Design critique · Fintech, EUA
How Yendo’s site and application read to a first-time visitor
Design critique · yendo.com and apply.yendo.com · reviewed 9 November 2026
I went through yendo.com the way a first-time visitor would, from the home page to the first screen of the application, and wrote down what helped and what got in the way. I did it on my own, as part of my application for the Design Chief role. Everything here comes from the public site, so the ratings and personas are my judgement and haven’t been checked against Yendo’s data.
Eduardo Ries · Product Design Lead · Product Owner, AI Products
Estudo escrito em inglês, idioma em que foi apresentado. Capturas de tela de novembro de 2026.
On the calculator
$6,000
- about $205/mo
- 60 months
- 35.70% APR
One click later, in the application
$10,000
- 60 months
- 15.45% APR
Both values are hard-coded in the link. The amount the visitor picked is lost, and the APR doesn’t match the one disclosed on the page.
See F01- findings
- 16findings
- critical
- 2critical
- high
- 7high
- medium
- 4medium
- low
- 3low
01 · Overview
What I found
Yendo lends to people who own a car or a home and have thin or damaged credit. People in that spot read credit offers closely, and the site’s copy mostly explains the product well. It falls down at the points a careful visitor checks hardest. The terms change between the calculator and the application, the total cost isn’t shown, and it’s hard to tell who is actually lending. The visual inconsistencies matter less, so I rated them Medium or Low.
- 01
The offer changes when you click apply
The calculator shows $6,000 at 35.70% APR. The application it opens starts at $10,000 and 15.45%. The calculator never shows the total you’d repay either, and the ratings and download counts on the home page don’t agree with each other.
- 02
The home page doesn’t say what the product is
The hero talks about limits, speed and cashback. It never mentions the car, which the whole offer is built on, and “Is my vehicle at risk?” is the last question in the FAQ.
- 03
At the application, it’s unclear who you’re dealing with
The application opens in a new tab, looks like another company, shows Mastercard and Credit Karma logos and promises funds “with eligible banks and credit unions”. The site talks as if Yendo were the lender, and the small print names Cross River Bank.
About the offer itself
- 01
People apply to build credit, and the hero barely mentions it
Most customers say they apply to build their credit. The hero leads with limit, minutes and cashback, and bureau reporting and automatic limit reviews only show up halfway down the page.
Related: F03 · Positioning
- 02
The strongest argument is in the FAQ
Pew found a typical title-loan payment eats about half the borrower’s monthly income. Yendo’s minimum is 1% of the balance plus interest and fees, or $50. That gap is the best case for the product, and it isn’t near the top of any page.
Related: F05 · Positioning
- 03
Low minimums keep people in debt for a long time
At an APR near 30%, paying only the minimum keeps a balance open for years. For people whose goal is a better score, a payoff view (“pay $X a month and you’re clear in Y months”) would help them get there.
Product idea · Plan
- 04
The smallest loan is bigger than many needs
Pew puts the average title loan at about $1,000. Yendo’s Personal Loan starts at $3,100 over five years. For smaller amounts the Auto Card probably fits better, and the site doesn’t say so.
Related: F06
- 05
Yendo says it picks the product for you. The site makes you pick.
The home page says one application finds the best option, and Yendo describes routing customers to the right product. On the site, visitors still go through separate product pages and a comparison with no buttons.
Related: F06 · F13 · Plan
- 06
Moving between products is invisible
Someone who sells their car can move to the unsecured card and keep their history and rewards. Someone on the unsecured card can add an asset later. That answers “what happens to me afterwards?”, and it doesn’t come up anywhere in the visit.
Product idea · Plan
- 07
People know they could lose the car
Press coverage mentions that a car can be sold at auction if Yendo can’t reach the customer for a long time, along with the hardship programs meant to prevent it. I’d say both early. Visitors will find the first part elsewhere anyway.
Related: F05
Every finding, most serious first
I used Nielsen’s 0 to 4 scale (4 critical, 3 high, 2 medium, 1 low), judged on how many visitors hit the problem, how much it affects their decision and whether it keeps happening. Anything that changes what people understand about the offer, the cost or the lender ranks above visual inconsistency. IDs go by severity. Section 03 lists them in the order you’d meet them.
