The work

The problem

A qualified lead cost roughly $30,000. In a business where every relationship starts with skepticism, the acquisition math simply didn't work at scale.

Constraints

• A regulated category — every claim reviewed before publication
• A small, wealthy, hard-to-reach audience
• Long trust cycles before any money moves
• Heavy dependency on external agencies

What I did

• Brought paid media in-house and rebuilt campaign structure around lead qualification, not volume
• Repositioned marketing internally as an acquisition engine — awareness in service of qualified pipeline, not the other way around
• Mapped every channel to a funnel stage (Lead → SQL → SAL → Client) so each asset had one measurable job
• Segmented nurture journeys by investor profile and intent stage
• Introduced AI-assisted multilingual production (EN/AR/FR), cutting asset production time ~50%

What I'd do differently

[Placeholder — one honest paragraph from Joel]

Act I — Build the engine (2021–24)

Wealth management content sits in Google's most scrutinized category — YMYL finance — for an audience that barely searches. The site started at Domain Rating 7, ranking for ~30 keywords.

• Content gap analysis that 5×'d ranking keywords in the first 5 months
• Technical SEO shipped with the dev team — structured pages, schema, automated crawl monitoring
• Publisher partnerships for editorial authority, lifting referral traffic 40%
• Built and coached the content team; week-on-week roadmaps reported to CEO level

Act II — Search changed

AI assistants and AI Overviews began answering questions before users ever reached a results page. Most brands are still deciding whether to care. I instrumented it instead — engaging a third-party AI-analytics platform to audit, for the first time, how the brand actually surfaces inside LLMs across our core markets and business lines.

Act III — Win the new surface (2025–26)

• The content system built for Google turned out to be the source material AI models cite: the brand measured #1 in AI-search visibility — ahead of J.P. Morgan, BlackRock, Fidelity, Goldman Sachs, and UBS
• #1 on Google AI Overviews; on ChatGPT, visibility of 22.6% vs 1.6% for the next-best wealth brand
• Reoriented organic around brand-demand capture: a four-phase measurement design at a flagship Jan 2026 event isolated a +108% lift in event-day clicks, CTR up 0.61%→0.90%, and a durable +48% post-event floor

The honest framing, up front

I worked on these accounts from 2019 to mid-2020, building their SEO and content foundations — keyword-gap analysis, content pipelines, technical audits, backlink groundwork. The shaded band on each chart is my tenure. The curves extend past it on purpose: I claim the foundation and the inflection, not everything that followed.

Turtlemint — insurance, India

YMYL before I knew the acronym: insurance content for first-time buyers. Keyword-gap analysis and a structured content pipeline took organic clicks and impressions up ~700% within 7 months — and the engine kept climbing.

Freakins — D2C denim, India

The opposite problem: a young D2C brand with no search footprint at all. Category and product-cluster architecture laid down in 2019–20 became the structure its later growth ran on.

Why this is here

It's where I learned the thing I still believe: organic growth is a system you install, not a campaign you run. Done right, it keeps paying after you leave the room.

The starting point

A 25-year-old communications agency whose digital offering was social media management — posts, calendars, community. No media buying. No performance marketing. No search. No email. I joined as a digital account executive.

What I built

• Pitched and stood up an entirely new service line: paid media buying and performance marketing across search and social
• Extended it into SEO and email marketing — four capabilities the agency had never sold before
• Ran it end to end: proposals and tender submissions, campaign architecture, customer journey maps, KPI frameworks, and client reporting
• Managed a multinational client's regional digital presence (Hempel) while building the practice

The result

The new offering drove a +45% increase in the agency's net revenue intake in its first quarter, cut client CPCs ~25%, lifted retention ~20% through email re-engagement — and earned me the promotion to Performance Marketing & Growth Specialist.

Why this is here

Because this is the closest thing marketing has to founding: no playbook, no existing revenue, sell it before you build it, build it while you run it. That's the muscle early-stage teams actually need.

The hypothesis

[Placeholder — Joel's original hypothesis paragraph: what problem OurSplit tests and for whom]

The build

• Full-stack build: Next.js frontend, Supabase backend, shipped solo with AI-assisted development
• Core product decisions locked; currently in live trial with real usage
• Every screen shown is the product as it exists — no mockups

Why this is here

Most marketing portfolios claim "AI-native." This is what I mean by it: not prompts for ad copy, but the ability to take an idea to a working product — and then apply growth thinking to something I own end to end.