Vol 43. Carvana: Getting personal with AI 📽️

Imagine you have a business and that business reached the 1 million customer mark. 

You’re probably going to want to celebrate that huge milestone, but not just with a sheet cake at the office. Why not celebrate with your customers who made it all possible—like, all of them. 

That was the idea Carvana had when it sold its 1 millionth car. 

This week, Case Studied® explores how Carvana gave 1.3 million customers their own personalized, AI-generated video.

The Brief:

Pardon the pun but Carvana’s had quite a wild ride since its founding in 2012. It had a bumpy IPO in 2017, going public at $15 a share and traded at $8 during the first week. 

It saw a meteoric rise during the COVID-19 pandemic—after all, it was the only place you could buy a car without coming in contact with another person—and by August 2021 it was trading at over $370 a share. 

Then came the tough times. Spending accelerated but growth didn’t follow. Logistical issues dented its reputation. Carvana reported a $806 million loss in Q4 of 2022 and its share price closed at an all time low of $3.72 on Dec. 27, 2022. 

Despite this tumultuous ride, Carvana managed to stick around. It took cost-cutting actions in 2023 and its stock price recovered. As of this writing, it was trading at $150 per share. And along this wild ride, the company racked up one million cars sold in 2023. 

To celebrate, Carvana wanted to take “each and every customer on a joyride of the day they bought their Carvana car with an AI-generated video tailored to their car-buying journey.”

Hire marketing agencies like a Fortune 500 CMO 📈

Tired of running marketing agency searches only to end up with an agency that can’t deliver results.

Discover why Microsoft, Shopify, Spotify and 100’s of brands use Vendry to gather proposals, vet and hire marketing, design and dev agencies for free.

If you’re running an agency search before the end of 2024, book a 15-minute call and tell us about it. Plus, we’ll send you a $100 Uber Eats gift card after the call to enjoy lunch on us.

_Promoted

The Execution:

To make the 1.3 million personalized videos for their “Joyride” campaign, Carvana teamed up with the agency 1stAveMachine. Together, they decided what data points to include, what the images made sense, and what scripted words could be relevant. 

“From animating customer names to making flags for national holidays and beyond, we spent months brainstorming ways to create an all-encompassing snapshot that could help customers re-live the moment they met their car,” Carvana said in a blog. 

As a standard practice, Carvana photographs the vehicles they sell from every angle to make a 3D model of the vehicle. Fortunately for them, these assets seamlessly integrated into the videos themselves, in addition to animation and motion graphics. The final piece was voice over narration. 

“[We] knew that asking a voice actor to record millions of lines to cover every possible combination of “name, vehicle, place, season, and cultural events” wasn’t exactly feasible,” the company said in a blog. 

Instead, they hired a voice actor to record different lines, and Carvana used “two emergent technologies—text-to-speech and re-voicing—to turn scripts into audio bits that sounded close to our actor. After applying a filter to blend the two, we ended up with the best of both worlds—AI voiceover, fluidly combined with our voice actor’s original work to create a unique narration for every video.”

The final result (see an example here) is roughly one minute and 47 second-long and highlights the bond between an owner and their car. 

The voiceover (which, to us, sounded like Paul Rudd at first) narrates the story from the perspective of the car, with opening lines like, “I’ll never forget the day we met. I was so excited. It was September 29th, 2019. Everyone was talking about the ending of Game of Thrones. It was National Coffee Day—yum.”

The video goes on with lines like, “I was a wide-eyed and innocent 2019 Chevrolet Corvette and they had just told me the name of my new owner—Holly. I hope I’m pronouncing that right” and “I know this is crazy but do you believe in fate? I mean, out of Carvana’s millions of cars and tens of millions of online visitors, you picked me. Me.”

Vehicle owners received an email with their unique video that included highly personalized information, down to the local spots that the customers could go in their vehicles. 

Interestingly, there was no big advertising push to hype up Joyride. All Carvana did was promote it on social media and encourage owners to post their videos themselves.

The Results:

Carvana ended up producing 1,345,948 videos, which shakes out to about 45,000 hours of personalized content—back-to-back, that’s over five years worth of video. 

For this to even be possible, Carvana’s cloud-based system was capable of rendering up to 300,000 per hour.  

It’s worth noting that with such a high volume of content, not every video is perfect. Factual inaccuracies did occur, which Carvana co-founder and chief brand officer Ryan Keeton acknowledged in an interview with AdAge. Regardless, Keeton was keen to take advantage of AI technology, saying, “AI is so topical right now, so it’s cool to take our own swing at it.”

The Takeaways:

This campaign is packed with lessons on personalization, AI, and experimentation. Here are some standout takeaways:

1. Leverage AI for Scalable Personalization

Carvana’s “Joyride” campaign is a great example of the potential of AI in creating highly personalized content at scale. Brands should explore AI-driven tools that can leverage customer data to deliver tailored experiences—whether through personalized videos, emails, or product recommendations.

Start by identifying key customer data points you already have that can be turned into personalized content. For smaller brands, using AI tools like chatbots, text-to-speech, and dynamic content creation to enhance engagement and deepen customer connections without overwhelming resources.

2. Embrace Experimentation and Learn from It

Carvana’s willingness to embrace AI, even with the risk of imperfections, highlights the need for experimentation, especially when tinkering with novel technology. Brands should adopt a test-and-learn mentality when it comes to testing new things and be prepared to adapt when things don’t go as planned.

Start small—pilot a new AI-driven initiative or personalized marketing effort, analyze the results, and refine based on feedback and data. This iterative approach allows businesses, big or small, to innovate continuously, stay adaptable, and improve their customer experience without committing large budgets upfront.

3. Create Content That Builds Emotional Connections

Carvana’s personalized videos were effective because they tapped into the emotional relationship customers have with their cars (and it’s price tag!). Brands can apply this principle by crafting content that resonates emotionally with their audience, reflecting shared values or significant moments.

This can be achieved through storytelling that centers around customer experiences, community involvement, or brand heritage. Emotional content is more likely to be shared, remembered, and talked about, giving brands a competitive edge in both customer retention and acquisition.

ICYMI:

Missed the last week of Case Studied? We’ve got you covered.

Last week, we covered how one of the greatest product collaborations of all time, the Taco Bell and Doritos Locas taco. Read about how this partnership came to be and how you can apply this case study to your brand.

Your perfect passenger: Who’s riding shotgun on the way to your next successful campaign? Get matched with vetted agency partners who will be there with you during every leg of the journey. Start here.