Financer Comparator Redesign — 5 Iterations

Context & approach

What we set out to do

The current Financer.com loan comparison pages rank among 26 lenders with a static list and a narrow sidebar filter. Competitor US sites (NerdWallet, Bankrate, WalletHub, Forbes Advisor, LendingTree) have moved past this pattern — their pages feel like tools the user is operating, not lists the user is scrolling. We're 1-2 generations behind on UX.

Before writing a single line of code, we ran an expert-level research pass: four parallel Opus agents covered CRO patterns for affiliate comparison pages, 2026 SEO/YMYL requirements, behavioral psychology of borrowers at the moment of click, and online marketing funnel strategy. Findings fed into a consolidated framework of design principles and priority patterns.

We then built five sequential prototypes (v1 → v5), each adding one layer of functionality. Every iteration was opened in the browser and validated visually before moving to the next. Early versions prove the core approach; later versions add the commercial, compliance, and SEO layers needed for a production-ready page.

The five iterations

Each card below describes what was added, why, and what we learned. The recommended path is v5 — the others exist to show the reasoning trail.

v1 is the biggest conceptual shift. The current financer.com page is a list: user arrives, sees 26 items, picks one. v1 turns the page into a tool: user arrives, sets their loan amount + term + credit tier + state at the top (a "hero input strip"), and every card recalculates in real time — monthly payment, total cost, APR bar with a "your likely rate" dot, and ranking reshuffle.

What's added versus the current live page:

Hero input strip with sliders and credit-tier buttons — the primary controls, not a narrow sidebar widget
Live "X of 26 lenders match your profile" badge with pulse animation
State-aware compliance alerts — if you select NY/CT/PA/MA, lenders who can't legally serve you are filtered out automatically with an explanatory banner
MLA 36% servicemember notice — federal law that current page doesn't reference
Triple byline: Written by + Edited by + Reviewed by [CFP®] + Fact-checked — 2026 YMYL E-E-A-T requirement
4-pillar methodology scorecard with transparent weights (Affordability 35% / Accessibility 25% / Transparency 20% / Speed 20%) — Bankrate-style
TILA representative example — legally required but missing from current page
CTAs changed: "Check My Rate" + "Won't affect your credit score" microcopy (replaces high-commitment "Apply Now")
Optional 4-step skippable match quiz — returns personalized ranking without collecting PII

Key decision Leave the filter sidebar intact for now. v1 intentionally keeps the content density low — we wanted to prove the "tool not list" shift without confounding it with filter changes.

How it works: slider/button input events write to a global state object, trigger a debounced recalc, which re-runs ranking + re-renders all 26 cards. Monthly payment and total cost use the standard amortization formula. Everything persists to localStorage so a returning visitor sees their last inputs.

The current production page has a right-side "Compare Loans" filter widget with: Loan amount, Term, Age (??), Features multi-select, Accept bad credit history, Weekend payout, Payment within 24h, Loan broker, Interest free loan, No income, Recommended company.

Honest take: most of these filters don't help visitors. Let's walk through why:

Age slider — Why would I filter lenders by my age? This is an eligibility condition, not a preference. No US competitor exposes this.
Loan broker — Technical jargon that consumers don't parse. Should be a trust signal on the card, not a user-controlled filter.
Interest free loan — Sets false expectations; in US bad-credit, almost nothing is truly 0% APR. Users filter, see nothing, bounce.
No income — Legally risky framing and confusing category.
Recommended company — "Recommended" by whom? Self-referential and erodes trust when pre-checked.
Weekend payout — Too niche to be a primary filter. Should fold into "Funding speed".

v2 proposes a replacement: keep amount, term, and credit tier in the hero, and add one compact "More filters" expandable with three genuinely useful controls:

Funding speed — Any / Same day / Within 3 days (pill group)
Loan purpose — affects ranking weight, not eligibility (a debt-consolidation user benefits from lenders ≤ 36% APR; emergency user benefits from same-day funding)
Safety & availability checkboxes: Soft credit check only · APR capped at 36% · Available in all 50 states

Direction we took Filters are table stakes — every US competitor has them. A richer filter sidebar is not a moat. The differentiator is the live interactive tool (v1) and everything stacked on top of it.

Why this matters: a shorter, more useful filter set reduces decision paralysis and builds trust. Users who see "No income" and "Loan broker" as filters are being shown implementation detail, not product benefits. v2 hides that complexity.

