Self-directed product

Little Something

Money, wrapped with care.

Little Something lets you give money as a gift you design yourself, with a photo, a background, a note and a reveal moment, so it still feels like a gift after it's opened. Money is often the right gift. But sent digitally, it arrives as a bank transfer with nowhere to put the effort, so it feels impersonal.

Research, product, design, business model: Gunay Built with Claude as my main tool Designed and prototyped, not launched

Walkthrough

90 seconds. A giver makes a gift for a friend, then the friend opens it on their phone.

How it works

A mobile-first web app for giving money as a gift. The giver makes the gift first and adds money second, so the effort comes before the amount.

For the giver

Make the gift

A title, a note, a photo, a background and an effect. The note can say what the money is for, or nothing at all. The app nudges ("Go on, get mushy") but never writes it.

Creating a gift with a title, photo and note

Add money

Then the amount and payment. Money screens stay quiet, with plain type and exact numbers.

Adding 50 euros with a choice of payment method

Preview and send

A last look at the gift exactly as the receiver will see it, then the giver decides how it goes out: a link through any chat or email, a printed card with a QR code that folds like a greeting card or slips into an envelope, or scheduled for later. More money can still be added here.

Preview of the gift with print, schedule and add money

The gift arriving in a chat with its own preview

Printing the card, folded in half or in quarters

For the receiver

Open it

No app and no account. The link opens on a sealed reveal, then the look, the note and the amount.

The opened gift with photo, note and amount

Move it to your bank

The money waits in the gift until they're ready. How they spend it is up to them, and the giver only ever sees "Opened".

Confirmation that the money is on its way

Also in the product

A home screen where givers see their drafts and the status of each gift (Draft, Ready to send, Sent, Opened), and an optional Receiving tab where receivers keep their gifts.

Left out on purpose

Group collection, brand-locked spending and AI-written notes.

TL;DR

A frustrating group gift made me ask why there's no good way to give money that still feels like a present. I took that question from first research to a designed product, with a business model and a plan to measure it.

Three calls shaped it

Compete on the experienceI cut group collection, my original idea, once research showed it was already solved.

No account to start, open or receiveGiving should be as easy as a transfer, and the receiver shouldn't owe anything for it.

Launch with a bankA standalone app owned by a bank that wants to be part of people's everyday lives, starting with gifts from the Netherlands to friends and family elsewhere in Europe. A startup would need about 8,000 gifts a month just to cover fixed costs.

The team would steer by gifts opened, and the bank would judge it by what each new customer costs. Before launch I'd test it with research panels, and I'd build it so every feature can be tested from day one. Next, I'm building it in Claude Code.

Part 2

How I got here

Why I started

It started with a friend's birthday. A group of us were chipping in for an expensive check-up she wanted, and I was the one collecting the money. A bank transfer was the most practical option.

But sending it straight to her bank account felt inconsiderate, even though it was what she wanted.

It would merge with everything else in her account, and I know from experience how easily gift money gets spent on other things. Once that happens, the reason behind the gift is lost.

How I handled it

SearchedOnly gift cards and prepaid cards. Nothing let me give digital money as a present.
Made a cardDesigned it in PicsArt and printed it for the party.
Sent the moneyA separate transfer that evening. It only worked because I had her bank details.

Why this problem

I kept wondering why nothing like this existed. When I decided to do a self-directed product case study for my move into product management, I didn't want to make up a problem. I already had one I had lived through, so I started there: what exists, and why is this gap still open?

The problem and who it's for

Many people want money as a gift and appreciate the freedom it gives, and 49% would like to receive digital cash but feel weird asking for it (Wells Fargo/Ipsos, 2025). The problem is how it's presented and perceived in a social context.

In person, we put money in an envelope or a card and say why we're giving it: "Go treat yourself" or "This is for the guitar you wanted." Digitally, the same gift becomes a bank transfer or a payment link, and there's nowhere to put the effort. That's when a money gift starts to feel impersonal. In the same survey, 57% said a digital gift can feel like no effort went into it.

