top of page

Personalized Nurture Campaigns

The opportunity

CarMax
I
Jan 2026

Car shopping doesn't happen in one session

​

Customers often move in and out of the shopping journey—saving searches, favoriting vehicles, returning to the site, or completing pre-qualification before they're ready to buy.

​

We saw an opportunity to use these signals to create more relevant communications and help customers pick up where they left off accross 6 distinct customer segments.

The strategy

Make the experience smarter as we learn more

We designed nurture journeys that became increasingly personalized as customers gave us more signals about their intent.

​

Behavioral signals
Return shopping · Account creation · Saved search · Favorited vehicle · GEICO total loss · Third-party shopping
​

↓
Relevant first touchpoint
↓
Personalized vehicle recommendations
↓
Continue shopping
↓
Reservation & purchase

Influencing the recommendation engine

Defining the signals behind personalization

The experience depended on more than designing the communications—we also needed to determine what made a recommendation relevant.

​

I partnered closely with our recommendation engine team, meeting weekly to help define how customer behaviors should influence vehicle recommendations.

 

Together, we established a hierarchy of intent signals, giving stronger actions—like saving a vehicle—more weight than lower-intent behaviors like viewing one.

Upper-funnel nurture

Bringing shoppers back with relevant inventory

Each audience entered through a communication tailored to their behavior or context. From there, eligible customers moved into a recommendation campaign featuring personalized vehicles designed to bring them back into the shopping experience.

Testing the experience

Would recommendations actually change shopping behavior?

​

We ran a 50/50 test comparing existing communications against an experience that added four personalized recommendation emails.

​

Control
Existing communications

Test
New communications → Recommendation 1 → Recommendation 2 → Recommendation 3 → Recommendation 4

6%

Click rate

+75%

Customer → sale

+6%

Online shopping activity

+35%

Customer → reservation

~$7M

Estimated incremental annualized value

Learning and iterating

One experience didn't work for everyone

Customers entering through Inventory Alerts saw a decline in customer-to-sale conversion.

 

Rather than scaling the experience universally, we excluded that audience and continued investigating how their needs and intent differed.

​

The successful experience was scaled to Account Creation and Shop No-ID audiences.

Extending personalization to Pre-Qual

What if we knew even more about the shopper?

Pre-qualification gave us additional context about both customer intent and the vehicle they were considering.

 

We extended the nurture strategy to these higher-intent customers, using the vehicle they pre-qualified from to surface personalized inventory recommendations and help them continue shopping.

Designing around customer state

The next message depended on where the customer was in their journey

​

The experience combined email and SMS across key moments after pre-qualification. Customers could receive messaging encouraging them to continue with their current vehicle or personalized recommendations to keep shopping.

​

Pre-Qual completed
↓
Financing communication
↓
Continue with this vehicle → Progress / reservation nudges

Keep shopping → Personalized vehicle recommendations

1.66%

Conversion to sale

~$~$8.8M

Estimated incremental annualized value

What we learned

Relevance meant more than recommending the right car

Performance varied based on customers' financing outcomes. Straight-approved customers responded positively, while several other segments were flat or negative.

​

This reinforced that effective personalization needed to consider both what a customer wanted to shop for and where they were in their journey.

What's next

Building toward a connected personalization ecosystem

The results gave us a foundation for expanding personalized recommendations throughout the shopping journey.

​

Next steps included adding more recommendation touchpoints, incorporating financing results into communications, introducing inventory and expiration nudges, and continuing to tailor experiences based on customer state.

next project

I

bottom of page