How SF City Limo Cut Customer Acquisition Costs 43% and Rebuilt Its Approach to Fleet Operations, Without Exposing Client Data to AI

At a glance

• Company: SF City Tours, Inc., operating as SF City Limo (LinkedIn)
• Industry: Premium ground transportation, sedan, SUV, Sprinter van, minibus, and motor coach service, national and international, through a vetted affiliate network (see the fleet)
• Teams using Nera: Marketing and client services
• Plan: Nera ChatApp, Business tier
• Headline result: 43% lower cost per acquisition, conversion rate more than doubled, in 6 months

The problem

SF City Limo serves a discreet, high-value client base, corporate accounts, repeat luxury travelers, clients who book through referral and loyalty relationships built over years. Understanding what actually drives that business, which campaigns convert, which clients are worth the most investment, where the fleet is underperforming, requires digging into exactly the data the company is most protective of: client identities, spend history, referral patterns, and internal operations.

"I wanted to use AI software with peace of mind, without worrying about exposing sensitive business or client data," said SF City Limo's CEO/COO.

That tension, wanting the analytical power of modern AI tools without the exposure risk, is what brought SF City Limo to Nera.

Marketing and growth: from guessing to knowing

SF City Limo's marketing and client services teams used Nera to analyze campaign performance across the full range of the business: corporate versus retail channels, bus versus sedan and SUV service, tours versus airport transfer bookings. They tracked cost per acquisition, conversion rates, referral percentage by client, and average spend, then compared how individual clients and accounts performed against each other and against their own history over time.

Over a six-month period, the results on the company's Sprinter and bus campaigns were direct and measurable:

• Cost per acquisition dropped from $102.67 to $58.72, a 43% reduction
• Conversion rate rose from 7.82% to 18.3%, more than doubling

But the deeper change was in how decisions got made. "It changed my investment pattern," the CEO said. "I allocated budget more efficiently between marketing, referral discounts, promo codes, and credits to loyal and high-ticket clients, and responded faster to clients coming through specific channels."

Asked what single insight captured the shift, the CEO didn't point to a dashboard, he pointed to a question he could finally answer directly: "It helped me understand where my investment matters more, where I should be allocating more budget, and which clients are most valuable to the business. It helped me scale at a faster rate."

Fleet and workforce: the operational layer underneath the numbers

Marketing wasn't the only place Nera got used. SF City Limo's team also turned it toward day-to-day fleet operations, the kind of workforce data that rarely gets analyzed systematically because pulling it together by hand takes too long to be worth it. That included driver call-out rates, how often drivers responded to wake-up call notifications on time, and patterns in driver lateness across the fleet.

Where marketing analysis answered "where should the next dollar go," the operational analysis answered a quieter but equally important question: where is the business exposed to service failures before they happen. Both required the same underlying guarantee, real analysis on real operational and client data, without that data ever being exposed to the AI model doing the analysis.

Getting started

"The onboarding process was simple," the CEO said. "Once connected, I was able to chat and get strong analysis that helped me make informed, critical business decisions, decisions that helped me grow and scale the business."

Asked what he'd tell a peer business serving a similarly discreet clientele considering the same move, the answer was direct: use a tool that lets you run AI with peace of mind, without worrying about exposing sensitive business or client data.

Where this goes next

The same approach that answered "where should the next dollar go" and "where is service at risk" extends naturally to questions SF City Limo hasn't fully explored yet: seasonal demand forecasting across service lines, deeper driver-safety and incident-pattern analysis, and expanding the same client-value and referral analysis to new markets as the company grows its affiliate network. None of it requires a different guarantee, only pointing the same architecture at the next question worth answering.

"This is exactly the outcome we built Nera for," said Rami Akeela, Ph.D., Founder and CEO of Nera Systems. "SF City Limo didn't need a slower, more careful version of AI. They needed the real thing, applied to the data that actually runs their business, without having to choose between capability and confidentiality."

Nera Systems builds confidential AI infrastructure for regulated and data-sensitive businesses. Client and business data is encrypted before it ever leaves your control, the AI model receives the question and the shape of the data, never the underlying values. Learn more or try it yourself.