- F01criticalThe calculator’s numbers don’t carry into the applicationstep 5
- F02criticalThe application looks like a different companystep 5
- F03highThe home page never mentions the carstep 1
- F04highThe ratings and download numbers don’t matchstep 2
- F05high“Is my vehicle at risk?” is answered at the very bottomstep 3
- F06highThe comparison table has no buttons and favours the other productstep 4
- F07highThe calculator never shows the total you’d repaystep 4
- F08highIt isn’t clear who is lendingstep 5
- F09highThe application form doesn’t say how long it is or why it asksstep 5
- F10mediumEach page has a different hero colour and logostep 1
- F11mediumOn mobile, the note about your credit score is almost invisiblestep 1
- F12mediumThe headline numbers are maximums, and a couple disagreestep 2
- F13mediumProduct names keep changingstep 3
- F14lowThe mobile hero is misaligned and partly coveredstep 1
- F15lowThe Personal Loan page uses three content widthsstep 3
- F16lowSome icons don’t match their labelsstep 3
About this page
- Audience
- Who Yendo lends to and what each group is looking for.
- The visit
- Every finding, in the order a visitor would run into it. Each step ends with what already works.
- Visitors
- The same walkthrough repeated for different kinds of visitor, using proto-personas I wrote.
- Message, sketches and plan
- What I’d say on the first screen, a few rough sketches, and the order I’d do the work in.
02 · Audience
Who Yendo is for
This is what I could find publicly about Yendo’s customers, set against what the site shows them. The customer facts come from the sources listed at the end. The gaps are my reading.
Yendo doesn’t publish demographics, so I used Pew’s research on title-loan borrowers as a stand-in: typical income around $30,000, borrowing mostly to cover household bills. The personas further down are built on that.
Segments
| Segment · product | Needs | What Yendo offers | Gap |
|---|---|---|---|
| Car owner with limited credit · Auto Card | Access to credit, a better score, keeping the car | Higher limits than typical subprime cards, reporting to three bureaus, a low minimum payment, hardship programs | The first screen sells limit, speed and cashback. Building credit and keeping the car come further down. |
| Urgent need for cash · Personal Loan | A specific amount, quickly, with a predictable payment | Same-day funding and 60 fixed payments | Loans start at $3,100 on a single 60-month term, more than many people in a pinch need. |
| Still paying off the car · Auto Card + Refinance | A lower car payment and some available credit | Refinancing of the auto loan together with a card | The site says little about what happens to the existing car loan, and the product’s name keeps changing. |
| Homeowner · Homeowner Card | Modest credit without putting the home at risk | Up to $2,000, no appraisal, backed by fixtures | It’s secured by “fixtures”, and the site never says what that means in practice. |
| No asset, or moving on · Premier Rewards | A simple card and a way to progress | Up to $750, cashback, option to add an asset later | There’s no sign of how to move to other products later. |
03 · The visit
Walking through the site
Step 1 · Discover
The first screen
Most visitors land on the home page, a lot of them on a phone. In the first few seconds they get a photo, a rating, a list of benefits and a button.
What the visitor wants to know
Is this a legitimate lender, and is it for someone like me?
- Discover
- Understand
- Compare
- Decide
- Apply
The home page never mentions the car
What I saw
The mobile hero says “Credit cards built to move you forward”, then lists pre-approval in 1 minute, limits up to $10,000, money in 2 minutes and 1.5% cashback. Nothing on the first screen says the credit is based on your car, even though that’s what makes Yendo different and what lets people with bad credit get approved. The browser tab title, “Credit Powered by What You Own”, says it. The hero doesn’t. Building credit, which is why most customers apply, shows up halfway down.
Why it’s a problem
Someone with damaged credit who reads a generic credit-card headline may assume they won’t qualify and leave. Those who stay find out about the lien later, and at that point it feels like a catch. It’s actually the reason they can get approved.
What I’d change
- Say it up front: credit based on the car you own, and you keep driving it.
- Put bureau reporting (“builds credit with all three bureaus”) on the first screen.
- Test the new headline against the current one, looking at hero clicks and drop-off at the lien explanation.
- Heuristics
- H2 Match with the real worldInformation scent
- How to tell it worked
- Hero click-through, drop-off at the lien section, and how many applicants say building credit is their goal.
Each page has a different hero colour and logo
What I saw
The home hero is a near-black photo. The Auto Card hero is burgundy. The Personal Loan hero is light blue-grey. The logo shows up in orange, white and black. Below the fold all three pages share a light cream background. The darkest version is on the home page, which most people see first.
Why it’s a problem
A colour per product is common and fine on its own. Two things bother me: the logo changes colour with the hero, which makes the brand harder to recognise from page to page, and the darkest first screen is the one most people land on. I don’t know whether a dark hero lowers trust for a car-backed product. The research on colour and judgement calls itself early-stage, so I’d test it before changing anything.
What I’d change
- Define theme tokens once: a light default for first screens, with product colour as an accent on card art and highlights.
- One logo treatment per type of background.
