Automotive e-commerce · Houston · 18 months

Rebuilt the storefront. Wired the funnel. Compounded LTV.

Hellhorse Performance engaged Gray Reserve in Q1 2024 as a full-stack growth partner. By Q2 2025, revenue was 3.4× — driven by a rebuilt Shopify storefront, unified paid media, and an audience augmentation layer compounding customer LTV monthly.

3.4× Revenue · 18 months
$23 Blended CPL
4.1× ROAS

Scaling, plateauing, and losing the thread.

Hellhorse hit the ceiling every growth-stage e-commerce operator eventually hits: the engine that got them to $3M/year was preventing $10M. Traffic flat. CPL climbing. A three-year-old Shopify theme. Three contractors optimizing their own attribution, none talking.

Enhanced Conversions unwired. Meta CAPI misconfigured at roughly 41% of purchases. TikTok untouched. Klaviyo with thirteen flows, four delivering. Twelve apps duct-taped onto a paid template. Core Web Vitals failing mobile. No analytics layer synthesizing a number anyone could act on.

Repeat purchase unmeasured. LTV guessed at ~$280. No lookalike model, no audience augmentation, paid media spraying cold traffic. The shop was doing fine and should have been doing much better — without a clear read on what was broken.

What we built. What moved.

  1. 01

    Diagnostic (Days 1–14)

    Two-hour working session. Every live data source pulled. Fifteen-page plan naming four leverage points: attribution, storefront, audience augmentation, LTV economics.

  2. 02

    Attribution rebuild (Days 15–30)

    Meta CAPI server-side, Google Enhanced Conversions, TikTok Events API. Purchase events into every platform within 200ms. CPL on Google down 31% in fourteen days.

  3. 03

    Storefront rebuild (Days 30–75)

    Shopify 2.0 custom theme. Apps removed. CWV to 94 mobile / 99 desktop. Product page conversion +38% desktop, +27% mobile in month one.

  4. 04

    Audience augmentation (Days 60–120)

    Four reservoirs in Neon: buyers, intent, high-LTV lookalikes, premium segments — refreshed hourly into Google, Meta, TikTok.

  5. 05

    LTV economics (ongoing)

    Per-cohort, per-product, per-channel LTV. Bid ceilings modeled, not guessed. ROAS from 1.9× to 4.1× over nine months.

Eighteen months. One compounding number.

3.4× Revenue growth
12k Repeat buyers
Retainer renewed
Become the next case study. Begin Private Audit