AI-First CX: SunTrust’s 2025 Strategy Shift

The pursuit of understanding and predicting customer behavior has long been a foundation of successful marketing. Today, an AI-first CX strategy isn’t merely about reacting to current customer interactions. It’s about proactively anticipating future user needs, shaping experiences before they even form. How can marketers design campaigns that truly resonate with an audience whose expectations are constantly shifting, often driven by the very AI tools they use?

Campaign Teardown: “Future-Fit Finance” with AI-Driven Personalization

In late 2025, a regional financial institution, SunTrust Bank (now part of Truist), launched its “Future-Fit Finance” campaign. The goal was to attract a younger demographic (25-40 years old) to its suite of digital wealth management and personalized lending products, specifically targeting those who prioritize financial planning tools that adapt to their evolving life stages. This wasn’t just about selling products. It was about building a relationship predicated on predictive support. The campaign ran for three months, from October to December 2025.

Strategy: Proactive Engagement Through Predictive AI

The core strategy revolved around an AI-first customer strategy that moved beyond simple personalization. SunTrust aimed to identify potential financial needs before customers articulated them. This involved creating predictive models based on anonymized transaction data, life event triggers (like job changes, home purchases, or family growth, inferred from aggregated public data and opt-in user profiles), and digital behavior within their banking app. The models were designed to predict which financial products or advice would be most relevant to an individual in the next 6-12 months. For instance, a user consistently saving for a down payment, indicated by specific transaction patterns, would receive tailored content on mortgage options and first-time homebuyer seminars, not just a generic savings ad.

According to a Statista report from 2024, companies excelling in customer experience generate 1.5 times more revenue than their competitors. SunTrust understood that superior CX wasn’t just about service recovery. It was about preemptive value delivery. Their approach sought to provide value before a need became an explicit problem.

Budget and Key Metrics

The total budget for the “Future-Fit Finance” campaign was $400,000. This was allocated across various channels, including programmatic display, social media (LinkedIn, Reddit, and emerging financial forums), and in-app notifications. Here’s a breakdown of the key performance indicators:

• Duration: 3 months (October 1 to December 31, 2025)
• Impressions: 35 million
• Click-Through Rate (CTR): 1.8%
• Conversions (Account Sign-ups/Product Applications): 18,500
• Cost Per Lead (CPL): $8.50 (for initial engagement, e.g., whitepaper download or tool usage)
• Cost Per Conversion (CPC): $21.62
• Return on Ad Spend (ROAS): 2.8x (measured by projected lifetime value of new accounts against ad spend)

These metrics, particularly the ROAS, demonstrated a solid return for a financial services campaign, where customer acquisition costs can often be significantly higher. The predictive element was key to this efficiency.

Creative Approach: “Your Future, Anticipated”

The creative strategy centered on the theme “Your Future, Anticipated.” Visuals avoided generic stock photos of smiling families. Instead, they featured abstract, clean designs with subtle animations that suggested foresight and smooth integration. For example, a common ad creative showed a digital interface subtly shifting to present a relevant financial option, like a personalized investment portfolio suggestion, just as a user might be thinking about it. The copy focused on benefits like “Financial Guidance That Knows You Before You Do” or “Proactive Planning, Effortless Growth.”

Interactive elements were important. On landing pages, users encountered a brief, AI-powered questionnaire (developed using Typeform‘s advanced logic features) that dynamically adjusted based on their responses, immediately providing a personalized “financial health score” and suggesting next steps. This immediate, tailored feedback loop was a significant departure from static lead generation forms.

Targeting: Micro-Segments Driven by Behavioral AI

Traditional demographic targeting (age, income bracket) formed a baseline, but the campaign’s effectiveness stemmed from its micro-segmentation. Using an internal AI model, developed with Amazon Personalize, SunTrust identified several distinct behavioral segments:

• “Growth Seekers”: Individuals frequently researching investment options, showing higher-than-average savings rates, and engaging with financial news.
• “Life Planners”: Users exhibiting patterns indicative of upcoming major life events, such as increased spending on home improvements, searches for family planning resources, or new job notifications on LinkedIn.
• “Debt Consolidators”: Those with multiple credit lines, consistent minimum payments, and searches related to debt management.

