AutoDrive’s 2026 EV Launch: 30%+ ROAS Boost

If you’re still using basic demographic segmentation for your digital ads, you’re leaving money on the table. For any real brand targeting, creating precise custom audiences is now the absolute baseline. But how does this advanced segmentation actually work to make a campaign successful? It’s about knowing your customer beyond a few simple data points.

We just wrapped a major digital campaign for “AutoDrive,” a big automotive brand, where we used LiveRamp RampID to build some seriously refined custom audiences. The goal was simple: get more qualified leads for their new electric SUV, the “Terra,” in a very crowded market. We ran the campaign for 10 weeks from Q1 to Q2 2026, working with a total budget of $1.2 million. We measured everything against cost per lead (CPL) and return on ad spend (ROAS).

Our whole strategy was to get past broad, interest-based targeting. We knew from experience that potential EV buyers aren’t some monolithic “eco-conscious” group. They’re often tech early adopters, they’re financially secure, and they’re usually engaging with other luxury brands or lifestyle content. Our starting point was to ingest AutoDrive’s own first-party CRM data, a mix of current owners, people who test-drove a car but didn’t buy, and customers who’d been to their service centers. This raw data, once anonymized and encrypted, was the foundation for our entire segmentation plan.

Initial Strategy: Layered Segmentation with First-Party Data

My team started by breaking down AutoDrive’s CRM data into three core buckets:

1. Existing Premium Sedan Owners: These were customers who had already bought high-end vehicles from AutoDrive, so we knew they had brand loyalty and an appetite for premium features.
2. Recent Test-Drive Participants (Non-Purchasers): These folks were clearly interested but walked away. We figured they were hung up on a specific feature, the price, or maybe the timing just wasn’t right.
3. Service Center Patrons (5+ Years Old Vehicles): These are customers with older AutoDrive cars who are probably starting to think about an upgrade.

We onboarded all this first-party data using LiveRamp’s platform, which converted it into privacy-safe, addressable RampIDs. This let us target these exact segments on different digital channels without touching third-party cookies, which is table stakes in today’s privacy-first world. The ability to connect our CRM data to actual people across the web with a persistent, person-level view was, frankly, the most powerful part of the tech stack for this campaign.

Creative Approach: Tailored Messaging

We built a totally different creative strategy for each segment, avoiding the “one-size-fits-all” trap that so many brands fall into. For the Existing Premium Sedan Owners, our messaging was all about how the Terra was a smooth tech upgrade from their current ride, focusing on the luxury interior and advanced infotainment. The ads featured slick visuals and a call to action for exclusive pre-order access.

For the Recent Test-Drive Participants, the creative was designed to overcome their likely objections. If they had test-driven a gas-powered SUV before, our ads hit on the Terra’s superior performance, lower running costs, and the expanding EV charging network. We even used testimonials from new EV owners who had the same initial doubts, with a call to action for a “second look” test drive backed by a better incentive.

Finally, for the Service Center Patrons with Older Vehicles, the ads focused on reliability, modern safety upgrades, and the environmental angle of going electric. We positioned the Terra as the smart, responsible upgrade with great long-term value, showing visuals of family safety features and the convenience of charging at home. The call to action pushed a trade-in appraisal and information on available EV tax credits.

Targeting and Channel Mix

We pushed the campaign primarily across programmatic display, connected TV (CTV), and social media. Because we used RampID, we could activate our custom audiences in all these places. For programmatic, we worked with DSPs that had deep LiveRamp integrations. On CTV, we bought premium inventory during prime time where we knew engagement was high. The social campaigns ran on platforms like LinkedIn and Meta, where we could pair their reach with our sharp audience definitions.

The $1.2 million budget was split like this:

• Programmatic Display: 40% ($480,000)
• Connected TV (CTV): 30% ($360,000)
• Social Media (LinkedIn, Meta): 25% ($300,000)
• Search Engine Marketing (SEM) Retargeting: 5% ($60,000)

That small SEM retargeting budget was our safety net. We used it to run ads specifically to people who had clicked on one of our other ads but hadn’t filled out a lead form yet, ensuring we didn’t lose them along the way.

Results: What Worked and What Didn’t

After the 10-week flight, the numbers spoke for themselves. We brought in 4,800 qualified lead submissions for the Terra, which was 20% over our goal. The overall cost per lead (CPL) ended up at $250. That’s way better than the $350 industry benchmark for luxury auto leads, according to a recent eMarketer report on automotive marketing trends. Our blended return on ad spend (ROAS) hit 3.8:1, which means we generated $3.80 in sales revenue for every $1 we spent.

Here’s how the performance broke down by segment:

The Existing Premium Sedan Owners group was the star performer, which just proves the power of targeting loyal customers who already know and trust your brand. Their higher average purchase price also helped pump up the ROAS for that specific segment. The Recent Test-Drive Participants group also performed very well, which confirmed our theory that the right message could win back people who were on the fence. Even the Service Center Patrons, which had the highest CPL, still delivered a solid 3.2:1 ROAS, making it a perfectly viable audience to target again.

So what didn’t work? For the first week, we ran a control group using broad demographic targeting without the RampID segments. The CPL was a painful $420 and the ROAS was a dismal 1.5:1. The huge difference showed just how much money you waste with imprecise targeting. We also learned that our generic CTV video ads, the ones without a specific call to action for the segment, had terrible completion rates and didn’t generate many leads. This just confirmed for us that personalizing the creative is just as important as segmenting the audience.

Optimization Steps Taken

This wasn’t a “set it and forget it” campaign. We were in there every week running A/B tests on ad copy and visuals for every segment. We learned some interesting things. For example, the Existing Premium Sedan Owners responded way better to visuals of the Terra charging at a fancy house than at a public station. For the Service Center Patrons, though, images showing the car’s safety features and big interior got more clicks.

We were also careful about frequency capping to avoid annoying people, especially in those smaller, high-value segments. We tweaked bid strategies on the fly, pushing more money to the channels and segments that were giving us the best results. Halfway through the campaign, when we saw how well the “Existing Premium Sedan Owners” segment was doing, we moved 10% of the budget over from the “Service Center Patrons” segment to double down on our winner. That pushed its ROAS even higher.

We also spent time refining the landing page experience. This was a big one. We made sure that for each segment, the landing page they hit after clicking an ad continued the same story. The page for “Recent Test-Drive Participants,” for instance, had a whole section busting common EV myths and a big button to schedule an extended test drive. You have to continue the narrative. The data showed that this consistency between the ad and the landing page was a major reason our CPL got so low.

Frankly, LiveRamp RampID was what made these kinds of granular optimizations possible. If you don’t have a persistent identifier to track users as they move between their laptop, their phone, and their smart TV, you can’t accurately attribute a conversion back to a specific ad or audience. You can’t optimize what you can’t measure, and I’m constantly telling clients that good measurement starts with good identity.

This campaign proved that putting the work into advanced identity resolution and building thoughtful custom audience segments pays off. It’s about reaching the right people with the right message at the right time to get the most out of every single ad dollar.

AutoDrive’s success with the Terra campaign really just points to where marketing is going in 2026. The brands that are winning are the ones that actually understand their customers as individuals and are able to build personalized campaigns at scale.