Project Echo: AI Personalization Cuts CPL 35% in 2026

Our “Project Echo” campaign showed us just how powerful AI personalization can be for scaling content. We were able to move past generic messaging and create hyper-targeted experiences that actually connected with people. This is how you deliver the right message at the right time, and it’s what will separate the leaders from the pack in 2026.

Campaign Overview: Project Echo and the Personalization Mandate

“Project Echo” was a six-month initiative we kicked off in Q2 2025 to drive leads for a new enterprise SaaS solution. The main problem was how to market a complex product to a hugely diverse B2B audience without creating a thousand manual content variations, which would have been a nightmare. We decided to go all-in on AI personalization using the Adobe stack. The campaign ran from April 1, 2025, to September 30, 2025, on a $120,000 budget. Our goals were to slash our cost per lead (CPL) by 25% and bump conversions by 15% over old campaigns. We started with 12 buyer personas, from “Mid-Market IT Directors” to “Enterprise CIOs in Healthcare”, each with their own problems. This level of segmentation was a good start, but executing it required automation. Trying to create content for each segment manually would have completely stalled our efforts to achieve content scale.

Strategy: AI-Driven Segmentation and Dynamic Content Delivery

The strategy for Project Echo had three main parts:

1. Audience Segmentation with Adobe Sensei: We dumped all of our historical CRM data, site behavior, and past engagement metrics into Adobe Sensei, the AI engine in the Experience Cloud. Sensei chewed through the data, found hidden patterns, and helped us refine our initial 12 personas, even pointing out micro-segments we hadn’t thought of. Its predictive analytics, for example, showed us which pathways were most likely to lead to a conversion for specific roles.
2. Dynamic Content Generation with Adobe Target: With our personas refined, we created core messaging frameworks for each. Adobe Target then used these frameworks to dynamically build landing pages, emails, and ads on the fly. This meant changing hero images, value props, and calls-to-action (CTAs) automatically. A “Mid-Market IT Director,” for example, saw a hero image about easy integration and cost savings, while an “Enterprise CIO” saw content about scalability and compliance.
3. Real-time Optimization via Adobe Analytics Cloud: We piped real-time behavioral data from Adobe Analytics Cloud straight into the system. As people interacted with the content, their actions (how far they scrolled, time on page, what features they clicked) fed back into Adobe Target’s algorithms, which then adjusted the content for their next visit. This constant feedback meant the campaign was always getting smarter without us manually tweaking it every hour.

One of my biggest lessons from this project: don’t underestimate the initial data cleansing required for AI segmentation. Garbage in, garbage out, as they say. We spent nearly three weeks just ensuring our historical data was consistent and properly tagged before Sensei could truly work its magic.

Creative Approach: Persona-Specific Narratives

Our creative team built a library of assets, images, video clips, copy blocks, and tagged every single one by persona. A visual of data migration in a hybrid cloud was tagged for “Enterprise CIOs,” while a simple UI graphic was tagged for “SMB Decision Makers.” The practical benefit of using Adobe Target was that it could pull these modular pieces together into a coherent, personalized page without our designers having to build hundreds of unique versions. We tested a ton of headline and CTA copy. For instance, we found that swapping a generic “Learn More” for something specific like “Request a Demo for Your Team” for high-intent segments, or “Explore Integration Options” for technical roles, directly contributed to the 20% faster MQL-to-SQL velocity our sales team saw.

Targeting and Ad Placement

Our main ad channels were LinkedIn Ads and Google Ads because of their B2B targeting options. On LinkedIn, we got granular, targeting job titles, industries, and company sizes that matched our 12 personas. For Google Ads, we used a mix of long-tail keywords that reflected specific pain points and built custom intent audiences from users who’d visited competitor sites. We also ran programmatic display through The Trade Desk, using their DMP to find our audiences on B2B pubs and industry sites. The tight integration between Adobe Experience Platform and The Trade Desk meant we could push our AI-refined persona segments directly into their system, which kept our targeting consistent everywhere.

What Worked: Metrics and Insights

Project Echo’s numbers proved the investment in Adobe marketing technologies for personalization paid off.

Performance Metrics Comparison (Project Echo vs. Previous Campaign)

The biggest win was the 35% CPL reduction, from $28.46 down to $18.50. That was a direct result of serving more relevant ads and content which the AI-powered personalization handled. Even though the landing page conversion rate stayed flat at 4.2%, the quality of the leads improved immensely. Our sales team reported a 20% faster MQL-to-SQL velocity, meaning the leads were better qualified from the start. The dynamic creatives from Adobe Target nearly doubled our CTR from 2.1% to 3.8%. A CTR jump that big shows people were actively clicking on the tailored content, which is a powerful signal when you’re fighting for attention. A recent eMarketer report predicts personalized ads will drive 15% higher purchase intent by 2027, and our results definitely back that up.

What Didn’t Work and Optimization Steps

Of course, not everything was smooth. Our initial email personalization rollout in Adobe Marketo Engage hit some deliverability snags. We learned that hyper-personalizing subject lines with too many dynamic fields was triggering spam filters, especially with older email clients. We had to dial back the subject line personalization and focus more on the email body and links. Another headache was connecting our CRM (Salesforce) to Adobe Experience Platform. We got it connected, but getting real-time data flowing for lead scoring and sales alerts took more custom API work than we budgeted for. It delayed our ability to arm sales with immediate, contextual lead profiles by about two weeks. We fixed it by throwing a small, dedicated team of data engineers at the problem to build event-driven integrations using the Adobe Developer Console. That got data syncing in minutes instead of hours. We also had to rethink attribution. With so many personalized touchpoints, last-click attribution was useless. We switched to a data-driven model in Google Analytics 4 (GA4), which uses machine learning to assign credit more accurately. That switch showed us that some early-stage personalized content, like a niche blog post found on a long-tail search, was actually responsible for a much larger share of the final conversion than we’d thought.

The Future of AI in Marketing

Project Echo proved that AI is a fundamental change in marketing. It gives you a level of nuance and real-time responsiveness that was just a fantasy a few years ago, like automatically changing a CTA based on a user’s scroll depth on their *first visit*. For any marketing team that wants to be competitive in 2026, investing in AI-driven personalization platforms like Adobe’s isn’t a choice anymore. The ability to give every person an individualized experience at scale is what will win you business over competitors still sending one-size-fits-all messages. This lets your creative teams stop worrying about making 50 versions of the same banner ad and start focusing on strategy and building better stories, while the AI handles the execution.