Today people work around it. Brand-locked gift cards are more intentional but take away the freedom, and prepaid cards feel thoughtless on their own. Both are easily lost, left at home or forgotten. Some people improvise, folding cash into origami or printing their own vouchers.

The survey is a US sample of 2,010 adults: evidence for the category, not proof it works in Europe.

Givers

Defined by the moment, not by age: they've already decided money is the right gift and want it to carry effort. For example, when they:

are close to someone but too far away to hand over a gift. This is where I'd start.
want to put money toward something specific, like a camera.
know someone well enough to give a gift, but not well enough to pick one.

Receivers

Friends and family, whatever bank they use. They never need the app: the gift arrives as a link and lives on their phone, so it can't get lost. The money is theirs to spend on anything, whenever they're ready, and the note and photo stay as a keepsake.

Not for

People who love choosing gifts and know exactly what someone wants, or givers who want to control how the money is spent. Little Something is another option for the moments when a present doesn't fit.

What I learned

Four interviews, about ten Reddit threads where people debate money as a gift, the Wells Fargo/Ipsos survey and a competitor map. Four findings shaped the product most.

The demand is bigger than I expected

Found: On Reddit, people openly ask how to give money without it feeling impersonal.
Changed: Confidence the problem was worth solving, and that awkwardness, not the money itself, is what to design for.

Nobody owns the giving moment

Found: Collecting and moving money are already solved, but nobody owns the giving moment. Could an app still stand out there? Partiful did it in a solved market by making the experience fun. And in Asia, money gifts like Hongbao and wedding envelopes are already a ritual, and Hongbao has moved into apps.
Changed: I positioned on the experience: the giver designs the gift before adding money.

AI writing the note would break the gift

Found: One interviewee said AI help with the message would remove the very effort the gift is meant to show.
Changed: The app prompts but never writes for the giver. AI can help behind the scenes, never as the voice of the gift.

Gift cards get forgotten

Found: Every interviewee had forgotten or lost a gift card, or never found the time to use it.
Changed: The gift lives on the receiver's phone, with an optional reminder.

What my research can't tell me yet

I spoke to a few people I know, most of my sources are qualitative, and the one quantitative survey is from the US. I don't have European numbers or a market size. I also can't tell what amount people are comfortable giving. €15 in money can feel small, while a €15 book feels thoughtful because nobody checks the price. People might also say they'd use the product and then hesitate when it comes to sending €25.

Next I would test real amounts with the working product, look for European data on gifting money, and ask receivers and givers how they feel about privacy.

Key decisions and trade-offs

Compete on the experience, not on moving money

The question

After the market scan, I asked myself whether this was worth building at all. The market looked crowded. But every product I found sat in one of two places: flexible but impersonal, or personal but tied to brands. Nobody offered both.

Why

Moving and collecting money are solved. What isn't is everything around it: to give money with a few words today, you send a transfer and then work out where the words go, in a chat, a call or a line in the transfer description.

What I chose

To compete only on the experience of giving: the intent and the reveal. Moving the money stays with whoever owns the product, a bank in my lead path.

What I gave up

Group collection. My own story started with a group gift, but Tikkie, bank payment requests and Collection Pot already do it well, so building it would mean competing where I have no edge.

What would change my mind

Testing that shows a money gift still reads as low effort even with a look and a note, or that it only works where money gifting is already a ritual.

No account to start, open or receive

The question

Where should sign-up sit? Most money apps ask for an account first. If I'm competing on experience, giving has to feel at least as easy as a bank transfer, and receiving easier than a gift card.

What I chose

Givers start as a guest and log in only when they add money. Receivers never need an account: they open the link and confirm a quick code to move the money. An account is offered later, as an option, to keep their gifts in one place.

