Skip to main content

​
Media & Distribution

​
Why Platform Defaults Fail

Every major ad platform — Meta, Google, TikTok — provides audience-building tools built on its own behavioral data. These audiences optimize for in-platform metrics: clicks, conversions, ROAS as defined by last-click attribution. They do not optimize for contribution margin, cohort LTV, or repeat purchase rate. The result is ad spend that looks efficient on the dashboard and performs poorly on the P&L. Brands that optimize for platform ROAS acquire customers who convert once. Brands that optimize for cohort LTV acquire customers who compound. The acquisition cost may be the same. The business outcome is completely different. DAS builds media strategy on the customer intelligence layer, not platform defaults. The audiences fed into ad platforms are derived from the analysis of the actual customer file. The attribution framework is designed to measure what actually matters to the CFO, not what the platform reports.

​
Audience Architecture

The media audience architecture at DAS flows directly from the RFM segmentation and cohort analysis built in the intelligence phase. Customer match audiences are built from the Champion segment — the 15–20% of the customer file driving 50–70% of contribution margin. These customers are excluded from acquisition campaigns (you already have them) and used as the seed for lookalike and similarity models. Lookalike audiences built from Champion-segment seeds consistently outperform lookalikes built from all-purchaser seeds or pixel-based audiences. The signal is cleaner because the seed is more precisely defined. Suppression lists are built from segments that should not receive acquisition messaging: recent purchasers, Champions, and customers with high predicted LTV who are better served by retention flows than acquisition ads. Retargeting audiences are structured by behavioral tier — site visitors who have not purchased are different from lapsed customers who have, and both require different creative and offers. This architecture requires ongoing maintenance as the customer file evolves. It is not a one-time setup.

​
Attribution Testing

Platform attribution is broken for DTC brands. This is not a controversial claim — it is a mathematical reality. Last-click attribution misses:
  • Subscription LTV from customers acquired through a specific campaign
  • Halo effects on retail or wholesale channels
  • Multi-touch journeys that begin on one platform and convert on another
  • Word-of-mouth and organic amplification generated by high-LTV customer segments
DAS builds reporting frameworks that clients own inside their Klaviyo and Shopify infrastructure — not black-box dashboards that disappear when the engagement ends. These frameworks report on contribution margin by channel, 90-day repeat purchase rate by acquisition source, and cohort LTV by campaign period. The goal is to answer the question the CFO actually cares about: which marketing activity produces customers who are worth having?

​
Media Allocation

Media budget allocation at DAS follows the customer intelligence findings. Channels that produce high-LTV cohorts receive more investment. Channels that produce high-volume, low-LTV cohorts receive less. This sounds obvious. It is rarely done. Most media allocation decisions are made based on platform ROAS, which rewards the channels that claim attribution last — not the channels that actually build the customer relationships that compound. DAS uses cohort data to trace the actual performance of each channel over 90 days, 180 days, and 12 months — and allocates accordingly. The result is a media mix that looks different from what the platforms recommend. That is the point.