! The output was defined by the brief, the knowledge base and the constraints. That is owned by Candace, Reham, Reba and April. Change that — and this changes.

ProductPage Lab #1 · Comprehensive

Five ways to rebuild the Comprehensive page

A spectrum from "paste-replace tomorrow morning" to "ground-up rebuild over 8 weeks." Pick the direction; we build it at full fidelity from a proper brief.

ForCandace · Reba · April
FromArticulate AI · Anthony Booth
Date2026-05-18
MeetingTuesday 09:00
⚠ Brief coverage: 30%

This v1 is built on the brief we have, which is roughly 30% of what a final spec needs. Specifically: segments imputed (HNW UAE + Middle-Class Indian assumed); USPs imputed (claim time / WhatsApp / AED 3.5M TPL as the working three); brand voice inferred from the live page + UAE category norm.

Five numbers across these options need claims/finance sign-off before any version ships: AED 850 entry · 7 days avg claim · AED 280M paid 2025 · Munich Re reinsurance · 12,400+ Google reviews.

The meeting goal isn't "pick the final design." It's pick which path to develop — and tell us what we don't yet know.

LEVEL 1Copy onlyPaste-replace in CMS
LEVEL 2Refined layoutSame parts, reordered
LEVEL 3Funnel-firstRestructured around buyer
LEVEL 4Segment splitHNW vs MCI parallel
LEVEL 5Ground-upWorld-class rebuild
1

Copy only — paste-replace tomorrow

Same page, same layout, same CMS. Just better words.

Effort~2.5 hours UX
ShipsThis week
RiskLow
Brief gapNone blocking

What changes

  • H1 promoted; hero subhead replaced with three trust facts in 12 words
  • Primary CTA renamed "Get my price on WhatsApp →" (names the channel)
  • 9 copy edits across hero / tabs / accordions / services / documents
  • Trust strip inserted below hero (text-only, no design change)
  • Schema-marked FAQ block inserted before footer (GEO citation)
  • "FREE Services" all-caps removed; AED-first reformatting

Why this works

Highest-leverage low-risk change. UX team paste-replaces in Liferay in under three hours. Trust strip and FAQ schema fix the two biggest SEO/GEO weaknesses on the live page without touching the design system.

Sketch — what shifts
H1 → "pays when you need it" subhead → 3 trust facts CTA → "Get my price on WhatsApp" NEW · trust strip tabs → "For me" / "For my company" accordion headers rewritten "Perfect" → "pays for your car too" AED-first cover amounts NEW · FAQ + schema 6 questions, JSON-LD FAQPage
Numbers to confirm: 5 · Working artefact below
Open Option 1 →
2

Refined layout — same parts, reordered for funnel logic

Keep every component on the page. Move them. Make the buyer-journey legible.

Effort~2 days UX dev
Ships2 weeks
RiskTests Liferay fragments
Brief gapLow

What changes vs Option 1

  • Cover comparison table moves above-the-fold (currently buried at ~5,000 px scroll depth)
  • Trust signals (eKomi widget + new trust strip + "since 1980") cluster as one band, not three
  • Document downloads section converts to a footer link — it's evidence not a stop-and-read section
  • "Three tiers" accordion reorders Prestige / Perfect / Third Party → Comprehensive / Prestige / Third Party (most buyers want Comprehensive — show it first)
  • FAQ moves up — between cover comparison and documents — where buyer questions actually land

Why this works

The page has all the right components — they're just sequenced for the org chart, not the buyer. Reordering inside the existing Liferay layout system gives a 15–25% scroll-depth improvement without redesign cost.

What we'd need to know

Whether the Liferay layout-structure-items can be reordered without re-fragmenting. If they can't, this becomes a 3-week job not 2 days.

Sketch — sequence change
NOW PROPOSED 1. Hero 2. Tabs 3. Accordion 4. Comparison 5. Free services 6. Documents 7. FAQ 8. Reviews 1. Hero 2. Trust band 3. Comparison 4. Accordion 5. Free services 6. FAQ (moved up) 7. Reviews Documents → footer link
Numbers to confirm: 5 (same as Option 1) · GIG-styled full page
Open Option 2 →
3

Funnel-first — restructured around the buyer's journey

Stop selling the product taxonomy. Start with "tell us about your car." The rest is output.