Why

For givers, it's adoption. People drop off when asked to sign up for something they don't understand yet, but once they've made the gift, they're likely to sign up to send it. For receivers, it's the moment. Handing over your details turns a gift into an obligation. And a receiver who enjoyed it is the most likely next giver.

What it costs

Fewer accounts and less data on receivers. Money still needs a check when it moves, so no account never means no verification.

What would change my mind

Guest drafts that rarely get finished, a bank that needs more than a code to pay a receiver, or an A/B test showing an account up front doesn't hurt sign-ups, gifts sent or receivers becoming givers.

The giver sees "Opened", and nothing after that

The tension

Givers hope a gift gets used the way they meant it, but the point of money is freedom. If someone gave you money for a guitar and a friend then passed you their old one, you should be able to spend it on something else without explaining yourself.

What I chose

"Opened" and nothing after it, the digital version of seeing someone unwrap your present. Everything after that belongs to the receiver.

What it costs

The giver can't see when a gift sits unused, and forgotten gift cards came up in every interview. For now the receiver sets their own reminder, and the giver can always just ask.

Where it fits

The need is real, but money gifting on its own makes very little money. What's expensive is the setup around it: on my own, a licensed payment partner's minimums mean about 8,000 gifts a month just to cover fixed costs (my estimate from published pricing), before building trust from zero. What's different is the experience. So the question is who it's worth the most to: a company that already moves money and wants a reason for people to choose it.

My lead

A standalone app owned by a bank that wants to be closer to people's everyday lives. Think of a bank that's strong in mortgages and business banking but rarely the app people open for everyday money. For that bank, Little Something is mostly a brand play: a warm moment that shows it's there for people, not only for big decisions. The app keeps its own brand and works with any Dutch bank, while the bank holds the licence and moves the money, the way Tikkie works for ABN AMRO.

First audience

People in the Netherlands sending gifts to friends and family elsewhere in Europe. It's the strongest giver moment I found, and euro to euro needs no currency exchange and no new rules.

What the bank gets

The giver pays, never the receiver, and the core experience stays free. Money comes from around it: brand offers tied to what the gift is for, and later a small giver fee. The bigger gain is reach beyond the bank's own customers and a reason for receivers to open an account, measured as new accounts per gift against what the bank usually pays to win a customer.

Other owners, and when they'd make sense

Owner	Why it fits	What it costs the product	When I'd choose it
A money transfer app (like Paysend or Remitly)	Holidays and celebrations are a big reason people send money abroad: in a Remitly survey, over a quarter of senders send for these occasions. A gift experience gives people a reason to choose that app and pay its fee	Receivers may collect cash or use mobile money instead of a bank, rules differ per country, and there's pressure to compete on exchange rates	Next step, for gifts outside the euro
A social or messaging app launching its own payments (like X Money)	Gifting is how WeChat got people to link their bank cards: over 5 million joined in the first two days of its digital red envelopes in 2014. X launched X Money in the US in July 2026 as part of becoming an "everything app". A gift is a warm first reason to use money inside a social app	The gift lives inside someone else's app and look. In the West these features often struggle: Snapchat closed Snapcash in 2018, and LINE Pay closed in Japan in 2025 after merging into PayPay. WhatsApp payments run in India and Brazil, not in Europe	If a social platform brings payments to Europe
A feature inside a bank's app	Cheapest and fastest to launch, and trust comes with the bank	Reaches only that bank's customers. The app gets simpler: sent gifts live in the transaction list instead of a home screen, in the bank's look and language. Making the gift and the preview, where most of the fun is, still works	If no bank wants a separate app
My own startup	Full control of brand and roadmap	High fixed costs from day one, and trust built from zero	Only with investment behind it

The biggest risk

Separate bank apps don't always survive. Zelle shut down its standalone app in 2025 once banks offered the same thing inside their own apps. The ones that last do a job the main banking app can't. For Little Something, that job is the experience: the look, the note and the reveal.

