Case study
Lifestyle brand, marketplace MVP, and ecommerce rebuild
New brand, marketplace, and headless commerce at scale
Sleepme Inc. ($50M+ revenue)
- Ecommerce
- Headless
- Shopify
- Growth
- Rebrand
- Marketplace
- ML
- Connected product
Summary
Role
Digital Product Lead
Timeline
Nov 2020 – Jan 2022
Team
Product, design, engineering, architecture, growth, marketing, hardware, and key vendors.
Domain
DTC ecommerce, lifestyle brand, hardware, partner marketplace, connected sleep insights.
Metrics moved
15% revenue growth during the global supply chain shortage.
Key constraints
Tight inventory; rebrand/URL migration risk; marketplace trust and partner onboarding on a fixed calendar.
Systems involved
Vue + Strapi; Shopify Plus checkout; RN for post-purchase/device; ML summaries for logged-in sleep data; curated marketplace.
Stack
Tech stack
Headless experience on commerce primitives; mobile parity; bounded ML on device-backed data.
Storefront & content
- Vue.js
- Strapi (headless CMS)
- Componentized PDP & marketplace modules
Commerce
- Shopify Plus
- Checkout integration
- Merchandising & pricing experiments
Mobile
- React Native
- Post-purchase / device flows aligned to web
Data & ML
Before chat UI was everywhere: short summaries grounded in what the devices actually measured.
- Authenticated sleep dashboard (web)
- Device telemetry
- In-house ML summaries
How the headless storefront carried the rebrand
Strapi for editorial velocity, Shopify Plus at checkout, RN where parity mattered.
Evidence-led prioritization
Legacy analytics and feedback chose PDP modules, marketplace surfaces, configurators, and checkout adjacencies under supply and the summer deadline.
Composable pages in Strapi
Marketing iterated landing/PDP blocks without monolithic theme releases; eng held schema and performance guardrails.
Shopify Plus at checkout
Cart, discounts, checkout on proven primitives—rebuild did not become a payments science project.
React Native in lockstep
Device setup and account claims matched web so post-purchase did not feel like a different company.
Authenticated sleep insights
Hardware metrics + ML summaries for logged-in owners—MVP scope and guardrails before chat LLMs were the default pattern.
Storefront
Sleep.me / Chilipad era: conversion-focused PDP with configurators (screenshot below).
Screenshot 1 of 1
Overview
Digital product lead for first Sleep.me site and marketplace through ChiliSleep→Sleepme rebrand and full storefront rebuild: lifestyle brand, partner products, not only hardware. Legacy analytics and feedback set MVP priorities. Shipped in under five months for summer peak. Owned prioritization, analytics/growth (instrumentation, funnels, leadership reporting), and commerce/analytics vendors. With eng/architecture: buy-vs-build across the stack. Logged-in MVP: sleep charts + short ML summaries—tight guardrails, pre–default-LLM pattern. One migration calendar aligned design, eng, growth, and hardware.
Business context
Premium sleep hardware + connected software via DTC; performance and organic search were existential. Supply shortage made inventory-led merchandising volatile. Leadership needed faster tests, richer PDPs, editorial velocity outside a monolithic theme, and clean handoff to Shopify Plus checkout.
Problem
Legacy funnel leaked on mobile; PDPs under-supported proof and configuration; pricing/bundles under-instrumented under constrained stock. Rebrand/URL risk to SEO. Needed credible PDP→cart→checkout while RN held device/account expectations. Hardware owners lacked one trustworthy web home for night-over-night data and plain-language interpretation.
Constraints
Global shortage on hero SKUs; finance margin discipline; SEO equity through content/URL migration; marketplace partner policies on top of commerce QA.
Role & ownership
Owned marketplace narrative: discovery → backlog → acceptance → stakeholders; analytics/growth strategy; vendor selection and contracts for analytics/commerce. With architecture/eng: buy-vs-build and TCO. For sleep dashboard: brokered scope across Hardware (telemetry, latency), ML (metrics, tone/claims guardrails), and web (auth, empty/partial states, performance on history). Funnel data drove PDP, configurator, checkout adjacency, and app handoff.
Goals & metrics
15% revenue in shortage; retention/satisfaction from insight-led hardware+software; post-migration search without revenue cliffs; marketplace GMV as a new channel.
Approach
Rebuild + rebrand = one program, shared metrics. ChiliSleep data sequenced marketplace and lifestyle surfaces against the summer deadline. Revenue-elastic pages first, then brand, then long-tail SEO—with obsessive redirect/parity on top URLs.
Decisions & tradeoffs
Strapi + Shopify Plus: own experience layer; buy proven checkout/catalog. Sleep dashboard: narrow authenticated MVP with hardware-backed contracts vs a full analytics product that would miss the rebrand window; ML with tone/factual guardrails. Deferred noncritical CMS flexibility for peak readiness; smaller marketplace launch for fulfillment SLAs and review quality over splashy SKU count.
Cross-functional leadership
Single program rhythm: rebrand assets, CMS migration, checkout, analytics validation, vendors. Marketing: landing/PDP modules match campaigns. Hardware: specs, support, pairing, telemetry for dashboard. ML: summary quality and review loops. Eng: spikes on checkout, Strapi, app sync, dashboard APIs. Screen recordings of homepage and PDP→checkout as async “source of truth” when staging access was uneven.
Execution
Headless: Vue + Strapi, Shopify Plus cart/checkout, coordinated RN where device/order context had to match web. PDP blocks (proof, financing, HSA/FSA, trials) and pricing experiments from legacy data; ML surfaces only where they moved conversion. Logged-in web dashboard: metrics shaped for comprehension + bounded ML summaries (pre–ChatGPT default). Rebrand with redirects and parity on high-value URLs; marketplace with curated sellers for trust.
Outcomes
15% revenue lift in the constrained period; marketplace beyond first-party catalog; search recovered/improved through cutover. Dashboard gave owners a credible on-site “what happened last night” link to ongoing product value.
What changed
Faster, clearer path to purchase; merchandising could iterate without monolithic theme releases. Owners validated hardware investment on the same domain they bought from. Internal teams one roadmap; marketplace a lever beyond owned inventory.
Lessons learned
Volatile supply: pricing and narrative discipline beat discounting. Instrument before the creative reveal so launch-day numbers are trusted. Full-journey screen captures helped async alignment. ML on owned data: Hardware, ML voice, and web UX must agree in one sentence—or “AI summaries” become a support problem.