Most of my web work doesn't fit a single case study — it's a body of growth and marketing pages produced at volume. Rather than design each from scratch, I built one visual system and a set of reusable patterns, then dialled the conversion intensity up or down to match where each page sits in the funnel. This gallery groups that work into four families: student-success proof, product and program pages, sessions and events, and careers.
A growth team ships constantly — a launch page this week, a success report next, a session registration, a careers post, a partner event. No single one of these is a flagship case study, but together they are the bulk of the public-facing web work, and they all have to feel like one brand while being produced far too fast to hand-craft individually.
Produce dozens of pages with no system and three things go wrong:
The answer wasn't one beautiful page. It was a system — shared patterns and a visual language — that any of these pages could be poured into, then tuned to the specific job it had to do.
A shared language — the same headers, footers, stat panels, icon-led explainer blocks, and proof patterns — so a launch page and a careers page read as the same brand even though no two are identical.
Conversion intensity scales with intent: a free top-of-funnel page reassures and educates; a high-ticket program page leans on scarcity and proof; a post-registration page switches from selling to retaining. Same system, different dial settings.
Patterns reusable enough that the next page is fast to build and on-brand by default — using a page builder for speed, and dropping to custom CSS and JavaScript only where the interaction demanded it.
Growth design is a systems problem disguised as a page problem. I built the system once, then spent my time on the decision that actually varies per page: what job is this page doing, and how hard should it push?
I designed and built these pages on a page builder for production speed, extending them with custom CSS and JavaScript wherever the native components couldn't deliver the interaction — filters, carousels, countdowns, and the like. The gallery below groups the work by the job each page does; every card opens the live page.
Four families of page, each tuned to a different moment in a prospect's journey — from proving the product works, to selling specific programs, to running live sessions, to hiring the team behind it all.
The proof pages turn a flood of student results into something a prospect can navigate and believe: an annual success report that re-cuts one dataset into several "wow" framings, a templated success-story system where only the names and numbers change — shown here across four students from a 705 to an 805-equivalent journey — and a filterable admits wall. The design job is legibility at volume: proof that scales without reading as sameness.
2025 Success ReportA year of outcomes plus the product investment behind them — one dataset re-cut into several achievement lenses.View live ↗
Success story — Divy, 735The templated case-study layout — hero score, before/after leap, per-section strategy, mentorship, recap.View live ↗
Success story — Anirudh, 765Same template, a different student — the 765 journey poured into the identical skeleton.View live ↗
Success story — Maria, 705The same system flexing to an international scorer's story — names and numbers swap, structure holds.View live ↗
Success story — Sankha, 715 (GFE)A GMAT Focus Edition variant of the same template — proof the system spans formats.View live ↗
M7 Admits proof-wallHundreds of elite-school admits turned into a filterable, self-segmenting credibility argument — a normalized outcome card plus a cohort filter.View live ↗Each product page takes something genuinely complex — a 1-on-1 mentorship program, a question-adaptive mock engine, a three-stage learning method — and breaks it into icon-led blocks, big-number stat panels, and head-to-head comparisons. The conversion dial moves with the price: a free mock says "try it"; a $499 program leans on scarcity and alumni proof.
Last Mile Push (LMP)High-ticket 1-on-1 program — a dual scholarship/purchase funnel and relative-advantage stats ("3X the success, half the time").View live ↗
Sigma-X Adaptive MocksA psychometric engine sold on insight, not score — translated into feature blocks and a comparison table against the official mock.View live ↗
GMAT Strategy GuidanceA top-of-funnel method page — a three-stage journey with a progression graph and free tools, soft-CTA instead of a sell.View live ↗The live-session pages run on a single registration layout — hero, countdown, details card, expandable content modules, host bios — content-swapped per topic. The sharpest decision is the post-registration state: the same template repurposed from acquisition to retention, swapping the register CTA for add-to-calendar and a guided pre-session checklist. Partner events (like the CAT decision-makers session) reuse the same system co-branded.
Weekly free sessionsA reusable webinar-registration template — countdown-driven, content-swapped per topic, with a post-registration retention variant.View live ↗
Partner event — CAT decision-makersThe session system co-branded for a partner event — seats, schedule, and speaker line-up on the same layout.View live ↗The careers pages apply the same growth thinking to a different audience: a homepage that pitches the company's mission and culture with the same warmth as a product page, and a job-description template that keeps every opening scannable — role, responsibilities, qualifications, and "what you'll get" in a consistent, accordion-style structure a candidate can read in a minute.
Careers homepageMission, what the team achieved, employee spotlights, and open roles — recruiting marketing built on the product visual system.View live ↗
Job-description templateA reusable JD layout — overview, key responsibilities, qualifications, and what you'll get — scannable in a minute.View live ↗