Your brand is on the road. Who’s driving?

I have a newish routine this year. A few days a week, I drive my daughter to school before heading to the office. It’s a hit-and-miss part of the day, a pocket of uninterrupted time together before the world takes over or infuriating chaos. 

The latter usually stems from witnessing some truly terrible driving. Tailgating. Running reds. Aggressive lane changes that make your heart rate spike. The ones that surprise me most are the cars that carry company branding. Phone numbers, logos, taglines, the works. Businesses I might have called one day. Some I’d been meaning to look into, but perhaps not anymore.

When I watch a branded vehicle cut someone off or blast through a school zone, the logo is the first thing I see, and just like that, the company’s reputation is shaped by a stranger behind the wheel.

This got me wondering how big an impact this really was, and it turns out that a branded vehicle generates between 30,000 and 70,000 daily impressions. WHAT! I need to rethink my LinkedIn impression goals. That's reach most paid digital campaigns would envy, so I wondered why every company wouldn’t do this. Well, it turns out that visibility cuts both ways. When the driver behind your logo behaves badly, every one of those impressions works against you.

I dug up some stats, (the ones that add the most drama to this article) and here’s what I found: 

I didn’t need to look too far either to quantify the reputational fallout. Thanks to dashcams and social media, bad driving in a branded vehicle becomes a PR nightmare pretty quickly. It’s also a deep rabbit hole I got stuck in on YouTube. 

Turns out when you are not behind the wheel, watching bad drivers is less infuriating and actually mildly entertaining.

‍

Of course, the car is just a metaphor

‍

Like always, all this got me thinking about this from a CX perspective, in particular, how AI and many of an organization’s entry points for customers are also vehicles in a way. Branded ones. Carrying their logo into the world, millions of times a day, without a human in sight.

A branded vehicle is just one example of a non-human entity carrying your brand into the world. Your IVR, your chatbot, your Voice AI agent, your automated email sequences, they’re all vehicles too. They carry your brand into millions of customer interactions, often as the first touchpoint someone has with your organization.

A poorly performing AI agent can destroy trust maybe as fast as a reckless driver behind the wheel of your branded van.

Forbes posted a stat that 96% of consumers say bad customer service affects their loyalty to a brand and the 2025 Zendesk CX Trends Report stated that more than half will switch to a competitor after a single bad experience So, when AI chatbots hallucinate between 3% and 27% of the time, confidently giving wrong answers to your customers, while wearing your logo, I think there is a need to pay attention.

When your AI runs a red light

We’ve already seen what happens when AI “drives badly” in public. An airline chatbot fabricating a bereavement fare policy. A delivery company's customer service bot swearing at a customer and writing a poem calling the company “the worst delivery firm in the world”, which went viral with 1.3 million views. A car dealership’s chatbot tricked into offering a $70,000 truck for $1.

Right now, these might be edge cases, sure, but they should serve as a warning. They’re the branded van running the red light, except the dashcam footage reaches millions, instantly. However, unlike a bad driver, a bad AI doesn’t make one mistake and pull over. It keeps going, at scale, around the clock.

One look at the stock market and you can see that the investment in AI is enormous. Without proper oversight, much of it goes to waste. Or worse, it actively harms the brand it was meant to serve.

You wouldn’t skip the dashcam. Don’t skip CX observability.

Many fleet managers solved this problem years ago. They installed telematics, GPS tracking, and dash cams. They monitor driving behavior in real time, set alerts for dangerous patterns, and intervene before a bad driver becomes a liability. They don’t wait for the accident or the complaint call from a fellow road user, no they watch for the warning signs.

AI-powered customer interactions need the same discipline. That’s where CX Observability comes in. It can see inside every layer of your customer experience stack, in real time, and understand what’s happening in each interaction.

The most dangerous failures in AI systems are the silent ones. Your monitoring dashboard shows green across the board while the model delivers bad answers, awkward silences, or policy violations to real customers. High-level metrics give you a superficial surface view. They’re lagging, not leading and they mask what’s really happening.

CX Observability changes this. It’s continuous and granular. It captures every interaction in real time, along with all the supporting voice, network, and telephony data. It tells you why something went wrong, where, and what to do about it.

Forrester’s analysis shows a 357% ROI from AI observability implementation over three years, with a payback period of under six months. So my question is: Why would you go without it?

Who’s behind the wheel of your customer experience?

Companies spend millions building their brand. They obsess over every pixel on the website, every word in the ad campaign, every detail of the in-store experience. Then they hand the keys to an unmonitored AI and let it interact with thousands of customers a day.

That’s the equivalent of spending a fortune on vehicle branding and then never bothering to check who’s driving, or their driving record, whether they have a licence or whether they have the capacity to drive.

Every non-human entry point in your customer experience is a branded vehicle on the road. Your Voice AI, your chatbot, your IVR, your automated workflows, they’re all out there, carrying your logo, shaping perceptions, building or breaking trust with every interaction.

CX Observability is the essential infrastructure and safety net protecting your brand reputation. The dashcam that catches problems before they go viral.

You can’t fix what you can’t see but you can bet that your customers are watching.

Until next time and as always, hooroo.
‍

Need a purpose-built dashcam for cloud contact center and AI?

See how Operata works 

‍

References

1. Grit Daily, "Is Your Brand on Your Vehicle Part of Your PR?" — https://gritdaily.com/is-your-brand-on-your-vehicle-part-of-your-pr/
2. Federal Motor Carrier Safety Administration (FMCSA), Large Truck Crash Causation Study.
3. SambaSafety, "Reasons for Fleet Vehicle Ownership" — https://sambasafety.com/blog/reasons-for-fleet-ownership/
4. NHTSA, Aggressive Driving Research Update (2009). Aggressive driving contributed to 56% of fatal crashes, 2003–2007.
5. Pine AI, "Customer Service Statistics" — https://www.19pine.ai/blog/customer-service-statistics/
6. Zendesk CX Trends Report (2025). Over 50% of consumers switch after one bad experience.
7. Qualifire, "When AI Gets It Wrong" — https://www.qualifire.ai/posts/when-ai-gets-it-wrong
8. CMSWire, "The Rise and Fall of Air Canada’s AI Chatbot" — https://www.cmswire.com/customer-experience/exploring-air-canadas-ai-chatbot-dilemma/
9. CX Today, "3 Times Customer Chatbots Went Rogue" — https://www.cxtoday.com/contact-center/3-times-customer-chatbots-went-rogue-and-the-lessons-we-need-to-learn/
10. Evidently AI, "LLM Hallucination Examples" — https://www.evidentlyai.com/blog/llm-hallucination-examples
11. MIT Sloan / MIT CSAIL research on AI pilot ROI (2024). 95% of pilots delivered zero measurable return.
12. Gartner, "Generative AI Project Forecast" (2024). 30% of GenAI projects abandoned post-POC by end of 2025.
13. Forrester, "AI Observability ROI Analysis" (via IBM) — https://www.ibm.com/think/insights/ai-agent-observability
14. TheTEIOfMonteCarlosDataAIObservabilityPlatform.pdf
    https://tei.forrester.com/go/montecarlo/dataaiobservabilityplatform/docs/TheTEIOfMonteCarlosDataAIObservabilityPlatform.pdf#:~:text=Monte%20Carlo%20commissioned%20Forrester%20Consulting%20to%20conduct,investment%20(ROI)%20358%25%20Net%20present%20value%20(NPV)