Performance Media That Trusts Its Own Numbers
Paid acquisition across Meta, Google, TikTok, and applicable channels — bid against true cohort value, with server-side tracking and Conversions APIs that survive iOS, ITP, and ad-blockers.
Industry · E-commerce
Marketing for e-commerce brands operating in the post-iOS attribution reality.
iOS 14.5 broke the attribution model most e-commerce brands were running their whole P&L against. Meta and Google still claim conversions the cohort math can't defend. Acquisition cost keeps climbing while reported ROAS stays improbably stable. Most agencies aren't equipped for any of this. We are — and we run the marketing operations that finally tell the truth, so spending decisions can make sense again.
The marketing reality
Why most agencies underperform for e-commerce now.
Reported ROAS no longer matches the bank account.
Since iOS 14.5, the gap between Meta-reported ROAS and revenue you actually book has widened materially for nearly every direct-to-consumer brand. Your team knows it's happening but can't quantify it precisely enough to act — which means budget keeps allocating against a metric that doesn't describe reality. The fix isn't to spend more; it's to rebuild measurement.
Cohort LTV is the only honest number, and almost nobody bids against it.
First-purchase ROAS and 60-day ROAS look very different across channels. Most teams bid against the first one — because it's the one Meta reports — and over-invest in channels that bring one-and-done buyers. The brands compounding right now are bidding against predicted 12- or 24-month LTV, modeled on first-party data. We build that capability.
Retention is the actual growth lever now.
When new-customer CAC rises faster than AOV, repeat rate and LTV become the lever. The brands winning are the ones treating lifecycle, retention, and post-purchase experience as growth investments — not as afterthoughts to acquisition. Most agencies still treat retention as Klaviyo configuration. It's much more than that.
What we do
Six services we deliver for e-commerce.
Cohort LTV & Predictive Bidding
12-month and 24-month LTV models trained on your real order data. Deployed back into Meta and Google as value-based optimization. Bidding moves from first-purchase ROAS to predicted cohort value — which fundamentally changes which customers your campaigns target.
Performance Creative at Velocity
Creative production at the volume modern testing actually requires — dozens of variants weekly, not monthly. AI-assisted UGC variation, platform-format optimization, and statistical-rigor variant testing.
Lifecycle, Retention & Subscription
Klaviyo, Postscript, Yotpo, and Recharge workflows built around real cohort behavior. Predictive churn segmentation, win-back, subscription where it fits. Retention treated as a growth lever, not as cleanup work.
Storefront & Conversion Optimization
Shopify, Shopify Plus, BigCommerce, or headless commerce engineering. PDP and checkout-flow optimization. Performance work so Core Web Vitals stop costing conversion. CRO tied to real revenue, not just session-level proxies.
Incrementality & Honest Reporting
Geo-lift testing, holdout designs, and marketing-mix modeling against real revenue. The math layer that finally answers what's genuinely incremental versus what's taking credit for organic demand.
What changes when we run your marketing
ROAS reports match the bank.
The gap between platform-reported revenue and actual revenue closes. You stop scaling channels that look profitable in dashboards but aren't in cash. Spend reallocates toward what's genuinely incremental.
Acquisition cost compounds down because LTV bidding is honest.
Bidding against predicted 12-month value instead of first-purchase value fundamentally changes which customers your campaigns target. Over a few quarters, the customer file tilts toward higher-LTV cohorts and blended CAC starts moving the right way.
Retention produces real revenue, not just open rates.
Lifecycle stops being a separate department and starts producing measurable second-purchase, third-purchase, and subscription revenue you can attribute. Retention becomes a number on the P&L conversation, not just a Klaviyo dashboard.
Tools & platforms
The e-commerce stack we work with.
We're platform-agnostic by design — we work with the tools your team already runs, and add only what's missing. The shortlist below is the stack we deploy most often for e-commerce engagements.
Storefront platforms: Shopify, Shopify Plus, BigCommerce, headless commerce on Next.js + Shopify Hydrogen.
Server-side tracking: Stape, server-side GTM, Snowplow — with deduplication and identity-resolution layers.
Lifecycle & retention: Klaviyo, Postscript, Customer.io, Yotpo, Loop Returns, Recharge — orchestrated against warehouse data.
Reviews & UGC: Okendo, Yotpo, Junip — joined to cohort retention modeling.
Data warehouse: BigQuery, Snowflake, Postgres — joining Shopify, payments, lifecycle, support, and returns.
BI & dashboards: Looker, Metabase, Hex — for true cohort, channel, and SKU-level economics.
Creative pipeline: AI-assisted production with structured testing variants and platform-format automation.
How engagements start
Three common ways e-commerce brands start with us.
Every engagement begins with a Discovery Audit — a six-week fixed-scope diagnostic of your current marketing operation. From there, e-commerce clients usually move into one of three paths, depending on where the biggest constraint is.
Path 01
We rebuild tracking, CAPI, and LTV bidding first.
The most common entry. Server-side tracking, Conversions APIs across Meta, Google, and TikTok, warehouse-grade attribution, and predictive LTV deployed back into bidding. 12 to 16 weeks, then operational handoff or retainer.
Path 02
We rebuild retention & lifecycle.
When acquisition is functional but repeat rate, AOV, or LTV is the actual constraint. We engineer Klaviyo, Postscript, and post-purchase workflows against unified warehouse data, plus the cohort and segmentation models that make lifecycle work as a math problem.
Path 03
We run the full e-commerce marketing operation.
For brands that need media, creative, engineering, analytics, and retention run as one continuous operation. Continuous creative pipeline, weekly experimentation, full warehouse and attribution ownership. Designed for brands above $10M revenue that can't staff this internally fast enough.
Get started
Get a marketing operation built for e-commerce.
Discovery Audit looks at your full e-commerce stack — storefront, tracking, CAPI, lifecycle, retention, and the warehouse layer behind them — and returns a clear roadmap. Six weeks, fixed scope, your document to keep regardless of next steps.