The current production page includes a "Financer's Choice" block promoting Ascent Funding at the bottom. Ascent is a student loan provider (APR 2.69%–15.26%, in-school deferred payments). On a page targeting "no teletrack tribal loans / bad credit," a user with a sub-600 FICO looking for $500 fast does not qualify for Ascent. That's a mismatch between traffic and offer — and it's probably a partnership that was placed without thinking about page-level relevance.

A static "our pick" block that recommends the wrong product for the audience is worse than no pick at all. It trains users to ignore the site's recommendations.

v3 fixes this by generating the "Our pick" dynamically from the user's inputs:

Lender chosen = the #1 ranked card from the current sort (which changes with inputs + filters + quiz results)
Reasons listed = 3-5 specific lines that reference the user's inputs ("Accepts your Fair credit — their minimum is 580", "Your $2,000 fits their $100–$40,000 loan range", "Licensed in Texas where you're located", "Funds the same day", "Uses a soft credit check — won't affect your score")
Savings delta = quantified comparison against #2 ("Estimated $47 less interest vs. #2 pick over your 12-month term")
Transparency footer: "Our pick changes when your inputs change. Ranking based on your credit, state, and amount — not commission."

Pattern we ported LendingTree quantifies "average $1,659 saved" as a headline CTA driver. Forbes Advisor's "Why we picked it" explains rank reasoning per card. We combine both into a dynamic, per-visitor recommendation block that explains itself.

Placement decision: the block sits below the comparison cards, not above. The user first scans 26 options (agency), then gets a verdict they can trust because they've seen the alternatives. Placing it above would feel like paternalism.

v1–v3 are tool-complete but would not ship to production. v4 adds the layers that our research surfaced as non-negotiable for YMYL finance pages in 2026:

Red Flags / "How to spot a loan scam" block — 6 warnings (upfront fees, guaranteed approval, pressure tactics, gift-card payments, hidden APR, missing license) with FTC/CFPB reporting links. Information-gain the competition doesn't match, and a trust signal that pre-empts the biggest objection a subprime borrower has.
FAQ section — 7 questions with direct 50-100 word answers, targeting Google PAA and AI Overview capture. Schema-backed.
JSON-LD schema — Article (with author credentials + reviewedBy + dateModified), FAQPage, ItemList → Product → Review → AggregateRating. Populated dynamically for all 26 lenders.

Why "Red Flags" here but not universal For bad-credit / subprime loans, fraud risk is elevated and borrowers are vulnerable — scam warnings are high-value and high-relevance. On a credit card or brokerage comparison page, the same block would feel off-topic. The structural slot stays; the content changes per category. For credit cards, it's "Rewards traps that shrink your return." For brokers: "Custody and insurance truth." For HYSA: "How banks silently drop your APY." Universal pattern, category-specific content.

What v4 doesn't yet have: NMLS IDs, live-verified timestamps per lender, dated benchmark statistics, methodology versioning, and sticky mobile CTA. Those come in v5.

v5 is what we'd ship. It adds 5 quick-win improvements from our gap-analysis research plus the single most differentiating per-card feature in the space: Approval Odds.

What v5 adds

Approval Odds badges per card — "Likely qualifies" (green) / "May qualify" (amber) / "Long shot" (red), computed from user credit tier + amount + state fit + broker rating. This is the closest open-affiliate analog to Credit Karma's "Approval Odds" moat. Label includes a tooltip: "Estimated fit based on your inputs — not a guarantee."
NMLS ID per lender — clickable to NMLS Consumer Access. For lead generators and brokers without NMLS, an honest "Lead generator — no NMLS" label. For tribal lenders, "Tribal lender — chartered".
"Verified N hours ago" timestamp per card — freshness signal, E-E-A-T win.
Benchmark stats section — 4 dated statistics with inline source links (Experian Q1 2026, Federal Reserve G.19, TransUnion, CFPB). This is the single highest-leverage AI Overview bait in the set — LLMs extract passages exactly this format.
Methodology versioning — "Methodology v2.3 · Reviewed April 2026 · View changelog" pill in the trust block. Aligns with Google 2026 YMYL guidelines.
Sticky mobile CTA — shows up on mobile viewport below 768px, summarizes current inputs ("$2,000 · 12 mo · Fair credit · 14 matches"), and offers a quick jump-to-cards button. Research shows 10–25% mobile CTR lift on long finance pages.

