Cookieless Ads: 35% Conversion Boost for TerraThreads

This is a breakdown of a media buying campaign we just ran completely blind, with no third-party cookies. The client was an apparel e-commerce brand in the sustainable fashion space, and the big question was whether we could hold the line on performance, or even improve it, as cookies disappeared and privacy rules got tighter. We had to change our entire approach to strategy and measurement to get real results. Here’s how we did it.

Campaign Overview: Sustainable Apparel E-commerce

We ran a six-month, $450,000 digital media campaign for “TerraThreads,” a direct-to-consumer sustainable apparel company. The main goal was simple: sell their new spring collection by getting new customers in the door and getting old ones to buy again. We couldn’t just use our old playbook. We had to fundamentally change how we found people, who we targeted, and how we measured whether any of it was working in this new privacy-first environment. The entire campaign ran from January 1, 2026, to June 30, 2026.

Strategic Pillars: First-Party Data and Contextual Relevance

Our whole strategy was built on two things: getting everything we could out of their first-party data and going all-in on contextual targeting. We knew that depending on third-party cookies for building audiences and retargeting just wasn’t going to work anymore. Luckily, TerraThreads had a solid customer relationship management (CRM) system packed with purchase history, email engagement, and loyalty data. We integrated that CRM data directly with ad platforms using secure hashed identifiers, which let us build custom audiences of existing customers for very precise retargeting without needing to track them across other websites. For example, we could take a segment of everyone who bought eco-friendly denim in the last year and show them ads for the new organic cotton jeans. Contextual targeting, which used to be an afterthought, became a core part of the plan. We found good placements on sustainability blogs, environmental news sites, and ethical living publications, aiming to put our ads where the audience was already thinking about things that aligned with TerraThreads’ values. We even used semantic analysis tools to make sure our ads appeared next to articles with a positive or neutral sentiment, not just the right topic.

Creative Approach: Authenticity and Engagement

The creative had to feel authentic and have a really strong call-to-action. We ditched the generic product shots and instead told stories that showed TerraThreads’ work in ethical sourcing and sustainable manufacturing.

• Video Ads: We found that short (15-30 second) videos showing a garment’s journey from an organic cotton field to the finished product worked incredibly well, especially since they were optimized with captions to be effective even with the sound off.
• Interactive Display: We also tried out HTML5 banners where people could play around with different fabric textures or see product variations right in the ad unit, and that really boosted engagement.
• User-Generated Content (UGC): We got permission to use customer testimonials and photos, which put real people in TerraThreads apparel. That kind of social proof builds a ton of trust when you can’t just follow someone around the internet with retargeting ads.

The creative team cranked out over 50 different ad variations in multiple formats. Having that much creative diversity to test and swap in and out is absolutely required now. The days of running three static banners and calling it a day are long gone.

Targeting and Audience Segmentation in a Cookieless World

Without third-party cookies, our targeting methods had to change completely.

• First-Party Audiences: As I mentioned, we took hashed email lists from the CRM and uploaded them to platforms like Google Ads Customer Match and Meta Custom Audiences. This let us build powerful segments like “High-Value Repeat Purchasers,” “Recent Purchasers (0-30 days),” and “Email Subscribers (Non-Purchasers).”
• Lookalike Audiences: We then used those first-party segments to generate lookalike audiences on platforms that still support them, expanding our reach to new people who behaved like our best customers.
• Contextual Targeting: We worked with demand-side platforms (DSPs) that had good contextual tech. Instead of building user profiles, these platforms just analyze page content in real-time. An article about reducing your carbon footprint might trigger our ad for TerraThreads’ recycled polyester activewear. It’s that direct.
• Geographic and Demographic: We still used the basics. Standard geographic targeting (like in cities with lots of environmentally-conscious shoppers) and broad demographic data (age 25-54, with stated interests in sustainability) were our foundational layers.

A big mental shift for the team was moving from hyper-specific, individual-level targeting to working with more aggregated, privacy-safe groups. We had to accept that we wouldn’t know every detail about every user, and then we built the whole strategy around that new reality.

