Active Intelligence: 3.5x ROAS in 2026

You can’t just stare at raw data and expect campaign performance to improve. You have to turn that data into actual decisions, which is where platforms like Active Intelligence Wavelength come in. This breakdown of a recent digital ad campaign shows exactly how our structured plan, paired with real-time analytical capabilities, steered the project to a successful outcome.

Deconstructing the “Spring Forward” Campaign

We ran the “Spring Forward” campaign for a mid-sized e-commerce client selling sustainable home goods back in Q1 2026. Their goal was straightforward: get qualified leads and sales for a new line of eco-friendly cleaning products. We had a three-month window from January 15 to April 15 and a total budget of $75,000. We planned it all out, of course, but you always have to be ready to adjust on the fly.

Strategy and Initial Setup

Our plan was to hit them on two fronts: Meta Ads (Meta Business Help Center) and Google Ads (Google Ads documentation). We front-loaded the budget with 60% on Meta to build awareness and grab some early leads, leaving the other 40% for Google to catch bottom-of-funnel traffic through search and shopping. On the Meta side, we built lookalikes from past buyers, targeted interests like “sustainable living” and “eco-friendly products,” and went after website visitors who had bounced. For Google Ads, it was all about high-intent keywords, things like “biodegradable cleaning supplies” and, naturally, the client’s brand terms.

Creative Approach: Beyond the Visual

For the creative, we went for authentic and effective. We made a bunch of short videos for Meta showing the products being used in real homes, no fake sets, and paired them with carousels that pointed out specific benefits. The copy was all problem/solution, hitting on worries people have about harsh chemicals and their environmental footprint. On Google Search, the ads were tight, packed with keywords, and had clear CTAs like “Shop Now for a Greener Home.” We intentionally skipped the glossy, perfect-looking photos for something that felt more like user-generated content. I think people chase high-production value too much. A real connection almost always performs better.

The Role of Active Intelligence Wavelength

This was the first time we used an Active Intelligence Wavelength platform across an entire campaign for real-time changes. The system just pulls in data from everything, the ad platforms, Google Analytics 4, our CRM, and gives you one dashboard to see what’s happening. Its machine learning spots trends and tells you what to do next. For example, it quickly flagged that our videos with real customer testimonials were getting a 25% higher engagement rate than the ones with a slick professional voiceover, so we immediately shifted our creative budget to make more of what was working.

Initial Performance Metrics and Challenges

The first two weeks were a mixed bag. Our Meta campaigns pulled in a ton of impressions (12.5 million) and a 1.8% CTR, which felt good since the industry average is around 1.5% for lead gen based on that Statista report. The problem was the cost. Our CPL on Meta was way too high at $12.50, and the landing page was only converting at 3.2%. Meanwhile, Google Ads had fewer impressions (3.2 million) but was bringing in much higher-intent traffic, with a much healthier CPL of $8.70.

Optimization Steps and Wavelength Insights

This is where the Active Intelligence Wavelength platform really started to earn its keep. It was constantly watching hundreds of data points, flagging ad sets that were bleeding money and telling us exactly where to shift the budget. Here’s a quick rundown of the big moves we made:

• Audience Refinement: The platform saw that one specific lookalike audience, people who viewed product pages for over a minute but didn’t buy, had a way lower CPL on Meta. So we pushed 30% more budget to that group and killed two broader interest audiences that were just too expensive.
• Creative A/B Testing: Wavelength made A/B testing copy and visuals super fast. We learned that creative with close-ups of the product textures and copy that mentioned “plant-derived ingredients” beat our generic lifestyle shots by 15% on CTR. We just started making more ads like that.
• Landing Page Optimization: The platform connected to our site analytics and showed us where the landing page was failing. Heatmaps showed people on mobile were scrolling right past the “add to cart” button. Whoops. We added a sticky CTA button and clarified the product benefits, which bumped the conversion rate for Meta traffic from 3.2% to 4.5%.
• Bid Strategy Adjustments: On the Google Ads side, Wavelength suggested we switch from a “maximize conversions” bid strategy to a “target ROAS” strategy (Google Ads Smart Bidding) for our best-selling product categories. This let Google’s bidding AI optimize for actual revenue, not just the number of conversions.
• Negative Keyword Expansion: The platform also kept finding search queries that got clicks but no conversions and automatically suggested them as new negative keywords for our search campaigns. That simple fix cut our wasted spend by about 8%.

Results and Final Performance

After three months of these data-led optimizations, the final numbers were solid. Our overall CPL fell from $10.64 to $8.30 (a 22% drop), conversions shot up, and we hit a 3.5x ROAS, well above our 2.8x goal. This came from reading the small signals in the data and moving fast. If we didn’t have the Wavelength platform’s constant feedback, hitting this level of granular optimization would’ve taken way more people and probably wouldn’t have worked as well.

What Worked Well:

• Data-Driven Agility: The Wavelength platform’s real-time analytics let us spot performance issues immediately and act on them. This was the key to keeping the campaign efficient.
• Creative Iteration: We were constantly A/B testing creative, using hard data to make sure our message was always the most effective one we could be running.
• Audience Segmentation: Being able to refine audience segments on the fly based on real engagement and conversion data directly led to the CPL reduction.

What Didn’t Work (or could have been better):

• Initial Budget Split: Meta gave us tons of impressions, but the high initial CPL was a classic reminder that big volume doesn’t always mean quality leads. In hindsight, a more balanced 55/45 or 50/50 split at the start might’ve been more efficient.
• Early Landing Page Experience: We definitely dropped the ball on the mobile friction on the landing page at first. More aggressive pre-launch testing across different phones would have caught that.
• Attribution Complexity: Even with Wavelength pulling all the data into one place, figuring out the exact credit for each touchpoint in a multi-channel journey is still a huge challenge. No tool has completely solved that yet.

The “Spring Forward” campaign proves that a good strategy and creative aren’t enough. The real edge comes from the constant, intelligent optimization of campaign performance. In this field, using a tool like Active Intelligence Wavelength for real-time insights and predictive analytics is a flat-out necessity if you want to maximize your return on investment.