How Zipcar's Gen AI Campaign Drove Brand Awareness and Site Visits

The car rental firm saw a 28% increase in brand awareness and 81% uptick in brand recall

Mark your calendar for Mediaweek, October 29-30 in New York City. We’ll unpack the biggest shifts shaping the future of media—from tv to retail media to tech—and how marketers can prep to stay ahead. Register with early-bird rates before sale ends!

Artificial intelligence took the front wheel in car rental firm Zipcar’s latest campaign, saving time on audience research and driving more people to its site to use its services.

Zipcar’s second “Zip If” campaign in the U.S. aimed to boost top-of-funnel brand awareness. The campaign ran from June to September with the goal of positioning the brand as the preferred rental car choice for diverse needs, whether it be running errands or embarking on out-of-town getaways.

Generative AI was ingested through the campaign planning and execution—including brand competitive analysis, strategy development, initial brief writing and drafting headline copy—all based on audience and category data gathered from its previous campaign in 2022.

“[Gen AI] saved our marketing team months of research and analysis and helped spark ideas, discussion and direction,” Justin Holmes, head of marketing and public policy at Zipcar, said in a statement. “While gen AI saved time upfront, the time it takes to contribute internal knowledge and the human touch is irreplaceable.”

In the past 10 months, a Gartner survey of 1,400 executive leaders found that more than one-half of firms have ramped up their gen AI efforts. Although some brands struggle to fully embrace AI, others, including Coca-Cola, have incorporated it into their marketing efforts. Ultimately, this technology is proving to save time for both brands and agencies while meeting campaign goals.

AI for audience personas

Zipcar, with the help of ad agency Supernatural, created a few-dozen gen-AI-driven synthetic audience representatives—a process that took a few weeks. These personas included characters like Samantha, a 31-year-old woman in Washington, D.C., and they were created from Zipcar’s existing data sets and designed to answer questions specific to the city they lived in. This informed the car-sharing brand on how residents across different U.S. cities used transport.

Supernatural’s AI tool, Merlyn, suggested bespoke ad copy, which was then checked by human copywriters, to encourage people to rent from Zipcar. For example, the tool generated the copy, “Zip if art’s cooler al fresco,” referring to D.C’s Glenstone Museum (where the art is outdoors).

The AI-generated ad copy was presented back to the audience to test messaging and identify areas for improvement.

“Just by making the creative more relevant and compelling, [Zipcar] was able to double its ad awareness and [saw] a 6% increase in site traffic [compared with] the year before,” said John Elder, co-founder and CEO of Supernatural. In 2023, the brand’s website saw a total of 1.3 million monthly unique visitors, compared with 1.23 million in 2022, according to Semrush, via Supernatural.

By surveying 500 people, one-half of whom were exposed to the ad, Zipcar saw a 28% increase in brand awareness while brand recall increased by 81% among those exposed, the company said. To that, 68% of participants reported an improved perception of Zipcar, and brand favorability saw an uptick of 119%. Zipcar wouldn’t share revenue specifics.

“All of this without an increase in budget [while using] AI to precisely uncover attitudes and interests of the target audience, develop a more compelling strategy and then create more relevant work,” added Elder. “This further improved marketing performance.”

Nuances of prompt engineering

Zipcar used Merlyn to understand how audience cohorts differed from the national average in 2022. The brand discovered several characteristics of its audience, including that they value avoiding the commitment of a mortgage, they prefer not to have a spouse or partner, and they tend to lean toward ultra-liberal views.

The agency extrapolated these data sets into synthetic personas, and then asked them questions.

Queries like “Why would you rent a car?” or “Why would you not use Uber?” led to responses such as “I am heading to Ikea, and Uber is too small,” generating details on why people opt for car rentals over ride-sharing services in certain scenarios.

However, prompt engineering, or querying the AI personas multiple times with specific questions, remained critical to brand strategy, helping tease out what makes someone use a car rental.

“The ability to ask questions is quite a difference maker,” said Elder. “The platform has moved eons beyond where it was the year before.”