Marina Ruiz Available · Los Angeles

HouseAmerica · Home lending · Design system and landing page architecture

Turning 25 contact cards into 25 lead-capture funnels

An independent mortgage lender with 25 loan officers, $11B funded in 2025, and a website untouched for five years. Every loan officer's page was a contact card.

Role

Sole designer

Client contact

Executive Director, Business Development

Scope

Competitive analysis, system architecture, design system, three landing pages

Status

Proof of concept, rollout queued

Every loan officer page was a headshot, a phone number, an email address, and a calculator. No positioning, no proof, no path to conversion. I analysed 10 competitor systems, reframed the brief from "rebuild one page" to "build a template system," and designed a complete design system plus a three-audience landing page architecture, deployed first as a proof of concept for the executive who championed it, then structured to clone across the remaining 24.

Nothing on this page is a measured outcome. The figures here describe the size of the problem, not movement produced against it.

Line 01

The brief was wrong, and that was the finding

The engagement started as "fix Bill's landing page." The competitive analysis ended it as something else. Every serious loan officer page in the benchmark set, from a solo operator running one high-converting page to a national lender running 2,800+ officer pages on a single template, treated the page as a conversion surface with purpose, proof, and path. HouseAmerica's 25 pages treated them as directory entries.

So the finding I delivered was not a design critique. The gap is not design, it is product thinking. And the fix is not one better page, because one better page leaves 24 broken ones and no leverage. The recommendation became a template system: shared modules, the flagship home-equity product block, lead magnet, FAQ, product grid, company stats, plus per-officer CMS fields for bio, testimonials, stats, and booking link. Deployed once, cloned 25 times.

The client's own words made the third audience explicit: rebuilt pages would double as a recruiting asset when pitching loan officers at competitor firms. One system, three audiences: homeowners, buyers and their agents, and recruitable loan officers.

Line 02

One wedge product, on every page

A fast home-equity product became the hero of every page rather than sitting inside a product list, because it was the single offer that worked on all three audiences at once. The flagship product card is the only one on the page carrying a button. Everything else earns a "Learn more" link. Hierarchy as a business decision, not a styling one.

What it cost

Every other product was demoted to a text link. A lender with a full catalogue accepts that its other offerings stop competing for attention on the page that matters most.

Line 03

Price the ask to the visitor's risk

The recruiting track was restructured into three tiers. Asking a currently employed loan officer to book a call as step one was the original design's mistake: commitment has to be priced to what the visitor is risking by taking it.

01An anonymous calculatorCosts nothing and reveals nothing. The hero points here.
02An email-gated PDFOne identifier, in exchange for something worth reading.
03A 15-minute callThe real ask, made only after the first two have earned it.

Line 04

Governance beats expression here

Officers lost the ability to make a page "theirs" beyond the CMS fields. The system accepts that deliberately: consistency, maintainability, and a five-year-stale site all argue that governance wins in this case.

What it cost

That is a real cost in an industry built on personal brands, and it is the constraint most likely to generate pushback as the rollout reaches officers who did not champion it.

Line 05

One gold number, and an accessibility catch

The system was built from scratch and locked before the build: a navy anchor connected to the existing brand, a golden wheat accent, soft linen backgrounds reserved for the flagship product sections only, and a full spacing, card, badge, and button vocabulary.

Two decisions worth pointing a reviewer at. The trust bar renders every stat in white except the single most impactful number, which takes the accent: one gold number reads as signal, five read as decoration. And gold eyebrow text failed contrast on light backgrounds, so eyebrows became navy text with a gold left border, keeping the accent's identity without failing WCAG. Body text on navy went through the same discipline, rejecting semi-transparent white for a solid tint that holds contrast.

The system shipped as a component reference document and a full-page mockup, then drove the production build. The iterations after implementation were the unglamorous kind that make a system real: fixing a hero card clipping into the stats bar, guarding count-up animations against double firing, de-emphasising a tag that read as a CTA, and repairing a comparison table's contrast so "them versus us" was legible at a glance.

Line 06

What is built, and what is queued

Proof-of-concept pages are built for the championing executive: officer profile, product landing page, and recruiting page, all on the locked design system with inline booking embedded. The second officer deployment is queued, chosen for his strong social presence, and the template's CMS fields make each additional page a content task rather than a design project. Lead and booking metrics will be added as the rollout proceeds.

Honest edges

What this page does not claim

There is no measured outcome here and none is claimed. Three pages exist and the system is locked, but no lead or booking figures have been produced yet, so none appear on this page or on the slip. Every number above is scope: how large the problem was, how many systems I read, how much has been built.

The competitor benchmark is described generically on purpose. This entry earns its place among the measured case studies when the rollout produces figures worth printing, and not before.