LandersRX How we actually work
How we actually work

LandersRX Playbook

Operating manual for everyone who runs LandersRX work. Written for a human strategist (a new hire who's never seen the system) and for Claude (the AI assistant that helps execute it).

If you're new to this: read sections 1 through 6 in order. Then skim the rest as reference. If you're Claude: this is the canonical map of the LandersRX operation. Treat it as the source of truth when a session lacks specific instructions. Defer to skill-specific SKILL.md files for execution detail.

1. What LandersRX Is

LandersRX builds high-converting matched landing pages for ecommerce brands, derived from the brand's own ad data. The thesis: a landing page isn't a website, it's the second half of the ad. When someone clicks an ad, the page must pick up exactly where the ad left off (same message, same promise, same hook).

LandersRX is a sister company of WRK (Brad's performance marketing agency). WRK uses LandersRX systems to deliver landing pages for agency clients, but LandersRX has its own clients and its own pricing.

The service has three legs: - Audit: 90-day Motion-driven account audit identifying the lanes that justify a landing page. - Roadmap: 5 page concepts with explicit page types, ranked by combined leverage (spend signal first). - Build: Design in Figma, approve, build in Replo, launch live, monitor.

2. Who Runs This and What Each Role Does

The strategist (human)

Claude (AI assistant)

The split, in practice

3. The Offerings

Pricing tiers (per the lead deck sales/decks/landersrx-lead-deck.html):

Tier Pages Price Guarantee Includes
BUNDLE (recommended — default sell) 3 pages $4,000/mo flat ROI Guarantee 3 matched landing pages, Pillars/Angles/Personas report, Replo build + launch
SCALE 3 pages + 30 statics $5,500/mo flat ROI Guarantee Bundle + 30 static ads matched to pages

Flat monthly pricing. No first-month discount. Confidence-led, not bargain-priced.

The hidden Starter (rep eyes only)

$1,500/mo for 1 page, no ROI Guarantee. Never advertised, never on the deck or the LP. Salesperson-only downsell, used when getting hard price pushback ("let me see what I can do — I can give you $1,500 for one page, but you also lose the guarantee"). Per-page price is intentionally higher than Bundle's per-page price to push everyone into the 3-page tier.

The ROI Guarantee

If the pages we ship aren't ROI-positive in the client's ad account, we deliver another month of pages — free. Same number, same scope, no questions asked. Applies to Bundle and Scale only. Replaces refunds: we can't give money back (commission already paid) but we extend the work.

Pricing discipline

We don't discount the Bundle to close. The two real moves: 1. Hold the line at $4K and lean on the ROI Guarantee + the calculator math. 2. Drop them to Starter ($1,500, no guarantee) — a downsell, not a discount.

Ideal-fit signals

4. The Sales Process

Full internal-rep notes are in sales/sales-notes.md. The short version:

Pre-call

  1. Build a free landing page for the prospect using only public info (their site, their ads, their product). This is the hook. No access needed.
  2. Drop the free-page render into slide 12 (Your Page placeholder) of landersrx-lead-deck.html.
  3. Save the deck as <prospect-name>-call.pdf. Don't overwrite the master.
  4. Skim their site so you can name-drop their actual products on the call.

The call (~30 min, 5 phases)

  1. Discover (5-10 min). Drive with the four core questions: - Tell me about the product. - How much are you spending on ads? - Where are you spending it? - Have you ever tried landing pages before?
  2. Position (5 min). Walk the deck. Who we are, the problem ("second half of the ad"), the process.
  3. Show (10 min). Walk the page on slide 12. This is the hook. Don't narrate every section. Let them react. Pause after each scroll.
  4. Price (5 min). Walk the tiers. Name the tier you recommend before they ask. Don't apologize for the price.
  5. Close (5 min). Confirm the next step. Send the agreement same-day if at all possible.

Common objections (full scripts in sales-notes.md)

Same-day decision incentives

Use sparingly. Pick one, never stack: - Free static ad pack with Bundle (~9-10 statics matched to first-month pages, normally a Scale-tier inclusion) - Audit kickstart (P/A/P report by tomorrow morning)

5. Client Onboarding

Access checklist (required day-one)

Ideal access (not blockers)

Brand kit

Before any page work begins, build a brand kit. It's the source of truth for every page produced for that client.

Kickoff deliverables to the client

  1. Welcome email with what's expected from each side
  2. Brand kit (link to Figma or markdown file)
  3. Calendar slot for the first audit + roadmap review
  4. Access requests submitted (Meta, Replo, Clarity)

6. The Delivery Workflow

The end-to-end flow from access granted to first pages live.

Step 1: Audit + Roadmap (week 1)

Invoke the landersrx-account-audit skill. Trigger phrases: "run the LandersRX audit on [client]", "build a page roadmap for [client]", "audit + roadmap".

