Glowout Salon
Chicago's premier spray tan brand, rebuilt as a living system.
Visit Live SiteOverview
A 19-year-old Chicago spray tan brand with 8 locations across 4 states needed a site that could compete with national chains on SEO, convert paid traffic, and run its own bookings without the team touching anything.
Scope of work
- Custom Next.js SmartSite
- Live "next available" booking API (Square integration)
- AI chat widget (GHL-trained on brand voice, hours, services)
- Blog system with Supabase CMS + admin dashboard
- Location-specific landing pages (neighborhood × service)
- Full schema markup — Organization, LocalBusiness, FAQ, HowTo
The Challenge
What we walked into.
The old site was a Squarespace brochure — no SEO, no schema, booking tucked behind three clicks. Paid ads were converting at ~2% because landing pages were generic. Staff spent hours fielding "what time do you open?" messages that should have been self-serve.
The Approach
How we solved it.
Rebuilt the entire stack on Next.js 14. Wrote custom Square webhook integration to pull live availability and surface a "next available" slot in the hero — visible before you scroll. Deployed Sunny, an AI chat agent trained on every FAQ, location detail, and service offering. Added structured data across every page so Google, ChatGPT, and Perplexity can answer questions about Glowout correctly. Built neighborhood-specific landing pages for paid traffic.
Outcomes
What shipped.
4.9★ · 190 reviews
West Loop GMB aggregate rating now surfaces in schema
~2s load
Page speed score 95+ on mobile
Live availability
"Next available" slot pulls live from Square API
24/7 AI chat
Sunny answers FAQs, hours, services around the clock
Highlights
- First SmartSite with a live booking API surfaced in the hero
- Schema deployed across every page — Organization, LocalBusiness, FAQ, HowTo, Event
- Admin-editable blog with full draft / publish workflow
- 3D particles on the Partners page using React Three Fiber
- Mobile CRO push (May 2026): hamburger nav, persistent BOOK NOW pill, press logos strip, Google Reviews badge
Technical stack
Reference Deployment · L0 → L5
From the chat window to a compressed operator stack.
PCC deployed Glowout up the five-level adoption ladder Anthropic published in April 2026 — and then past it. Each stage below maps to a specific level on the ladder. Read the framework essay for the full thesis.
L1 · Build
Months 1–2
A SmartSite deployed against Glowout's actual stack — Square Bookings, Gmail, Drive, Slack. The first deliverables were unglamorous and high-leverage: a weekly booking report with same-day cancellations and per-technician utilization, a monthly P&L narrative written driver-first (the way Jenn reads numbers), and DM/review response drafts in brand voice. No new tools for staff to learn — all output landed in the existing Drive folder structure.
L2 · Encode
Months 3–4
Workflows got encoded as skills. /retail-attach-prompt — a single-question consultative pattern with role-play scripts and objection handlers. /glowout-sop — the core training SOP, identical across all locations, which resolved the dilution problem at the source. /press-and-glow — the activation framework for the Press + Glow co-brand with Glamoré, encoded so the outreach pattern is repeatable. Each skill was built by pointing Claude at existing memos, training transcripts, and Slack threads. No new documentation required.
L3 · Automate
Months 5–6
Skills got scheduled and bundled into recurring agents. Monday 7:30 AM: weekly operator brief from Square, GHL, QuickBooks — bookings, retention deltas, retail attach by location, top three things Jenn should know before her day starts. Friday 4 PM: retail attach report by technician with coaching prompts surfaced for locations below benchmark. Continuous: review and DM monitoring with draft responses queued for approval. Monthly: vendor and partnership pipeline review.
L4 · Department
Months 7–9
Scheduled agents and encoded skills got bundled into full functional plugin packs. Operations plugin: SOPs, training, scheduling, multi-location consistency layer. Marketing plugin: voice skill, Canva integration, scheduled IG/email distribution, asset library tagged and live. Finance plugin: QuickBooks + Square integration, scheduled P&L, cash position, retail attach. Customer service plugin: iMessage + Gmail + Slack monitoring with drafted responses queued for approval.
L5 · Operator
Field State
What emerged past L4 is something Anthropic's published model doesn't name yet. Jenn's laptop is now a dedicated operator machine. Claude Code and Claude Cowork run simultaneously, with full connector authentication to GHL, Square, QuickBooks, Canva, Vercel, GitHub, iMessage, Gmail, Slack, Drive, and a growing list of others. Every function's plugin bundle is stacked into the same session — context carries between marketing, ops, finance, training, and CS without re-explaining. She doesn't open seven apps to run her business. She opens one operator session.
The Field State
This is L5 — the Compressed Operator Stack. It's what L4 becomes when the business doesn't have departments to distribute across — when all eight functions get stacked into one operator instead. It's the owner-operated endgame, and it's what PCC has been operating in production at Glowout for nine months.
What this proves
- 01Owner-operated businesses don't need enterprise infrastructure to climb the adoption ladder — and they go past it. Square + GHL + QuickBooks + Canva + Gmail is enough of a stack if someone operates it correctly.
- 02The operator role is real and external. Glowout did not hire an internal champion, run a six-month pilot, or appoint a department head. PCC ran all four published levels — and built the L5 state — on Glowout's behalf.
- 03Anthropic's framework hasn't caught up to the field yet. Their published model stops at L4 because their reference customer is enterprise. PCC has been operating L5 in production for nine months. The framework will follow.
08 / Your project
Want a build like this?
Every SmartSite starts with a conversation. Tell us what you're building.