- A/B test the home hero on a light background, measuring button clicks and bounce.
- Heuristics
- H4 ConsistencyBrand coherence
- How to tell it worked
- Home bounce, hero click-through, and a one-question trust survey on exit.
On mobile, the note about your credit score is almost invisible
What I saw
“Checking your offer does not impact your credit score.” sits under the main button in small grey italics, straight over the photo. Where it crosses the flowers I could barely read it in the screenshot. I can’t give an exact contrast ratio from a compressed image, but it’s well under the 4.5:1 WCAG AA asks for. The same sentence shows up under the calculator button and when the application opens from the home page, so visitors do get it later. The wording also switches between “does not impact” and “does not affect”, sometimes on the same screen.
Why it’s a problem
For someone worried about their score, this sentence removes the main reason not to tap, and the home hero is where most people would see it first. When legal-sounding text keeps changing its wording, careful readers stop and re-read it.
What I’d change
- Put it on a solid background at 4.5:1 or better, upright, at body size.
- Try a small label attached to the button: an icon plus “Soft check only. Your score isn’t affected.”
- Pick one approved wording and store it as a content token.
- Heuristics
- WCAG 2.2 · 1.4.3 Contrast (Minimum)H8 Aesthetic and minimalist design
- How to tell it worked
- Hero click-through on mobile.
The mobile hero is misaligned and partly covered
What I saw
On the mobile hero the headline is centred and the rating, bullets and note are left-aligned. In the benefits list, centred pills of different widths leave the check marks at different horizontal positions. A fixed “Feedback” tab covers the end of “Unlimited 1.5% cashback with autopay”. The inline “Check your offer” button and the sticky “Check your offer” bar can be on screen together.
Why it’s a problem
People scan lists down the left edge, so a ragged edge slows them down. The feedback tab hides part of a benefit, and two identical buttons on one screen compete with each other.
What I’d change
- Left-align the hero and lists on mobile, with the check marks in one column.
- Keep the feedback tab off content on small screens.
- Hide the sticky bar while an inline button is visible.
- Heuristics
- Gestalt: continuityH8 Aesthetic and minimalist design
- How to tell it worked
- Hero and benefits clicks on mobile, and how clicks split between the inline and sticky buttons.
Already working
The main button is obvious and repeated down the page, and each one has a note saying the check won’t affect your score.
Step 2 · Understand
Checking the proof
Before reading how anything works, careful visitors check the ratings, press logos and headline numbers, and compare them with what they’ve seen elsewhere.
What the visitor wants to know
Can I rely on these numbers?
- Discover
- Understand
- Compare
- Decide
- Apply
The ratings and download numbers don’t match
What I saw
The home hero shows “4.8 average rating | 555k+ downloads”. Lower on the same page, the app block says “4.8 average rating | 115k+ downloads”. The Personal Loan page shows Trustpilot at 4.2 out of 5 from 858 reviews, and “4.8 · 555k+ downloads” in its footer. The 4.8 has no source. The video testimonials use first names and note that customers were paid.
Why it’s a problem
A visitor who notices has no way to square these numbers. An unsourced 4.8 next to a sourced 4.2 makes both look less trustworthy, and two download counts on one page suggest nobody owns these figures.
What I’d change
- One social-proof component with a single data source, showing source and date on every figure (“4.8 on the App Store · Oct 2026”, “4.2 on Trustpilot · 858 reviews”).
- Show the app-store and Trustpilot ratings together, labelled, wherever either appears.
- Add a short written quote to each video testimonial so people can skim them.
- Heuristics
- H4 ConsistencyCredibility
- How to tell it worked
- Scroll depth past the proof block and clicks on the button right after it.
The headline numbers are maximums, and a couple disagree
What I saw
The home hero says “Credit limits up to $10,000” for cards whose stated range is $450 to $10,000, and the Homeowner and Premier Rewards cards top out at $2,000 and $750. The loan calculator slider starts at $3,000, while the small print says loans start at $3,100.
Why it’s a problem
A maximum sets an expectation most applicants won’t reach, and small mismatches make every number look approximate.
What I’d change
- Show typical outcomes where they exist (the “$6,000 average loan” on the Personal Loan page is a good example) and put the range next to the maximum.
- Pull every figure from the same configuration the product uses.
- Heuristics
- H2 Match with the real worldCredibility
- How to tell it worked
- Offer acceptance, and complaints about the limit offered against what people expected.
Already working
Trustpilot is cited with its review count, the footer rating has a date, and the site says the reviewers were paid.
Step 3 · Understand
Understanding the product
The visitor scrolls the home page and opens a product page to find out how a car-backed loan works, what they need and what the lien means.