Output: a single LandersRX-branded HTML file (and matching PDF) with: - §1 Account Snapshot (90-day KPIs, spend-state distribution) - §2 Messaging Angles (spend-first bar chart) - §3 Audience Cuts (persona cards + audience × angle matrix with primary/efficient/soft tiers) - §4 Hook & Format diagnostic - §5 Top creatives with thumbnails - §6 Opportunity Map (Dominant / Underbuilt / Emerging / Deprioritize) - §7 Page Roadmap Inputs (5 lanes that justify a page) - §8 Recommended Page Concepts (5 pages with explicit page types) - §9 What Happens Next (Round 1 workflow + Round 2 watchlist)

The strategist reviews the audit, dials in the page picks (the data ranks them, the strategist makes the call), and gets client approval.

Step 2: Design (week 2, days 1-3)

For each approved page concept: 1. Invoke the landing-page-creator skill. Trigger: "create a landing page for [client] [page-type]" or "build the [page-type] for [client]". 2. The skill reads the brand kit, the audit, and the relevant DTC library reference. 3. Output: an HTML draft + a rationale document. 4. Capture the HTML render to Figma (using the embedded capture script, see workspace CLAUDE.md). 5. Strategist reviews, marks up in Figma. 6. Iterate same-day.

Step 3: Approval (week 2, days 4-5)

Client reviews the Figma file. Markup, comments, async or call. Iterate until approved.

Step 4: Build (week 2, days 5-7)

Once approved, build in Replo. QA the page on the actual store.

Step 5: Launch (week 2, day 7+)

Pages go live. Each page gets 7 to 14 days of spend, with weekly check-ins for early signals and directional reads.

Step 6: Monitor (week 1 of spend)

Step 7: Game Plan (weeks 3-4 of spend)

We work with clients on a 30-day cycle. By the end of week 3, decide: - What's working, what's not - What to iterate on - What new pages to build - Set up the next 30 days

The audit refresh happens every 30 days (not 90). Each cycle builds on the last: new pages, iterations on winners, killing losers, expanding to new personas or ads.

Phase escalation

Win conditions for a page: 1. Beats the existing page on profit per session (CPA, ROAS, CVR are inputs, profit/session is the verdict). 2. Adds incremental volume: total spend goes up while efficiency holds. The page expanded the client's ability to scale.

7. The AI Skills

All skills live in landersrx/skill/ as packaged .skill zip archives, plus an unpacked source directory for landersrx-account-audit. To invoke, use the trigger phrases listed in each skill's description (or use the slash-command form when surfaced).

The skills, in order of use

Skill When to invoke Triggers Output
client-brand-onboarding First touch with a new client, before any page work "set up the brand for [client]", "onboard [client]", "create a brand kit" clients/{client}/brand-kit.md + optional Figma brand file
landersrx-account-audit After access is granted, to produce the audit + roadmap as a single deliverable "run the audit on [client]", "build the page roadmap for [client]", "audit + roadmap" HTML + PDF audit document with 5 page concepts
landing-page-creator For each approved page concept "create a [page type] for [client]", "build the [page type]", "Shopify landing page", "create a quiz" Standalone HTML page + rationale doc + optional Shopify OS 2.0 section
static-ad-creator ⚠️ WIP Scale-tier clients, after pages are built "create static ads for [client]", "build statics matched to [page]" Static ad creatives (HTML/PNG) matched to the page they support

Skill anatomy

Each skill is a zip with {skill-name}/SKILL.md + {skill-name}/references/ + {skill-name}/assets/. SKILL.md has YAML frontmatter (name, description) followed by the workflow body. References are markdown files the skill consults during execution. Assets are scripts/templates the skill uses.

To peek inside a skill:

unzip -p landersrx/skill/{skill-name}.skill {skill-name}/SKILL.md

Skill conflicts and the right one to grab

8. The Tools

MCP-connected tools Claude uses

Build/delivery tools (operated outside Claude)

Direct API access

Local dev server

For previewing pages before Figma capture (run from the HQ root):

cd companies/landersrx/work
python3 -m http.server 8765

Pages render at http://localhost:8765/clients/{client}/{page}/index.html.

PDF generation (audits and decks)

"/Applications/Google Chrome.app/Contents/MacOS/Google Chrome" \
  --headless --disable-gpu --no-sandbox --no-pdf-header-footer \
  --print-to-pdf="<output>.pdf" \
  "file://<absolute-path-to-html>"