Why v5 is "ready" but v4 isn't v4 is a demo. v5 has the compliance, trust, and SEO plumbing that YMYL finance needs to rank AND convert in 2026. The difference is mostly invisible to a casual scroll, but it's what the page needs to survive Google's 2026 ranking updates + FTC/CFPB scrutiny.

What v5 deliberately doesn't have

We stopped short of adding more content. No expanded "what is a bad credit loan" section. No second FAQ. No 2,000-word "Understanding subprime lending" editorial block. This is a deliberate content-density ceiling — see the plan below.

Next: category pivot — Romanian investment brokers

After v5 proved the framework for US bad-credit loans, the next test was whether it ports to a completely different product category in a completely different market. v6 is a Romanian broker comparison built on the same frame — and v6.1 brings it up to international parity with Brokerchooser / NerdWallet / Investopedia.

v6.1Category pivot — Romanian investment brokers

Proving the template is portable: same frame, different product, different market, different regulator, different tax regime.

🇷🇴 Romanian market FIT engine (not RATE) 2026 tax reset aware 4 investor profiles Brokerchooser benchmark Trust tiers + sub-scores AUM dated stats

v6 takes the v5 framework (tool-heavy comparison, live recalc, "Our pick for you", category-specific TIGB, FAQ, methodology, sticky mobile CTA) and adapts it to Romanian investment broker comparison. It's a concrete test of the category-adaptation matrix we developed in research: if v5 works for a totally different product in a totally different market, the framework is genuinely reusable.

What fundamentally changed

RATE engine → FIT engine. Loans compute monthly payment from a single formula. Brokers need multi-dimensional matching — cost estimate + asset coverage + platform complexity + profile-weighted ranking. The same broker (e.g., XTB) can be #1 for a Passive investor and #5 for a Trader. Weights change per profile, not just per filter.
Credit tier → Investor profile. Replaced "Excellent/Good/Fair/Poor/Bad" with "Investitor pasiv / activ / Trader CFD / Investitor BVB" — four personas, each with distinct asset preferences, ranking weights, and default capital/frequency values.
US state filtering → Romanian tax regime split. Romania passed a fiscal reset effective 1 January 2026 that makes broker type a first-class signal: 3%/6% withheld at source for Romanian brokers / RO branches (no Declarație Unică), vs. 16% on net annual profit + mandatory Declarație Unică for foreign brokers. This is shown as a prominent side-by-side explainer block and a coloured tax-regime badge on every card. No US / UK / DE template has anything like it.
FTC / CFPB compliance → ASF / ESMA compliance. Different regulators, different mandatory disclosures. The ESMA-mandated "74–89% of retail CFD accounts lose money" warning sits above the comparison. FCI coverage (€20,000 per investor) appears on every card. ASF registry link replaces NMLS Consumer Access.
Approval Odds → Fit %. Brokers approve 99%+ of applicants, so approval-odds doesn't translate. Replaced with a 0–100% "potrivire cu profilul tău" dot (Excellent / Good / Partial match) that changes when the profile changes — not just when filters change.
Red Flags → TIGB adapted for brokers. Same block slot, completely different content: offshore licenses, double RON conversions, CFD leverage traps, copy-trading opacity, Declarația Unică responsibility, hidden inactivity/withdrawal fees — all specific to Romanian investor reality.

What v6.1 adds on top of v6 MVP

After a second research pass against international best-in-class (NerdWallet, Bankrate, Brokerchooser, StockBrokers.com, Investopedia, Forbes Advisor, ForexBrokers.com, LendingTree, Credit Karma, WalletHub), we identified 8 features our RO prototype was missing. All 8 shipped in v6.1:

Benchmark trades table (Brokerchooser signature move) — the same 3 trades (100 AAPL, €2k VWCE ETF, 1 lot EURUSD CFD) compared across all 10 brokers, ranked cheapest-first, with "Nu oferă" grey cells for assets a broker doesn't support.
Trust Tier badges (ForexBrokers.com "Highly Trusted" pattern) — ★★★ Listed publicly (IBKR, XTB, eToro, Plus500, Freedom24) · ★★ EU-regulated private (Trading 212, Revolut) · 🇷🇴 ASF-authorized (Tradeville, BT Capital, Goldring).
"Best for [X]" badges per card (NerdWallet pattern) — 2–3 categorical tags per broker ("DCA automat", "Copy trading", "BVB pur", "Capital mare", etc.) for scannable long-tail SEO.
Per-broker sub-scores (Brokerchooser pattern) — 4 mini bars per card: Costuri / Reglementare / Acoperire / Platformă (0–5 scale, gradient fill). Breaks down the composite Financer Score into defensible pillars.
Fit score as percentage (Credit Karma Approval Odds pattern, adapted) — replaces vague "Likely match" text with a specific "94% potrivire — excelentă" number.
Micro-persona line per card (WalletHub "Who Should Get It" pattern) — one sentence: "Potrivit pentru: investitori români începători–intermediari care vor 0% comision pe acțiuni/ETF + suport în română".
Quantified savings headline (LendingTree $1,659 average savings pattern) — computed live from benchmark: "Cu top brokerul nostru pentru profilul tău (XTB), economisești în medie €X/an comparativ cu brokerul median".
AUM + global accounts dated (Investopedia pattern) — hard trust signals per card: "3,2M+ conturi globale · $620B AUM · mar 2026" for IBKR.
Stats Authority Bar (NerdWallet "60+ data points" pattern) — navy-gold banner above results: "10 brokeri · 40+ criterii · 1.200+ puncte de date · Conturi reale testate · Actualizat lunar".

The 10 brokers compared

XTB, Tradeville, BT Capital Partners, Goldring, eToro, Trading 212, Freedom24, Interactive Brokers, Revolut Trading, Plus500. All with real regulator IDs (KNF, ASF, CySEC, CBI, FSC, LB), compensation scheme coverage amounts (FCI / ICF Cyprus / Polish KNF / Irish ICS / Lithuanian LB), actual 2026 commission structures, and Romanian-market accurate data: RON account support, BVB access flag, RO-language tax report availability.

Biggest insight from the pivot The structural frame (hero input strip → trust block → compliance alerts → live recalc cards → "Our pick" dynamic → TIGB → FAQ → methodology → sticky CTA → schema) ports cleanly. What changes: input fields, output metrics, TIGB content, ranking weights, regulator/tax context. Everything else stays. This is the reusable template for 10+ product categories across 23 markets.

What v6.1 deliberately doesn't have yet

Dedicated per-broker review pages (v6.2 structural work — blocks user reviews, side-by-side compare, long-tail SEO capture for "XTB review Romania" and similar branded queries). Also deferred: AI-generated personalized explainer tiles, live rate deltas, "Save my comparison" email capture, embedded platform screenshots. These are v6.2–v6.3 scope.

Direction going forward

My plan is not to have more content than v5. I want to keep comparison pages tool-heavy, not content-heavy. The comparison page is a decision tool — live recalc, smart filters, dynamic verdict, per-card badges, short dense trust blocks. That's the product.

For users who want to learn the underlying topic ("what is a personal loan," "how do credit scores work," "differences between subprime lenders"), we'll have dedicated informational pages — guides, wiki, blog articles — and we'll link to them from the comparison page. Mixing a 2,000-word guide into a comparison page hurts both jobs: it dilutes the tool and it competes with itself on SEO intent.

Priorities after internal review:

v6 confirmed the framework is category-agnostic. Same frame handled a totally different product (brokers) in a totally different market (RO) with totally different regulators (ASF/ESMA) and a market-unique tax regime. Next categories up: credit cards (feature grid engine, "rewards traps" TIGB), savings accounts (APY-focused), business loans (factor-rate engine).
Port v5 + v6 to the Next.js codebase — componentize the hero input strip, cards, "Our pick" block, TIGB, FAQ, benchmark trades, sub-scores. Hook ranking logic to Sanity-sourced data.
Build v6.2 — dedicated per-broker review pages (XTB / eToro / IBKR / Trading 212). Unlocks branded-query SEO, user reviews integration, side-by-side compare, and deeper content without bloating the comparison page.
Instrument + A/B test — baseline CTR on current live pages vs v5 / v6.1. Target: +15–25% mobile CTR, +10% downstream partner conversion.
Localize across 22 remaining markets — the v5 US template ports to DE / NL / FR / ES / IT / PL / CZ / SK / HU / BG / EE / LT / FIN / SE / DK / NO + LATAM (BR / MX) + Asia (ID / KZ). Adapt regulators, tax context, compensation schemes; keep the structural frame. v6's Romanian adaptation is the reference for how much changes (modest) vs. stays (most).

Financer Comparator Redesign — 6 Iterative Prototypes