What the visitor wants to know
How does this work, and what happens to my car?
- Discover
- Understand
- Compare
- Decide
- Apply
“Is my vehicle at risk?” is answered at the very bottom
What I saw
“Is my vehicle at risk?” is the first FAQ question, at the bottom of the home page. The Personal Loan page has specific commitments (no tracking devices, no starter-interrupt devices, a repayment plan before any action, lien released within 10 business days), but only far down the page.
Why it’s a problem
With a car-backed loan, the first thing most people want to know is whether they can lose the car. The risk is real: press coverage says a car can be sold at auction if Yendo can’t reach the customer for a long time. Yendo already has good answers, and at the end of the page most visitors have decided before they get there.
What I’d change
- A short “Your car stays yours” block under the hero button, with the commitments and a link to how liens work.
- The same block on every product page.
- Heuristics
- H10 Help and documentationH6 Recognition rather than recall
- How to tell it worked
- Hero click-through, and FAQ opens, which should drop if the block does its job.
Product names keep changing
What I saw
The page is titled “Personal Loan”. The phone mock-up in its hero shows “Auto Equity Loan”. The menu says “Auto Card + Refinance” and the comparison table says “Auto Card + Refi”, which not everyone will read as refinance.
Why it’s a problem
When the name changes, people wonder whether the terms change too.
What I’d change
- A naming glossary owned by content design and applied through the design system.
- Heuristics
- H4 ConsistencyH2 Match with the real world
- How to tell it worked
- Search and support queries that mix product names.
The Personal Loan page uses three content widths
What I saw
Measured on a 2,560 px screen, the left edge of the content jumps between about 90 px (full-width use-case cards), 290 px (“What happens to your title”) and 530 px (“What you need”) from one section to the next. On a phone everything collapses to one column and the problem mostly goes away, so it only affects desktop readers.
Why it’s a problem
Every time the left edge moves, the eye has to find a new starting line. On a page about liens, terms and eligibility, a steady grid makes it easier to scan.
What I’d change
- One 12-column grid with one content width and at most one wide variant.
- Document the allowed widths as tokens and check them in code review.
- Heuristics
- Gestalt: alignment and continuityCognitive load
- How to tell it worked
- Time on page against scroll depth, and success in a five-person scan test (“find what you need to apply”).
Some icons don’t match their labels
What I saw
“Consolidate debt” has a car icon. “Make a repair” and “Plan a move” share the same credit-card icon. The icons are grey-blue on a beige card, so they’re also hard to see.
Why it’s a problem
When an icon repeats or doesn’t fit the label, the reader has to stop and work out what it means.
What I’d change
- One icon set with one meaning per concept, checked for 3:1 contrast.
- Heuristics
- H6 Recognition rather than recallWCAG 2.2 · 1.4.11 Non-text contrast
- How to tell it worked
- Design-system audit coverage.
Already working
The lien section is concrete: no trackers or starter-interrupt devices, a plan before any action, automatic release. The FAQ explains how this differs from a title loan, and “within 16 minutes of signing” is a real number.
Step 4 · Compare and Decide
Choosing between loan and card
Now the visitor compares the two car-backed products and tries the calculator.
What the visitor wants to know
Which one fits my situation, and what would it cost?
- Discover
- Understand
- Compare
- Decide
- Apply
The comparison table has no buttons and favours the other product
What I saw
Each need maps to a pill (“Personal Loan”, “Auto Card”, “Auto Card + Refi”) that looks like a button and does nothing. The section is there to help people choose, then gives them no way to act on the choice. On the Personal Loan page the Personal Loan pill is pale grey and the Auto Card pill is dark burgundy, so the page pushes the other product harder than its own.
Why it’s a problem
People get to this table when they’re ready to pick. With nothing to tap, they have to scroll to find the next step, and the colours may steer some of them away from the product the page is selling.
What I’d change
- Rebuild it as two columns (Personal Loan | Auto Card) with a check per row and a button under each column.
- Let each row expand to show the detail: rate, term, how repayment works.
- Give both products the same visual weight.
- Heuristics
- H4 Consistency (affordance)H3 User control and freedom
- How to tell it worked
- Clicks from the comparison by product, and the product mix of started applications.
The calculator never shows the total you’d repay
What I saw
The calculator shows $6,000, about $205 a month, 60 months and 35.70% APR. Sixty payments of $205 come to about $12,300 for $6,000, and that total isn’t on the page. The 5% origination fee and a worked example ($10,000 with a $500 fee, 60 payments of $341.44, about $20,500 in all) are only in the grey, centred small print under the calculator. It doesn’t say whether the fee comes out of the money you receive.
