How we rebuilt the storefront,
wired the funnel, and compounded LTV.
Hellhorse Performance, a Houston-based automotive performance parts operator, 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 across Google, Meta, and TikTok, and an audience augmentation layer compounding customer LTV monthly.
Scaling, plateauing, and losing the thread.
Hellhorse Performance came to Gray Reserve in Q1 2024 with the problem every growth-stage e-commerce operator eventually hits: the engine that got them to $3M/year was actively preventing them from getting to $10M. Traffic was flat month-over-month. CPL was climbing on every paid channel. The storefront was three years old, built by a freelance Shopify developer who had since gone silent. And the growth team was three contractors running Google Ads, Meta Ads, and email independently of each other — each optimizing their own attribution, none talking.
The technical gaps were measurable. Google's Enhanced Conversions wasn't wired. Meta CAPI was installed but misconfigured, reporting roughly 41% of actual purchases. TikTok had never been touched. Klaviyo had thirteen flows, of which only four were delivering. The Shopify theme was a paid template with twelve third-party apps duct-taping around missing functionality. Total site Core Web Vitals: failing on mobile, barely passing on desktop. No analytics layer was synthesizing any of this into a number anyone could act on.
The business gaps were worse. Repeat purchase rate had no measurement. LTV wasn't calculated — just “we think it's about $280.” There was no lookalike model, no audience augmentation, no view into which customer segments were actually profitable. Paid media was running flat against the entire addressable market instead of ranked audiences. In the founder's words on the briefing call: “we're doing fine. We should be doing much better. I just don't know what's wrong.”
Rebuild the stack. Wire the funnel. Compound the LTV.
- 01
Diagnostic (Days 1–14)
Two-hour working session, not a kick-off call. Gray Reserve pulled every live data source — Shopify, GA4, Google Ads, Meta Business Manager, Klaviyo, and two years of order history — and returned a fifteen-page written plan in the second week. The plan named the top four leverage points: attribution rebuild, storefront rebuild, audience augmentation layer, and LTV economic modeling.
- 02
Attribution rebuild (Days 15–30)
Meta CAPI reinstalled server-side through the Shopify event layer. Google Enhanced Conversions wired. TikTok Events API deployed for the first time in company history. Every purchase event now fires into every platform within 200ms. Paid media teams gained 2.3× the conversion data overnight, and CPL on Google dropped 31% inside fourteen days.
- 03
Storefront rebuild (Days 30–75)
Shopify 2.0 migration. Custom theme built on Dawn with progressive enhancement. Twelve third-party apps removed. Core Web Vitals moved from failing to 94 mobile / 99 desktop. Product page conversion rate on desktop jumped 38% in the first month post-launch, 27% on mobile. Site search, facet filtering, and the checkout flow rebuilt from scratch.
- 04
Audience augmentation (Days 60–120)
Four reservoirs built in Neon Postgres: past buyers (90d rolling), intent signals from browsing + cart behavior, high-LTV lookalikes derived from the top revenue decile, and premium-segment custom audiences. Refreshed hourly, pushed to Google + Meta + TikTok via CAPI. Paid media stopped spraying against cold traffic and started ranking against proprietary intent.
- 05
LTV economics (Days 90–ongoing)
The LTV model was the unlock. Calculated per-cohort, per-product, per-acquisition-channel, refreshed monthly. Gray Reserve could now tell Hellhorse exactly how much a Q1 Google Ads customer was worth at month 6, month 12, and month 18 — and therefore exactly how much to spend to acquire one. Bid ceilings moved from guessed to modeled. ROAS moved from 1.9× to 4.1× over nine months.
- ✓ Diagnostic Days 1–14
- ✓ Attribution rebuild Days 15–30
- ✓ Storefront rebuild Days 30–75
- ✓ Audience augmentation Days 60–120
- ✓ LTV economics Days 90–ongoing
- Compound retainer Month 6 → now
Eighteen months of compounding.
Top-line revenue from Q1 2024 baseline to Q2 2025. Compounded, not spiked.
Original 6-month contract renewed twice. Slack channel still hot daily.
Past buyers, intent signals, high-LTV lookalikes, premium lookalikes. Hourly refresh.
Blended return-on-ad-spend at month 18, modeled against customer LTV per cohort.
Gray Reserve didn't give us a deck or a checklist. They took the wheel. Eighteen months later we've shipped more infrastructure than the previous three years put together, and I'm not paying five different vendors to blame each other when a number moves. One retainer. One number. One partner.