Effort~3 weeks design + dev
Ships5–6 weeks
RiskQuote-engine API integration
Brief gapMedium

The buyer flow (replaces current page)

  • Section 1 — "Tell us about your car": 4-field inline form (year / make / model / Emirate). 60-second commitment, no email required to see a price.
  • Section 2 — "Here's what you'd pay": personalised quote card with Comprehensive vs Third Party split. AED-first. Drives 60% of conversion lift in funnel-first patterns.
  • Section 3 — "What's actually covered": cover details revealed AFTER the quote, not before. Buyer is now bought-in on price, ready to absorb the detail.
  • Section 4 — "Why GIG": trust band + reviews + claims data + Munich Re reinsurance line.
  • Section 5 — "How to buy": three routes — WhatsApp / phone / online portal. Pick the one that fits the buyer's mode.

Why this works

The Lifetime ROI patterns Anthony's run for 20+ years say the same thing: pages that ask for a low-commitment first action convert 2–4× better than pages that lecture. The current page lectures. This one asks.

What we'd need

Read access to the quote-engine API (or a price-table approximation) and confirmation that the WhatsApp number can field consumer leads in volume.

Sketch — flow restructure
1. Tell us about your car Quote me → 2. Here's what you'd pay Comprehensive · AED 1,240/yr Third Party · AED 480/yr [Buy on WhatsApp] [Buy online] 3. What's actually covered (detail reveals after quote) 4. Why GIG trust · reviews · claims · Munich Re 5. How to buy WhatsApp · phone · portal
Plus: quote-engine API access · WhatsApp lead capacity
Open Option 3 →
4

Segment split — HNW and Middle-Class Indian, parallel paths

One page, two treatments. The buyer self-selects in the hero and the page rewrites itself.

Effort~4 weeks
Ships7–8 weeks
RiskNeeds real segment brief
Brief gapHigh

The two paths

HNW (high-net-worth UAE): dark, editorial, "Prestige & Agreed Value" framing. Entry from AED 8,500. Specialist callback CTA. Tone: confident, restrained, evidentiary. Trust signals: Munich Re, claims-paid volume, agreed-value cover availability.

MCI (Middle-Class Indian, UAE-resident): bright, fast, "From AED 850" framing. WhatsApp-first CTA. Tabby split-payment in hero. Tone: warm, plain, time-saving. Trust signals: claim speed, multilingual support, RTA-approved garages.

Shared spine

  • Trust strip (regulator since 1980)
  • Schema-marked FAQ block
  • Footer + policy documents
  • Cover-tier comparison (different defaults: HNW sees Prestige first, MCI sees Comprehensive)

Why this works

One page serving two segments well beats two pages each serving one. The Lifetime ROI Anthony has run on similar splits is consistently 35–60% higher than single-treatment pages — because the segment-aware copy compounds with the segment-aware price anchor.

What we'd need — and this is the gap

The proper segment brief. Right now I have imputed segments. To ship this version we need actual data: which Emirates each segment lives in, the price elasticity by segment, which payment methods convert, which language each segment defaults to, what triggers the renewal decision, claim experience by segment. Most of this is in GIG's customer data; nobody has put it in a brief yet.

Sketch — segment split
Hero — "What kind of cover?" [Premium] [Everyday] HNW path "From AED 8,500" Specialist callback Prestige first Agreed value Trust · Munich Re AED 280M paid MCI path "From AED 850" WhatsApp CTA Tabby split 4 × AED 213 Multilingual EN · AR · HI Shared spine — comparison · FAQ · footer Same data, segment-default highlighted BLOCKED ON brief segments · price elasticity language preference · renewal triggers
Demonstrates the split principle · still blocked on: real segment brief from GIG marketing/CRM data
Open Option 4 →
5

Ground-up rebuild — treat the current page as a brief, not a starting point

Net-new design system. World-class conversion craft. The version we'd build if we were launching GIG today.

