Case Study Liquid Rubber DIY · Gold Coast, AU · DTC + Trade

A waterproofing brand, re-engineered to scale.

Liquid Rubber DIY sells Australian-made waterproofing direct to homeowners and trade. We took a product people already trusted and built the growth engine around it — paid media across three channels, a custom AI storefront agent, lifecycle email, CRO, and the pricing and margin work underneath. In eight months, monthly revenue nearly tripled and orders more than tripled. Execution, not advice.

+0% Monthly revenue growth
peak against engagement start
+0% Order growth
135 → 427 per month
0× Conversion rate
roughly doubled and holding
$0 Meta cost per purchase
across 1,900 purchases

Nov 2025 – Jun 2026, recomputed from the client’s raw Shopify, Meta and Google exports and reconciled to source. Absolute revenue is withheld and shown as movement against an indexed baseline. Cost per purchase is an acquisition-cost metric. Efficiency is shown as blended MER, not platform ROAS — see results.

(01) — The brief

A good product, under-instrumented.

Genuine demand and a strong review base, with a thin growth layer on top: one generic ad account carrying everything, new channels running without conversion tracking, and no automated support underneath.

A / What we walked into

A Shopify store with a product customers rated highly and bought repeatedly — and almost none of the machinery a brand needs to scale that reliably.

A single undifferentiated ad setup did all the acquisition. New channel spend ran with no conversion tracking, so its return could not be judged. Support was entirely manual. And margin and freight had never been modelled by region, so nobody knew which orders actually made money.

B / What we set out to build

Not a campaign — an engine. Disciplined paid media across Meta, Google and the emerging ChatGPT channel, sitting on top of infrastructure built to compound: a production AI storefront agent, connected Klaviyo flows, a hardened Shopify theme, and a regional margin model to protect profit as volume grew.

The test: could a lean brand run three ad channels, always-on AI support and a real CRO layer without adding headcount? That is the whole thesis — marketing that performs, on systems that scale it.

(02) — What we did

Six systems, one operator.

The range a single client can demand — performance media through to full-stack engineering — delivered by one team.

01 Performance media

Meta, Google & ChatGPT Ads

Three channels on a disciplined creative-variation system, rebuilt from raw platform data with Shopify as the revenue source of truth.

1,900 Meta buys @ $12 · Sealant 5.96× PMax ROAS
02 AI engineering

Custom storefront agent

A production shopping and support assistant, live on the storefront — answering product questions, sizing the job, recommending the right bundle and capturing trade leads.

Always-on selling and support
03 CRO & front-end

Shopify theme engineering

Rebuilt the coverage calculator, fixed the quote form so enquiries stopped being lost, and brought tap targets and contrast up to accessibility standard.

Underwrites the conversion-rate gain
04 Lifecycle

Klaviyo email flows

Connected Klaviyo and built sample-to-purchase follow-up flows across the core products — turning one-time buyers into repeat ones.

Returning customers up 84% in June
05 Margin intelligence

Regional P&L & pricing

Modelled freight and margin by state, drove an approved price increase, and mapped an eParcel freight-saving opportunity.

Price rise approved · margin modelled
06 Channel expansion

Applicator & trade arm

Built an installer network with a quote-routing system, and stood up the trade arm with its own AI agent and help centre — opening commercial revenue alongside retail.

New trade channel opened
(03) — Build spotlight

The store that answers back.

Flagship engineering

A production AI agent, not a chat widget.

We forked Shopify’s reference agent and rebuilt it into a live shopping and support assistant. It works the way a good salesperson does: asks what the job is, calculates how much product it needs, recommends the right bundle, answers the questions that stall a purchase, and adds it to the cart.

When a visitor turns out to be trade rather than retail, it captures the lead and hands off to a human. It runs every hour of every day, and it was hardened against prompt injection and abuse before it went live.

(04) — Results

Instrumented, multi-channel, and compounding.

Monthly revenue, indexed · Oct 2025 – Jun 2026

Monthly revenue, nearly tripled in eight months.

As the store moved to three managed channels, an always-on AI agent, connected email and a hardened storefront, revenue and order volume climbed together and held. June closed on 427 orders, the highest order volume on record, and +12.6% revenue against June the previous year on flat traffic — growth coming from conversion, not from buying more visitors.

5.96× Flagship Sealant
Google PMax ROAS
~4× Blended MER, store-wide
not platform ROAS
$0 Meta cost per purchase
across 1,900 orders

Recomputed from raw Shopify, Meta and Google exports (July 2026 analysis, all totals reconciled to source). Absolute revenue and spend are withheld: the chart is indexed to the engagement start at 100, and efficiency is expressed as ratios plus cost per acquisition. MER is store revenue divided by total ad spend across the overlapping reporting window — a blended measure that includes organic and returning-customer revenue, not platform-attributed ROAS: neither ad export carried a conversion-value column, so true channel ROAS is not calculable. Meta purchases are pixel-attributed on a 7-day click window; the Sealant figure is Google-reported ROAS at product level.

(06) — Start a conversation Let’s build your engine.
BasedGold Coast, Australia
WorkingGlobally · a handful of clients at a time