Why it’s a problem
People reading this closely care most about the total. A $205 payment looks manageable. Paying back about twice what you borrowed is a different conversation. Truth in Lending disclosures already include a total of payments at signing, so Yendo has the number. If it sits next to the estimate, nobody has to work it out on their own.
What I’d change
- Add “Total of payments” and “Origination fee” rows to the estimate, next to the monthly payment.
- Say whether the fee is taken from the amount you receive or added to the balance.
- Keep the worked example, at body size, left-aligned, right under the estimate.
- Heuristics
- H1 Visibility of system statusH2 Match with the real worldCredibility
- How to tell it worked
- Calculator-to-application rate, offer acceptance, and support contacts about total cost or fees.
Already working
APR and term sit right next to the estimate, and the small print under it has the fee, a worked $10,000 example and the average loan.
Step 5 · Apply
Starting the application
The visitor taps “Check your offer”. A new tab opens on apply.yendo.com and asks for a phone number and personal details.
What the visitor wants to know
Am I still on Yendo, and is this the offer I just looked at?
- Discover
- Understand
- Compare
- Decide
- Apply
The calculator’s numbers don’t carry into the application
What I saw
On the Personal Loan page the calculator starts at $6,000 and shows about $205 a month over 60 months at 35.70% APR, with the 5% origination fee in the small print. The “Check your offer” button opens a new tab that says “Requested loan amount $10,000 · 60 months · APR 15.45%”. The amount I picked is gone and the APR is different. Both values are hard-coded in the link (amnt=10000, apr=15.45).
Why it’s a problem
Within one click the visitor sees two sets of terms and can’t tell which one is real. With a loan, that makes people suspicious of the whole offer. Since it involves disclosed terms, product, legal and compliance should look at it. It might just be an old link.
What I’d change
- Pass amount, term and estimated APR from the calculator to the application and show them at the top: “You’re checking $6,000 · about $205/mo · est. 35.70% APR”.
- Make sure the APR in the application matches the one disclosed on the page the visitor came from.
- Take the fixed offer values out of marketing links and generate the links from one place.
- Add an automated check that compares the site’s disclosures with what the application shows.
- Heuristics
- H1 Visibility of system statusH2 Match with the real worldH4 Consistency
- How to tell it worked
- Bounce on the first application screen, phone verification completion, and support contacts that mention rate or amount.
The application looks like a different company
What I saw
The application opens in a new tab. First comes a pixelated logo and “Loading application” with no progress, then a screen with a different typeface, colours and layout from the site. Coming from the Personal Loan page it also shows Mastercard and Intuit Credit Karma logos (channel=CK in the URL), although the visitor came straight from yendo.com. F08 covers how it describes Yendo’s role.
Why it’s a problem
This is where people type their legal name, date of birth and phone number. If the page suddenly looks different, a careful visitor stops to check they’re still on Yendo, and some of them leave.
What I’d change
- Open the application in the same tab, in a shell built with the site’s tokens: SVG logo, same type, colour and spacing.
- Replace the loader with a branded screen that says what’s happening (“Checking eligibility. This won’t affect your credit score”).
- Show partner logos only to visitors who came from that partner, by referrer or UTM.
- Heuristics
- H4 Consistency and standardsH8 Aesthetic and minimalist design
- How to tell it worked
- Bounce on the first application screen, time to first field, and completion rate by entry page.
It isn’t clear who is lending
What I saw
The site talks as if Yendo is the lender (“Yendo vehicle-secured loan”). The small print under the calculator says Yendo credit products are issued by Cross River Bank, Member FDIC, under licence from Mastercard. The application headline offers “instant access to your funds with eligible banks and credit unions”, which sounds like a marketplace. On the mobile home page, “Issued by Cross River Bank, Member FDIC” is styled as a benefit pill, next to “Contactless payment”.
Why it’s a problem
On the home page the visitor’s first question is whether this is a legitimate lender. The site gives three different answers, and the line that actually answers it is dressed up as a feature.
What I’d change
- Write one plain sentence about who issues the credit and what Yendo does, have legal sign it off, and use it on the site and in the application.
- Put the issuing-bank line next to the first button, away from the benefits list.
- Change the application headline to match, and drop “eligible banks and credit unions” unless offers really are routed that way.
- Heuristics
- H2 Match with the real worldH4 ConsistencyCredibility
- How to tell it worked
- Support contacts asking who the lender is, and first-screen bounce on the application by entry page.