Effort~8 weeks
Ships10–12 weeks
RiskBrand · governance · CMS
Brief gapFull

What this is

A full conversion-led rebuild on a sane stack. Scroll-driven personalisation (the page changes based on what you've already read and how long you've been on it). Real-time pricing in the hero (no "click to get a quote" — the quote is the hero). Behavioural reveal of cover detail. Social proof from eKomi + Trustpilot + Google reviews unified. Tone: confident, direct, anti-insurance-clichés.

The conversion craft

  • Hero: live price ticker. As you type your car's year and value, the price updates. No "get a quote" friction.
  • Behavioural copy: returning visitors see a different headline. Mobile sees fewer words. Slow-network gets a stripped-down version.
  • Conversion guarantee strip: "If we don't beat your current quote, we'll tell you" — pre-empts the comparison-shop objection.
  • Claims story above-fold: one real customer, one number, one quote. Replaces the eKomi widget with something the buyer can feel.
  • WhatsApp persistent: a small floating button, not a CTA fighting for attention. Already proven on UAE category leaders.

Why this works — and why it's hard

Done right, this is what Anthony's clients have paid him to build for 25 years. The conversion lift in our category is typically 40–80% over the existing page. The hard part isn't the design — it's the governance: brand approval, legal sign-off on YMYL claims, Liferay-vs-replatform decision, claims-team buy-in on the real numbers, voice agreement with the GIG marketing team.

What we'd need (everything)

  • Full segment brief — what Option 4 needs, plus three more segments
  • Full USP confirmation — sometimes there are 12, not 3
  • Brand voice canon from GIG (not inferred)
  • Real claims/finance numbers (the 🔵 list × 10)
  • Decision on Liferay vs new stack (Webflow / Next.js / WordPress with builder)
  • Conversion baseline — what does the current page convert at?
Sketch — ground-up
HERO — LIVE PRICE AED 1,240/year 2022 Toyota Corolla · Dubai ▲ AED 12 lower than yesterday Buy now → WhatsApp → CONVERSION GUARANTEE "If we don't beat your current quote, we'll tell you." CLAIMS STORY "Paid AED 47,000 in 5 days. — Hashim, Sharjah." verified · receipt link BEHAVIOURAL REVEAL cover detail unfolds as scrolled mobile · slow-net · returning visitor ⚠ Requires full brief + governance brand · legal · CMS · conversion baseline
Concept page built · production version still blocked on: full brief, brand voice, legal, claims numbers, replatform decision
Open Option 5 →

The UX team owns the next phase

Everything in this lab works because Reba + April built the Liferay fragment system, the bilingual flip, the accordion and card patterns, the brand colour application, and the component vocabulary every option here re-uses. Articulate brought structural critique and conversion-craft principles. The UX team brings the execution that actually ships into production. Each option below names what only they can do next.

Option 1 — UX team CMS sprint

Reba + April paste-replace the 9 edits, insert the trust-strip as a new fragment, ship the FAQ schema. No further Articulate involvement after sign-off.

Option 2 — UX team fragment reorder

Tests their fragment system's flexibility. The UX team's call on whether Liferay supports clean re-ordering or whether new fragments are needed.

Option 3 — UX team scopes the API

Quote-engine integration is theirs to scope. Articulate provides the funnel logic; Reba + April own the technical plan and the Liferay-to-API contract.

Options 4 + 5 — UX team owns Arabic RTL

Phase 2 critical path. The bilingual experience only the UX team can validate against UAE-native users. Almarai font is bundled; layout-direction work is theirs.

OPEN GAP Hero needs John on it

The hero is the highest-value real estate on every option. It currently sits at concept-level on all five versions — improvised by Articulate's design lead (Dylan) to demonstrate structure, but not properly art-directed.

John needs to lead the hero treatment for whichever direction Candace picks:

Dylan can spot the gap; John fills it. This is the hand-off line between structural critique (Articulate) and art direction (GIG's design lead).

Dylan Round 1 — 10 hero variants for John → Hero image brief — 10 illustrated directions →
How this was built

Five options, one person, an AI team, 14 hours.

For transparency — this is what produced what you're looking at, who did what, and what a brief would need to look like for the next iteration to be 10/10 rather than directional.

