Detaching From AI to Build Meaningful Customer Relationships: How TFL Is Reframing CX
At the Call and Contact Centre Expo 2025, AI took center stage. Transport for London offered a different playbook. Fola Olafare, Senior Contact Center Delivery Manager, made a simple point: build trust first, then use automation where it actually helps. Less chasing hype, more serving people.
Less AI, More Human
For years, TFL put unnecessary distance between customers and help. "We had 15 different phone numbers and 35 different emails." That's friction no one needs.
They stripped it back. "Telephone is a great starting ground for us - now we have one phone number for customers." The focus is clear: listen, respond, and make it easy to get support. That includes nuanced help for visitors and dedicated aftercare for people impacted by serious incidents on the network.
Simplify the Front Door
TFL's move from dozens of contacts to one clear entry point is the kind of decision that compounds. Fewer options, faster routing, better outcomes. Their ethos: "listen to our customers and demonstrate that we care." You don't need a fleet of bots to do that-you need clarity and accountability.
Results That Matter
The model shows up in the data. TFL supported roughly 30 million daily trips across the network in 2024, ran 9,000 buses, operated 1,400+ e-bikes, and managed a quarter of London's roads-while improving experiences for residents and visitors. The direction is set: get more people confidently using public transport and keep the experience human at the moments that count.
Want to see how they present help and travel information publicly? Start here: Transport for London.
Automation Still Has a Seat at the Table
TFL hasn't ditched automation. They've just put it in the right place. Olafare noted they run strong algorithms in the background to detect anomalies and surface alternative route messages when something goes wrong.
Here's the key: they expect automation to fail sometimes. That's why they invest in people who can jump in quickly, own the issue, and resolve it with care. Automate the repetitive. Humanize the exceptions.
How CX Leaders Can Apply This Now
- Consolidate contact points. One number. One help center. No guessing.
- Define "moments that matter." Route high-stakes or sensitive cases to trained humans fast.
- Measure conversation quality, not just handle time. Empathy and resolution beat speed alone.
- Use automation behind the scenes first. Detect anomalies, prefill context, and reduce agent toil.
- Design for aftercare. Some events require follow-up over days or weeks-own that responsibility.
- Publish your values. Be open and transparent about how you support people and what you're improving.
The Principle Underneath
AI can't fix a messy experience. Simplification does. TFL's shift is a reminder: make it easy to reach you, listen deeply, and build systems that back your agents when it matters most.
Need to Level Up Automation Without Losing the Human
If you're defining what to automate and what to keep human, this resource helps teams make smarter trade-offs: Practical AI automation ideas.
Your membership also unlocks:
