Most banking apps look fantastic when the balance is healthy.

Sunlit gradients. Confetti for rounding up. Little fireworks when you tap “paid.”

Then the number hits £0 (or worse), and the polish falls off. Tricky language, greyed-out buttons, pop-ups with moral undertones. The UI that was keen to celebrate a £3 cashback suddenly becomes shy about fees, holds, and deadlines.

If a neobank hired me to fix their banking app UX at zero, here’s what I’d look for. And what I’d fix first.

Because bad days aren’t edge cases. They’re where banking app UX actually earns its keep.

What I’d Audit First: The £0 Experience

I’d create a test account, set the balance to £0, and spend an hour clicking through everything.

Here’s what I’d check:

Does the app state the balance plainly, or hide it?
Does it explain what “available” vs “pending” means, or assume I know?
When I try to make a payment, does it warn me about fees BEFORE I confirm?
If I’m overdrawn, can I see exactly what triggered the fee and when?
Is there an obvious path to hardship help, or do I need to hunt through three menus?

Most banking app UX design celebrates the good times and panics during the bad. The app that threw confetti for £3 cashback starts scolding you with “Uh-oh! Looks like you overspent!” when you hit zero.

That’s where trust breaks.

What £0 Really Looks Like (To a Human)

I’d design for these four forces explicitly, or they’ll design the experience for me:

Ambiguity: “Why is the payment pending?” “Why did the balance jump?” “Where did the refund go?”

Time pressure: Rent due Friday, salary on Monday. A weekend sits in the middle like a trapdoor.

Fear of penalty: Fees, interest, collections, credit score damage — usually explained after the damage.

Shame: The quiet dread of opening the app and being told off by a tone-deaf pop-up.

My job isn’t to cheer any of this up. It’s to reduce fear by making outcomes legible, options obvious, and next actions quick.

If I’m fixing UX for a financial product, clarity beats cheerfulness every time.

First Fix: Dignity at Zero

I’d treat £0 as a state, not a warning. The empty state isn’t a blank canvas for motivational quotes. It’s a message from the bank. Make it a professional one.

Say the number plainly.
“Available £0.00 • Pending £42.17”
No euphemisms, no hide-and-seek in three menus.

Name what will happen next (and when).
“Two payments scheduled for Monday. You can pause one.”

Offer one dignified action.
“Prioritise payments” or “See hardship options.”

No scolding.
I’d ban language that implies failure. “Uh-oh,” “Oops,” “Looks like you overspent” — delete all of it.

A calm empty state is product reassurance. It tells people they’re still welcome here, even at zero.

Second Fix: Negative Balances Without Theatre

Overdraft UX is where some teams role-play as moral philosophers. I wouldn’t.

I’d be exact:

Separate ledger states clearly:
Initiated → Pending → Settled. Show the bounds of the overdraft and what happens if a pending item settles first.

Explain fees at the moment of risk, not after:

“This payment would incur £6 today. You can:
– Pay now and accept the fee
– Schedule for Monday (no fee)
– Use your arranged overdraft (£127 remaining)”

Make receipts behave like evidence:
Every fee gets a line in the feed with plain description and a link to the rule. No FAQ safari. No PDF treasure hunt.

Clarity isn’t kindness. It’s legality and trust. That’s banking app UX in grown-up mode.

Third Fix: Hardship Flows That Work

If the neobank’s “help” route starts with a phone queue at lunchtime, they don’t have help — they have PR.

I’d build in-app hardship flows that:

Start from the empty/negative state. No hunt through “More” tabs.

Offer realistic options immediately. Payment pause, interest freeze, temporary limit increase, human callback.

Capture context once. Amount, due dates, income dates — the things a human will ask anyway.

Give an instant interim outcome. “You won’t be charged fees for 72 hours while we review.”

Leave a trace. Dated entry in the feed so users have proof if policy forgets later.

Use neutral language. No “hardship” banner that announces itself on shared screens. Privacy matters.

This is where good UX design stops being about interfaces and starts being about dignity.

Fourth Fix: Payment Prioritisation

When money is scarce, the UI must explain consequences, not assign blame.

I’d build a payment queue that:

Sorts by consequence, not alphabet.
Rent and utilities at the top. Gym membership at the bottom.

Explains the trade-off inline.
“Pause gym (no fee). Pay council tax (avoids £60 penalty).”

Allows one-tap re-order.
Drag handles that immediately show the new projected balance and fee exposure.

Includes a “why” link in English.
One sentence per item. No legalese. Just clear explanation users can trust.

You’re designing decisions under pressure. Fewer taps, more truth.

Fifth Fix: Fee Warnings at the Tap

If a tap can trigger a fee, the warning belongs on that tap.

Before confirming: “Transfer will complete Monday. Paying today costs £6.”

After settling: “£6 overdraft fee charged (rule link).”

Never: A polite monthly statement with an asterisked mystery charge.

I’d convert fee policies from PDFs into pages that load fast on 3G and link from the place they matter.

Hidden fees are UX debt. Fix them before they compound.

Sixth Fix: Credit-Build Nudges Without Shame

When a balance is often near zero, suggestions should feel like options, not judgment.

I’d offer the smallest helpful step:
“Set a £50 buffer alert.”
“Report rent to credit file.”
“Build a 3-month history with a £200 secured card.”

And explain the risk in the same breath:
“This may increase your credit utilisation and can lower your score short-term.”

Never celebrate debt. Confetti for borrowing is tone-deaf. A tick for on-time payments is enough.

Real help is specific and boring. That’s fine.

Seventh Fix: Notifications as Evidence

Push and email aren’t marketing channels during hardship. They’re part of the account record.

I’d make every notification:

Reconcile with the feed. Tap → takes you to exact transaction state.

Use ETA ranges, not fake clocks. “Expected Monday 09:00–12:00” is less pretty than a spinner, far more useful.

Switch to essential signals only. No “you might like” during hardship periods.

In banking app UX, certainty beats theatre. Every time.

The Metrics I’d Track

Time to certainty: From payment → “settled/failed” with explanation.

Fee visibility rate: Percentage of fee-eligible taps that showed pre-tap warning.

Hardship access time: From zero state → help confirmation.

Reversal/waiver satisfaction: A human reads these. The comments are the roadmap.

If these numbers improve, NPS takes care of itself.

The Tone Guidelines I’d Write

Neutral, not chirpy.
Specific, not motivational.
Outcomes, not adjectives.
Options, not lectures.

If a line would embarrass me to read aloud to someone having a bad month, I’d rewrite it.

This is basic product design work for financial products: treat users like adults, especially when things go wrong.

The Order I’d Fix Things

If the neobank only had budget for three fixes, here’s the priority:

Week 1: Fee warnings at the tap. Stops immediate trust damage. Quick win, high impact.

Week 2: Empty state dignity. Changes tone from scolding to supportive. Affects everyone who hits zero.

Week 3: Hardship flow. Helps users who need it most. Shows you’re serious about financial wellbeing, not just acquisition.

Everything else can wait. These three fixes stop users from leaving when they most need the product.

What Good Looks Like

A banking app with good zero-state UX:

States the number plainly (“Available £0.00”)
Explains what’s pending and when it clears
Warns about fees before you trigger them
Offers one obvious path to help
Uses professional language (no “Uh-oh!”)
Makes every notification reconcile with the feed
Prioritises clarity over celebration

The app doesn’t panic when you’re down. It stays calm and useful. That’s the product.

The Bottom Line

Most banking app UX is built for good days. Gradients and confetti and celebration.

Then the balance hits zero and the UI gets weird. Guilt trips. Hidden fees. Passive-aggressive notifications. Buttons that grey out with no explanation.

If a neobank hired me to fix this:

I’d audit the £0 experience first — every tap, every notification, every state.

I’d fix fee warnings (show them before the tap, not after).

I’d rebuild the empty state (professional, not scolding).

I’d add hardship flows (in-app, not phone queue).

I’d track certainty, not engagement (time to settled/failed, fee visibility rate).

Bad days aren’t edge cases. They’re where users decide whether to trust you or leave.

Real banking app UX plans for empty, negative, and awkward states first. Party graphics later.

The rest is just decoration.