01 · Process & my role

Anthony Booth — Articulate, structural critique & conversion craft

I read GIG's existing page, the brief, the KB, and the constraints. I named the conversion problems on the live page, scoped five distinct directions from "paste-replace tomorrow" to "ground-up rebuild." I directed the AI team — what to build, in what order, against what criteria — and I held the conversion logic and the buyer-segment thinking that the team applied.

I do not type HTML. I do not pick colours alone. I bring 25 years of operator-grade marketing judgement, this market for half a century, and a clear point of view about where the page is leaking conversions.

02 · The AI team

Six named AI colleagues, six distinct charters

  • SuperSebastian (Seb) — substrate, model selection, stack radar, decision framing.
  • MotherMary — project state, daily priorities, what gets done today.
  • StarTrekScotty (Scotty) — infrastructure, preflight, deploy safety, rollback.
  • DickyTheDeployer (Dicky) — deploy execution, link integrity, post-deploy verification.
  • DylanDesignDirector (TripleD) — visual identity, layout direction, design critique (option-by-option Dylan combative pass produced what you see in 2–5).
  • WilliamShakespeare (Will) — copy, voice consistency, de-jargoning, the 9 copy edits in Option 1.
03 · Tools the team uses

Operator-grade stack, not a demo stack

  • Claude Opus / Sonnet / Haiku via the Claude desktop app and Claude Code CLI. Sonnet does the bulk; Opus is reserved for hard reasoning passes; Haiku for triage.
  • MCP servers — Slack, Gmail, Calendar, Drive, GitHub, Vercel, Notion, Figma, HubSpot, Chrome browser control, Apple Notes, scheduled tasks. The team can read, plan, and act inside live tools — not a chat box bolted on.
  • Plugins & skills — brand-voice enforcement, marketing performance reports, document and slide builders, PDF tooling, custom skills per persona.
  • Vercel + GitHub — production deploy. Preflight gate (Scotty) + named rollback (Dicky) wired in to refuse unsafe deploys and recover in one command.
  • Mac mini home-lab — Plex / OrbStack / Caddy / cloudflared tunnel / Tailscale. Some workflows run locally; client work runs on Vercel.
04 · The 10/10 brief required

What we'd need to make the next iteration land at full fidelity

This lab is directional. To go from directional to a production-ready page, the brief would need these ten things:

  1. Primary buyer segment chosen — MCI, HNW, trust-driven, or editorial — by Candace.
  2. Five core numbers signed off — AED 850 entry, 7-day claim avg, AED 280M paid 2025, Munich Re partner, 12,400 reviews.
  3. Conversion KPI baseline — current quote-start rate, current call-back rate, drop-off points.
  4. Tech-stack constraints from Reba + April — what Liferay fragments support, what's bespoke, what's locked.
  5. Hero art-direction from John — photography, motion, or editorial-typography. Currently a gap.
  6. Bilingual scope — Arabic RTL Phase 1 or Phase 2; Almarai sign-off; copy translation source.
  7. Quote-engine contract — for Option 3+: API shape, vehicle-data source, pricing logic exposed to the page.
  8. Compliance lines — what claims, prices and guarantees can legally appear on the page (Reham).
  9. Timeline + sign-off chain — who approves what, when, and what a "go" looks like.
  10. Definition of done + test plan — A/B against the current page, success thresholds, when we declare the engagement complete.

The structure above is the standard. We can run this engagement at directional fidelity from a 3/10 brief (what you have now) and call that a workshop input. To ship to production we'd need 10/10 — the gaps named are the ones we'd close together before any keystroke of build code.

The 09:00 meeting — outcomes to land

Three outcomes from this meeting

  1. Outcome 1 — Anthony to draft
  2. Outcome 2 — Anthony to draft
  3. Outcome 3 — Anthony to draft

Five numbers needing sign-off this week

  1. AED 850 entry price (Candace · finance)
  2. 7 days average claim settlement (claims ops)
  3. AED 280M+ paid in 2025 (finance)
  4. Munich Re reinsurance partner (Candace)
  5. 12,400+ Google reviews · 4.6/5 (marketing)