The application form doesn’t say how long it is or why it asks
What I saw
The application is one long page inside a fixed-height card that scrolls on its own: phone verification, amount, then legal first, middle and last name, suffix, date of birth and email, with no progress indicator. The SMS consent text is small and light grey.
Why it’s a problem
A card that scrolls inside a page that scrolls is awkward on a phone, and identity fields with no explanation make people hesitate. If they can see the steps, they can judge how much is left. This is also where the funnel is narrowest, and the page already looks different from the site (F02), so hesitation here costs more than anywhere else.
What I’d change
- Short steps with a progress indicator and a time estimate.
- One line under each sensitive field saying why it’s needed.
- Consent text at a size and contrast people can read.
- Heuristics
- H1 Visibility of system statusH5 Error prevention
- How to tell it worked
- Completion at each step and abandonment per field.
Already working
Coming from the home page, the application repeats that checking won’t affect your score, and lets you pick up an earlier application.
04 · Visitors
How it changes by visitor
Section 03 follows one generic visitor. To see how things change with the goal, I wrote proto-personas based on what the site says about its customers, grouped them by goal and entry point, and walked each group through the live site on desktop and mobile, up to the first application screen. I didn’t submit any data.
- A · Cash today
- B · Rebuild with a card
- C · Still paying off the car
- D · The researcher
- E · No car, has a home or nothing
Trust from 1 to 5 at each stage is my own estimate, from walking the site as each kind of visitor. I haven’t tested it with anyone.
Scenario A
Cash today
Personal Loan · 6 personas
- DiscoverSearch “loan using my car” on mobile3.0
- UnderstandLands on /personal-loan4.0
- CompareUses the calculator ($6,000 · 35.70%)3.5
- DecideReads “Which one suits you”; the pills are not interactive3.0
- ApplySelects the CTA: new tab, partner logos, $10,000 at 15.45%1.5
Where they’d probably leave · At the hand-off, when the amount and APR change.
Scenario B
Rebuild with a card
Auto Card · 5 personas
- DiscoverTaps a social ad on mobile2.5
- UnderstandDark home hero, general headline2.5
- CompareScrolls product cards and video reviews3.0
- DecideOpens “Is my vehicle at risk?” at the bottom3.5
- ApplyStarts the generic application3.0
Where they’d probably leave · In the first screen, before the car question is answered.
Scenario C
Still paying off the car
Auto Card + Refinance · 3 personas
- DiscoverArrives on /personal-loan from search3.0
- UnderstandReads it needs a car owned outright2.0
- CompareFollows “Auto Card + Refinance”2.5
- DecideMeets “Refi” and “Auto Equity Loan” on the way2.0
- ApplyUnsure which product they are applying for2.0
Where they’d probably leave · At eligibility, when names stop matching.
Scenario D
The researcher
Any · 4 personas
- DiscoverComes from a review site or Credit Karma3.5
- UnderstandSees 4.8 · 555k+ on the home hero3.0
- CompareSees 115k+ further down and 4.2 on Trustpilot2.0
- DecideSearches “Yendo vs title loan”2.5
- ApplyLeaves to compare elsewhere2.0
Where they’d probably leave · At the proof blocks, when the numbers disagree.
Scenario E
No car, has a home or nothing
Homeowner / Premier Rewards · 2 personas
- DiscoverSees “up to $10,000” in the home hero3.5
- UnderstandLearns the car is what unlocks it2.5
- CompareFinds $2,000 or $750 limits2.5
- DecideWeighs a small limit against effort2.5
- ApplyApplies with low expectations or leaves2.5
Where they’d probably leave · At the product cards, when the headline maximum turns out not to apply.
