Where it started
The founder of an LA-based premium kitchen-appliance brand reached out in April 2025. Forty years of factory expertise behind them, ETL/CSA-certified product, a real product line ready to go — and an August 1st live-operations deadline that came from a hard commitment they'd already made to dealers and channel partners.
What they had on the technical side: a Shopify store with a theme installed and not much else. No B2B wholesale pricing wired up. No inventory management. No serial-number tracking. No SOPs. No team yet.
What they wanted to call us was: "a Shopify guy." What they actually needed was someone willing to architect the operational spine of a small business that wasn't yet a business.
What we found by listening
Half a day on a Zoom with the founder, then another half-day mapping the actual operational flow they had in their heads. The picture that emerged:
- The "Shopify build" was the smallest piece. The real work was end-to-end: inventory intake from the warehouse → barcode/serial assignment → order capture (online + admin-built draft for dealers) → fulfillment → tax handling → SOPs documenting all of it so a team could run it without the founder.
- Dealer pricing wasn't a feature toggle. It was a tier system, with volume discounts, that needed to live in third-party apps integrated with Shopify and not break when an admin built a draft order.
- The serial-number tracking the founder kept describing wasn't an ERP feature — it was a fulfillment constraint (premium appliances ship with serials that need to be captured per-unit at the warehouse, then preserved through the order record forever).
- Most critically: nobody on their team yet knew how any of this would work. The build had to ship with playbooks the team could actually follow, not a how-it-works deck that would gather dust.
We didn't quote a Shopify build. We quoted a phased operational launch.
What we built — phased on purpose
Three phases, scoped before we signed, priced once. Each phase ended with SOPs that matched the live system, not the planned one.
Phase 1 — Core launch (May → August 1, 2025). Shopify front-end and back-end. Wholesale pricing tiers and volume discounts via vetted third-party apps. Inventory management software integrated with Shopify. Manual serial-number tracking via order metafields (deliberately not over-engineered — barcode scanning came later). Shopify Flow automations for both ops (order routing, exception alerts) and marketing (cart events, dealer notifications). Multi-channel payment handling — Shopify checkout for D2C, invoice flow for the larger dealer orders.
Phase 2 — Barcode infrastructure (August 2025). Once live operations were stable, we layered in the barcode system: research, selection, labeling workflow, scanning workflow, integration with the inventory ledger. SOPs updated to match.
Phase 3 — Serial registration (September 2025). Final layer. Customer-facing serial registration, audit and trace processes for warranty service, integration into fulfillment.
At every phase: a one-on-one walkthrough with the founder where we sat together and used the live system end-to-end before declaring it done. The SOPs were rewritten after that walkthrough, not before — because what gets written into a doc before the system runs is fiction.
The 12-month follow-up — a real audit, not a sales pitch
May 2026: a year after launch, the founder came back. The business was generating real revenue — north of $400K in dealer revenue tracked through the infrastructure we'd built, across 350+ orders and 80+ active dealer accounts spanning 12 states. The top dealer alone had run 30+ orders and $25K+ in lifetime spend through the new system. They wanted to know what wasn't working.
We did a four-phase audit nobody asked for and the brand needed:
- Baseline data pull. Shopify Admin API across catalog, customers, discounts, theme, scripts. 32 raw data dumps cross-referenced. (Even Shopify's 60-day order-history scope limit was itself a finding — flagged with a fix path.)
- Page-level walkthrough. 17 URLs × mobile + desktop in Playwright, 34 screenshots, Lighthouse scored against a 10-dimension rubric. We learned the dealer signup page — the literal front door of the business — was the lowest-scoring page on the site (mobile LCP 13.7s, no H1, no value proposition, rendered nearly empty above the fold on cellular).
- Strategic gap analysis. Brand positioning, competitive landscape (every direct mid-market competitor benchmarked), customer-voice mining from forums (Houzz, Reddit, Quora) for the language real buyers use when shopping the category.
- A 95-finding RICE-prioritized backlog, sequenced into a 6-month execution plan: foundation (week 1), quick wins (30 days), build tier (60 days), strategic bets (90–180 days).
The most uncomfortable finding: the brand was priced 30%+ above its closest D2C peers, with the heritage story that justified the premium buried in the About copy nobody was reading. The fix wasn't to drop price — it was to surface 40 years of factory expertise in the hero, the PDPs, and every dealer-facing surface.
The outcome
- Live ops shipped on the August 1, 2025 deadline. Phase 1 deliverables stood up the back-end of a real business — not a launch-day demo.
- Three phases delivered as scoped. Barcode in August, serial registration in September. Each phase ended with handoff SOPs the team could actually use.
- $400K+ in dealer revenue through year one. 350+ orders, 80+ dealer accounts, 12 states. None of it would have been operationally possible on the founder's pre-launch setup.
- Zero data-integrity incidents. Inventory ledger never disagreed with itself. Serial numbers attached to every shipment, recoverable at any time.
- A post-launch audit that turned into 6 more months of work. 95 RICE-scored findings, sequenced into a 30/60/90/180-day plan. Including the things we got wrong on launch day (the warranty PDP being public-listed; H1s missing on every PDP; the dealer signup page being the slowest page on the site) — flagged honestly and prioritized.
- A brand whose founder still calls. A year after launch, when the question is "what's broken now," we're the call they make.
"We thought we needed a Shopify guy. What we got was someone who built the back-end of our business and then came back a year later to tell us what was still wrong with it. Including the stuff they shipped. That's the part nobody else does." — Founder, the LA premium kitchen-appliance brand
Why it worked
Three things.
The first was that we refused to scope the engagement as "a Shopify build" when the actual problem was operational architecture. Saying yes to a smaller scope would have shipped a website that the team couldn't run.
The second was phasing on purpose. Phase 1 deliberately under-engineered serial tracking (manual via metafields) so we could ship the launch on time. Phases 2 and 3 layered in the heavier infrastructure once the team was running real operations and could feel where the gaps were. Trying to ship all of it at once would have missed the deadline.
The third was that we showed up a year later. Most of this category disappears after launch — the agency cashes the check, the founder is left holding a Shopify store and a stack of half-applicable SOPs. We came back, audited everything we'd built (and everything we'd missed), and turned out work for the next six months. None of it was urgent. All of it was the next set of right things to do.
A scoped, milestone-based engagement for the launch — every deliverable defined and priced before any work began. A separate scoped audit a year later. No retainers. No hourly creep. Both engagements ended with a clear handoff to the team, with everything we built living on their infrastructure, owned by them.