the key questions CX leaders need to be asking? Those are:
- Where is your team spending time that customers never see?
- How quickly can you identify emerging issues?
- Are your agents supported, or overloaded, by systems?
AI-minded organisations are quietly winning when they aren’t chasing hype. They’re the ones removing friction and protecting the human touch.
The most successful CX strategies are grounded in outcomes. It should never come down to tools.