Ver todas as proto-personas
| Name | Age | Location | Situation | Goal | Entry point | Device | Path |
|---|---|---|---|---|---|---|---|
| Marisol | 34 | Phoenix, AZ | Home health aide, paid-off 2016 Corolla, 580 score | Cover a $4,000 dental bill | Google search | Android | A |
| Derrick | 41 | Atlanta, GA | Rideshare driver, car paid off, irregular income | Fix transmission to keep working | Google search | iPhone | A |
| Tanya | 29 | Columbus, OH | Retail supervisor, single parent | Deposit for a new apartment | Facebook ad | Android | A |
| Luis | 52 | San Antonio, TX | Contractor with a paid-off truck | Consolidate two store cards | YouTube ad | iPhone | A |
| Kendra | 38 | Tampa, FL | Medical biller, past late payments | Emergency vet bill | Google search | iPhone | A |
| Ray | 47 | Las Vegas, NV | Casino technician, credit union declined him | $8,000 roof repair | Credit Karma | Desktop | A |
| Aaliyah | 26 | Charlotte, NC | Thin file, owns her late grandmother’s car | First real credit card | Instagram ad | iPhone | B |
| Brandon | 31 | Indianapolis, IN | Bankruptcy discharged 2 years ago | Rebuild score | TikTok ad | Android | B |
| Grace | 58 | Tucson, AZ | Widowed, fixed income, 2014 Camry | A card for emergencies | Daughter sent the link | Tablet | B |
| Omar | 35 | Houston, TX | Delivery driver, wants autopay and cashback | Card for fuel | App Store search | iPhone | B |
| Jess | 23 | Denver, CO | Student, car gifted by parents | Build credit history | Reddit thread | Desktop | B |
| Paul | 44 | Detroit, MI | Still owes $6,000 on his auto loan | Lower his car payment | Google search | Android | C |
| Nina | 30 | Orlando, FL | Two years into a 6-year auto loan at high APR | Refinance and get a card | Facebook ad | iPhone | C |
| Earl | 61 | Birmingham, AL | Co-signed truck loan with his son | Cash for medical travel | TV / radio | Desktop | C |
| Sandra | 49 | Sacramento, CA | Office manager, careful researcher | Compare against her credit union | NerdWallet | Desktop | D |
| Mike | 39 | St. Louis, MO | Burned by a title loan in 2021 | Make sure this is different | Trustpilot | Desktop | D |
| Priya | 33 | Dallas, TX | Personal-finance follower | Check the catch before applying | Credit Karma | iPhone | D |
| Jamal | 27 | Memphis, TN | Gig worker, reads every review | Find the real rating | App Store reviews | Android | D |
| Linda | 55 | Albuquerque, NM | Owns a home, no car | Card for home repairs | Google search | Desktop | E |
| Tony | 36 | Chicago, IL | Uses transit, no car, 600 score | Any card with cashback | Instagram ad | Android | E |
05 · Positioning
The message on the first screen
The offer is credit based on something you own and keep using. The home hero doesn’t say that.
“Credit cards built to move you forward”
Could be any credit card. Nothing about the car, about owning it, or about how the credit is secured.
“Credit Powered by What You Own”
This one says how it works, and it only appears in the browser tab.
What visitors compare it with
- Title loans. The CFPB’s 2016 study reported a typical APR of about 300%, and that one in five single-payment borrowers had their vehicle seized.
- Secured cards, which require a cash deposit.
- Unsecured cards for limited credit, usually with low limits.
- Credit unions and banks, which may already have declined the applicant.
What sets Yendo apart, in the site’s own words
- The vehicle backs the credit line, and the customer keeps driving it.
- No tracking or starter-interrupt devices.
- A repayment plan is offered before any action as lienholder.
- No deposit, reporting to all three bureaus, issued by an FDIC-member bank.
For people who own a car and can’t get much credit elsewhere, Yendo turns the car’s value into a credit card or a fixed-payment loan, and they keep the car. They see the terms before they commit, and unlike a title loan the payments are fixed or revolving over a set term.
What to say first
- 01
What you own counts.
Evidence · limits based on vehicle value; average loan and typical limit shown next to the maximum.
- 02
Build credit that counts.
Evidence · reporting to Equifax, Experian and TransUnion; automatic limit reviews. This is the reason most customers give for applying.
- 03
Your car stays yours.
Evidence · no trackers, plan before action, automatic lien release. Placed beside the first CTA.
- 04
See your numbers first.
Evidence · the calculator’s numbers carried into the application, unchanged.
“Lower rates” needs a comparison
35.70% APR is low next to a title loan and high next to a prime card. Saying which one you mean makes the claim checkable.
06 · Sketches
Rough sketches
Rough sketches of a few of these changes. They show how things would behave. The visual design would come later.
- Yendo Soft check onlyYou’re checking a Personal Loan$6,000about $205/mo60 monthsest. 35.70% APRChange amountMobile phone numberContinue · step 1 of 4
- YendoUnder each hero CTAYour car stays yours
- You keep driving it. No trackers and no starter-interrupt devices.
- We talk first. If you fall behind, we work out a plan before anything else.
- Lien released automatically within 10 business days of your last payment.
How vehicle liens work - YendoLoan or card?You needLoanCardCash in your bank✓–Same payment monthly✓–Fixed end date✓–Use anywhere, reuse–✓Cashback–✓Check loanCheck card
Still paying off your car? See Auto Card + Refinance
07 · Plan
Order of work
Sorted by how much I expect each item to help applications, then by what stops the same problems coming back. Each item lists the finding it fixes and what to watch.