9. File and Folder Conventions

companies/landersrx/             # in HQ — source of truth, synced to the team
├── knowledge/PLAYBOOK.md        # This file. Master operations doc.
├── clients/                     # Per-client folders (brand kits, ad data, round-1, round-2, etc.)
│   └── {client-slug}/
│       ├── brand-kit.md
│       ├── ad-data.md
│       ├── round-1/
│       │   ├── pages/
│       │   └── audit.html / audit.pdf
│       └── round-2/
├── lead pages/                  # All client-facing landing pages, flat directory
├── deliverables/                # Reusable client deliverables (testing framework, etc.)
├── internal/                    # Internal docs (service-roadmap, design system, unit econ)
│   ├── service-roadmap.md       # Earlier version of the workflow doc; PLAYBOOK.md supersedes
│   ├── enhancement-plan.md      # Active enhancements roadmap
│   ├── 100m-unit-economics.md   # Working-backwards model
│   ├── landersrx-design-system.md
│   └── landersrx-master.html    # The LandersRX sales site (monolithic version)
├── sales/                       # Sales materials
│   ├── sales-notes.md           # Internal rep playbook (call structure, objections, levers)
│   ├── meeting-notes-*.md       # Strategy meeting notes
│   ├── offer-development.md     # Working doc for offer iteration (Hormozi GSO framework)
│   ├── offer-todo.md            # Cascade of offer-update action items
│   ├── offer-scorecard.html     # Visual GSO scorecard
│   ├── page-math-calculator.html # Local calculator (also lives on landersrx.com)
│   └── decks/
│       └── landersrx-lead-deck.html / .pdf  # Lead-facing sales deck
├── skill/                       # Skills (zip archives + active source for the audit)
│   ├── SKILL.md                 # Top-level skill router for the LandersRX site + service
│   ├── *.skill                  # Packaged skills (zip)
│   ├── landersrx-account-audit/ # Unpacked source for the audit skill
│   ├── references/              # Skill reference files (page-types, design-system, etc.)
│   └── templates/               # Onboarding / Production / Testing template DOCX files
├── references/
│   └── dtc-library/             # 30 annotated DTC pages with section-level "what works" notes
├── sections/                    # Shopify OS 2.0 sections for the LandersRX sales site
└── snippets/                    # Shared liquid snippets

Naming rules

10. Quality Bars and Output Rules

These rules apply to everything Claude produces for LandersRX work. Hard rules, no exceptions unless the user overrides explicitly.

Visual / branding

Voice / writing

Audit-specific rules

Brand-specific compliance

Every client has its own compliance and brand rules — voice constraints, restricted claims, regulatory requirements (FDA, FTC, etc.). Before drafting any client ad copy or landing page, read the per-client memory file and brand kit. Per-client rule sets live in ~/.claude/projects/-Users-bradploch/memory/ (referenced by name in MEMORY.md).

11. The Reference Library

Page types

landersrx/skill/references/page-types.md. Templates and section flows for each:

DTC reference library

landersrx/references/dtc-library/. 30 annotated DTC landing pages across 10 verticals and 7 page types. Each file has section-level "what works" commentary. Query during page creation for structural patterns (hero formula, comparison table structure, trust bar composition, social proof placement, pricing/offer framing).

Rule: pull structural patterns only. Don't copy headlines or body copy. Ad data always wins over library patterns when they conflict.

Design system

landersrx/skill/references/design-system.md. CSS variables, typography, component patterns. Adapt to the client's brand kit, don't replace.

Section catalog

landersrx/skill/references/section-catalog.md. All Shopify OS 2.0 sections with schemas and usage notes. For the LandersRX sales site itself, not for client pages.

12. Working with Claude

Session shapes

Trust-but-verify

Claude can produce a lot quickly. The strategist verifies before anything ships: - Numbers in the audit (cross-check against Motion directly if anything looks off) - Page copy (read it as if you were the client) - Brand kit (check [VERIFY] items) - Page renders (preview in browser; capture to Figma at BOTH desktop and mobile widths — every push lands a desktop + mobile frame pair so the client approves both views, never desktop alone; look on mobile)

When Claude should stop and ask

How Claude protects the bar

13. Templates

skill/templates/ holds the active client-facing template deliverables. As of 2026-04-30 these are still in .docx format; expect a migration to PDF as the canonical delivery format.

The previous Page Roadmap template has been replaced by the landersrx-account-audit HTML + PDF output, which combines audit and roadmap into a single visual deliverable.

Appendix: For Claude

When Brad opens a session and says "run the audit on [client]", do the following without asking for permission:

  1. Read this PLAYBOOK if not already in context.
  2. Resolve the client's Motion workspace via mcp__claude_ai_Motion__get_auth_context.
  3. Invoke the landersrx-account-audit skill, following its SKILL.md.
  4. Produce the deliverable HTML + PDF in clients/{client-slug}/.
  5. Copy both to ~/Downloads/ if Brad asks (he typically does).
  6. Summarize in 3-5 lines what's in it: top spend lane, biggest underbuilt opportunity, the 5 page types proposed.

When Brad says "build the [page-type] for [client]":

  1. Confirm the brand kit exists and is approved (clients/{client}/brand-kit.md).
  2. Confirm there's an audit on file. If not, ask whether to run one first.
  3. Invoke landing-page-creator with the page type and the relevant audit context.
  4. Save the page to lead pages/{client-slug}-{descriptor}.html.
  5. Verify the Figma capture script is included.

When Brad says "I'm working on [other thing]" with no LandersRX trigger:

Don't auto-invoke LandersRX skills. Skim the PLAYBOOK for context if the work is in this domain, otherwise stand down.

When in doubt about a decision (pricing, scope, branding), ask. The strategist owns judgment calls. Claude owns execution.