Performance Metrics and Data Analysis

The campaign ran for 26 weeks, from January 1 to June 30, 2026. Here’s the raw data:

The numbers speak for themselves. The CTR went up by 0.4%, which told us the contextual placements and creative were hitting the mark. But the real story is the massive 34.4% drop in CPA and a 5.0x ROAS. This proves that a well-planned cookieless campaign can actually beat the performance of the old methods.

What Worked: Data, Creativity, and Measurement

Here’s what really moved the needle for this campaign:

• First-Party Data Activation: Activating TerraThreads’ CRM data was the most important thing we did. We used those hashed email lists to build custom audiences for retargeting and to seed lookalike models, which let us spend money very efficiently on users we knew were high-intent (or looked just like them). An IAB report (“First-Party Data in a Privacy-Centric World”) notes that advertisers doing this well see a 2.9x revenue lift, and our results certainly back that up.
• Server-Side Tagging and Enhanced Conversions: We set up server-side tagging via Google Tag Manager and turned on enhanced conversions for Google and Meta. This meant conversion data was sent directly from TerraThreads’ server to the ad platforms, which patched a lot of the measurement holes left by disappearing cookies and gave the algorithms much better signals to optimize on. This was the single most impactful technical change we made.
• Diverse and Engaging Creatives: The budget we spent on making lots of different creative, especially video and interactive ads, paid for itself many times over. We were constantly A/B testing everything, headlines, images, CTAs. The ads that focused on the brand’s sustainability mission always beat the generic product-on-white-background ads.
• Privacy-Centric Measurement Framework: We finally got away from last-click attribution and switched to a data-driven model inside Google Analytics 4 (Google Analytics 4 Help). We also implemented Google’s Consent Mode v2, which helped us model and recover some of the conversion data we would have otherwise lost from users who declined cookies. This gave us a much more complete picture of the whole customer journey.

What Didn’t Work and Optimization Steps

Of course, not everything worked right out of the gate.

• Initial Broad Contextual Campaigns: Our first attempt at contextual targeting was way too broad. Just targeting general “fashion” or “eco-friendly” sites led to a lot of wasted impressions and pretty low engagement. It was a mess.
• Over-reliance on Automated Bidding (Early Stages): Automated bidding is great, but in the first couple of weeks, the algorithms were struggling because they had so little cookieless data to work with. We found ourselves having to jump in and manually adjust things way more than we expected, particularly when launching new campaigns.

We made adjustments on the fly:

• Refining Contextual Targeting: We got much more specific, narrowing our contextual segments to niche topics like “organic cotton production,” “sustainable fashion certifications,” and “circular economy clothing brands.” This move alone boosted CTR by 0.8% in those placements.
• Phased Budget Allocation: We stopped launching with a full budget. Instead, we started with just 30% for the first two weeks. This gave the platforms time to learn and let us spot the losing segments before we’d burned a ton of cash. This probably saved us about 15% of the initial budget that would have gone to bad targeting.
• Manual Bid Adjustments for New Segments: When we launched a new audience or contextual placement, we started it on manual bidding. This gave us direct control to find the right performance baseline before we felt comfortable handing the keys back to an automated strategy and let us feed the algorithms better starting signals.
• Landing Page Optimization: We saw some contextual ads driving great clicks, but the conversions weren’t happening on the corresponding landing pages. So we ran A/B tests on headlines, product descriptions, and even CTA button colors which squeezed out another 0.2% CVR increase on our key product pages.

The Future of Media Buying

This whole cookieless transition forces you to build smarter, more durable ad programs that don’t depend on creepy user tracking. The TerraThreads campaign showed us that a solid foundation of first-party data, smart contextual plays, and genuinely good creative will beat old cookie-based tactics. The advertisers who are building these skills now are going to have a serious advantage as the rest of the industry plays catch-up. If you want to dig deeper into maximizing your return, you should check out Programmatic AI: 5 Steps to Maximize ROAS in 2026. What we learned about creative variety here is also directly applicable to the discussion around AI Ad Creativity. And it all helps to have a realistic view of the ROI reality of AI media buying to frame these big strategic decisions.