Now
2–4 weeks
Pass the calculator values into the application and remove hard-coded offer values
Watch · Apply step-1 bounceShow total of payments and the origination fee in the estimate
Watch · Calculator-to-apply rateOne sentence on who issues the credit, used on the site and in the application
Watch · Support contacts about the lenderSame-tab application with site tokens and a branded loading screen
Watch · Time to first fieldOne social-proof component with source and date
Watch · CTR after proof blockMove the score note into a readable chip beside each button
Watch · Mobile hero CTR
Next
1–2 quarters
Test a hero that leads with the car and with building credit
Watch · Hero CTR, later-step drop-offTheme and colour tokens, with a light first screen (A/B tested)
Watch · Bounce, hero CTR12-column grid and container tokens across templates
Loan vs. Card comparison with a button per column
Watch · Comparison CTR by product“Your car stays yours” block on every product page
Watch · Hero CTR, FAQ opensProduct naming glossary and content tokens
Application in steps, with progress and a line on why each field is asked
Watch · Step funnel
Later
2–4 quarters
Product finder: answer 3 questions, get Card or Loan with your numbers
New
Watch · Start rate, product fitDesign system v1 with documented component states and an owner
Watch · Coverage, defectsMonthly usability sessions with recent applicants
New
Watch · Insights shipped / quarterPayoff view in the app: how much a month clears the balance by a chosen date
New
Watch · Share of customers paying above the minimumShow how customers can move between products (unsecured, home-backed, car-backed)
New
Watch · Cross-product moves, retention
If I joined: the first 90 days
Days 1–30 · Listen
Understand the funnel
- Funnel data split by entry page and device, from first visit to signed offer.
- Ten conversations with recent applicants and five with people who didn’t finish.
- Support and complaint themes, with the CX and compliance teams.
- An inventory of the components, states and tokens that exist today.
Days 31–60 · Fix
Ship the first changes
- Ship the “Now” column, with engineering and compliance involved from the start.
- Start running two-week experiments on the hero and the hand-off.
- Agree theme, grid and container tokens with front-end.
Days 61–90 · Systemize
Document and maintain
- Design system v1 with documented component states.
- Test the comparison and the product finder with applicants.
- A weekly review with product, content and compliance, so disclosures and UI get checked together.
08 · Method
How this was done
Method
- Heuristic evaluation against Nielsen’s usability heuristics, the WCAG 2.2 contrast criteria and basic Gestalt principles.
- I walked through the home, Auto Card and Personal Loan pages and the first screens of apply.yendo.com in Chrome at 1,440 px, plus mobile captures, on 9 November 2026.
- Contrast was estimated from screenshots, so the values are approximate.
- I didn’t submit any data to the application.
- Severity weighs how many visitors hit the problem, how much it affects their decision and whether it keeps happening. Anything that changes what people understand about the offer, the cost or the lender ranks above visual inconsistency.
Limits
- One evaluator. Nielsen recommends three to five, so this should be repeated with the team.
- Most of it was done on desktop at 1,440 px with mobile captures, while most visitors use a phone. I’ve probably missed mobile-only problems, and F15 was measured at 2,560 px.
- I didn’t look at how eligibility is explained (which cars and titles qualify, existing car loans, which states). That’s the next thing I’d review.
- No access to analytics, research or the logged-in product. Severity is my estimate of impact, and none of it is measured.
- Personas and trust levels are my assumptions. They should be checked against funnel data, five moderated sessions per scenario and a one-question survey in the application.
- Anything touching disclosures (F01, F07, F08) needs product, legal and compliance review before anyone draws conclusions.
Sources
- 01Nielsen Norman Group: 10 Usability Heuristics for User Interface Design
- 02W3C: WCAG 2.2, 1.4.3 Contrast (Minimum) and 1.4.11 Non-text Contrast
- 03Elliot, A. J. (2015). Color and psychological functioning: a review of theoretical and empirical work. Frontiers in Psychology, 6:368
- 04Labrecque, L. I. & Milne, G. R. (2012). Exciting red and competent blue: the importance of color in marketing. Journal of the Academy of Marketing Science, 40, 711–727
- 05CFPB (2016). Single-Payment Vehicle Title Lending
- 06Fortune (Jan 2025). A new borrowing option: a credit card secured by your car
- 07Yendo press release (May 2026). Launch of the Yendo Unsecured Card
- 08Pulse 2.0 (2023). Yendo: $24 million Series A
- 09CardRates. Interview with Jordan Miller, Yendo CEO
- 10The Pew Charitable Trusts (2015). Auto Title Loans: Market practices and borrowers’ experiences
- 11Yendo: Home, Auto Card, Personal Loan and apply.yendo.com, reviewed 9 Nov 2026