Each micro-segment received highly customized ad copy, creative variants, and landing page experiences. The AI continuously refined these segments based on real-time engagement data, adjusting bid strategies and creative rotations to maximize relevance. This dynamic optimization was a continuous feedback loop, not a set-it-and-forget-it approach.

What Worked: Precision and Proactivity

The most successful element was the campaign’s ability to deliver highly relevant content at precisely the right moment. The CPC of $21.62 for a financial product sign-up was notably efficient. This was largely due to the AI’s predictive capabilities. Users felt understood, not just targeted. For example, one creative variation targeting “Growth Seekers” with an ad discussing robo-advisory services saw a CTR of 2.1%, significantly higher than the campaign average, and a conversion rate of 12% directly to account sign-ups. This demonstrated the power of anticipating a user’s intent rather than simply reacting to past behavior.

Another strong performer was the AI-powered chatbot on the landing pages, built using Google Dialogflow. It handled initial queries, qualified leads, and guided users to the correct product or human advisor. This reduced the load on customer service teams and contributed to a smoother user journey, which in turn lowered the overall CPC by minimizing unqualified leads. I always advocate for using AI in these initial touchpoints. It’s where you can make the biggest impact on efficiency and user experience without sacrificing the human touch for complex issues.

What Didn’t Work: Over-Reliance on Implicit Data

Not every aspect was a resounding success. An initial attempt to infer “life event” triggers solely from broad, implicit behavioral data (e.g., changes in online shopping habits) without explicit user consent or additional data points proved less effective. For instance, an algorithm might interpret increased online baby product purchases as an impending birth, leading to family savings plan ads. While often correct, a significant percentage of these inferences were inaccurate, leading to irrelevant ad impressions and a higher CPL for those specific segments. This particular sub-segment had a CPL of nearly $15, almost double the campaign average. It became clear that while AI is powerful, a blend of explicit user data (opt-in surveys, profile preferences) and inferred data creates a more strong and ethical predictive model. You can’t just rely on the black box. Transparency and user control are paramount.

Optimization Steps Taken: Balancing AI with User Consent

Recognizing the limitations of purely implicit data, the SunTrust team implemented several optimization steps:

1. Enhanced Opt-in Mechanisms: They introduced clearer, more appealing in-app prompts for users to share life stage information (e.g., “Planning for a major purchase?” or “Tell us about your financial goals for personalized insights”). This significantly improved the accuracy of the “Life Planners” segment.
2. A/B Testing of Predictive Models: The team continuously A/B tested different predictive algorithms. One test involved comparing a model heavily reliant on transaction data against one that balanced transaction data with explicit user preferences. The hybrid model consistently outperformed, showing a 15% increase in conversion rates for the same ad spend.
3. Dynamic Creative Optimization (DCO) Refinement: While DCO was used from the start, the team refined the triggers for creative variations. Instead of just changing headlines, they began to dynamically adjust the entire ad layout and call-to-action based on real-time engagement metrics and the predicted user need. For example, if a user spent more time on a retirement planning article, subsequent ads would highlight features of their retirement accounts, complete with a direct link to a personalized projection tool. This led to a 0.3% increase in overall CTR in the last month of the campaign.
4. Feedback Loop Integration: They integrated a discreet “Was this helpful?” feedback option on personalized content. This qualitative data, while small in volume, provided valuable insights for further training the AI models, helping to fine-tune the understanding of true user intent versus algorithmic inference.

By the end of the campaign, the refined AI models were achieving an accuracy rate of 88% in predicting the next most relevant financial product for a user, a significant jump from the initial 75%. This enhanced accuracy directly translated into better ROAS and a more positive customer experience, illustrating the iterative nature of successful AI implementation in CX.

The “Future-Fit Finance” campaign demonstrated that an AI-first CX strategy is not just about automation. It’s about intelligent anticipation and continuous refinement. By understanding and predicting the future needs of their users, SunTrust was able to deliver a highly personalized, efficient, and in the end successful campaign.