Ruled out

Selling to businesses, since cash to employees counts as salary in the Netherlands, and earning from gifts nobody opens, which clashes with EU e-money rules.

Success metrics

The team would steer by one number, gifts opened per month, and the bank would judge the product by what each new customer costs.

Every target here is a rough estimate, anchored where I could on research into similar situations rather than identical apps. The sources and the math are in my metrics working notes.

Bank outcome

What each new account costs

(yearly running cost + cost per gift × gifts) ÷ new accounts

At low volume, running costs (team, support, marketing) are much bigger than the cost per gift, so each account gets cheaper as gifts grow. The product earns its place once this falls below what the bank usually pays to win a customer, a figure I'd get from the bank.

North star

Gifts opened per month

Opening is the moment both sides get something: the giver's effort lands and the receiver gets the gift. For a receiver who isn't a customer, it's also the first time they meet the bank. Gifts sent only counts the giver's half. New accounts is what the bank wants, but if the team chased it, the product would start asking for sign ups during the reveal, which goes against my decision on accounts.

Inputs

Each input is tied to a decision, so if one moves the wrong way, I know which decision to revisit. Gifts happen a few times a year, not every week, so the time windows are long.

Input	Decision it checks	Starting target (rough estimate)
Started gifts that get sent	No account to start	60%
Days from opening until the receiver moves the money	The forgetting problem, and how long money stays as deposits	Half moved within 7 days, under 5% unmoved at 90 days
Receivers who send a gift of their own within 12 months	No account to receive	10%
Givers who send another gift within 12 months	When a giver fee can switch on	20%
Gifts kept as a memory receivers who press save or share on the opened gift	Competing on the experience	25%
Non-customer receivers who open a bank account within 90 days the bank's side, multiplied by the north star rather than feeding it	The bank-owned path	2%

10% sounds small, but at about 3 gifts per giver a year, every 10 givers the bank wins bring about 3 more for free. Reminders could lift returning givers above 20%, as they do at Moonpig, where reminders drive four in ten orders.

Everything else, from drop off at each step to which looks people pick, is tracked without targets. The full list is in my metrics notes.

Guardrails

What must not get worse

Payment success rate stays at 98% or higher (assumption). A failed payment hurts more in a gift than in a transfer, because it breaks a moment.
Drop off at "move it to your bank" doesn't rise. It's the one step where the receiver is asked for anything.
The effort score doesn't fall, measured with a skippable one question survey after sending and after moving the money.
Support contacts per 1,000 gifts don't rise.

Stop signs

What would make me rethink

Fewer than 10% of givers send another gift within 12 months. People try it once and don't come back, so the experience isn't worth repeating and competing on it doesn't hold.
Fewer than 5% of receivers become givers in the first year. Growth would then depend on paid marketing, which weakens the case for the bank.
The cost per new account stays above the bank's usual cost. Then the feature inside a bank app or the transfer app partner becomes the better owner.

How I'd test it

Research panels before launch, then a product built to run experiments from day one.

Before launch: research at scale

Before there's real traffic, I'd test the prototype with panels of people living in the Netherlands, recruited through a research platform such as Maze or Lyssna. Each test is written up before it runs: what I believe, what I'll measure and what I'll do with each result. The pass lines are rough estimates.

Test	What it checks	Sample	Pass line
Unmoderated prototype test: make a birthday gift for a friend abroad	People can make and send a gift on their own	100 to 150	80% finish, ease score of 6 out of 7 or higher, no screen loses more than 10%
Sign in A/B test on the prototype: account first versus account at payment	No account to start	About 150 per version	Account at payment finishes at least 15 points higher
First click test on the WhatsApp message and the sealed reveal, in two versions	Receivers trust a link that says they got money	200	80% would tap the version written in the giver's own words
AI-moderated interviews right after the task	The gift feels like enough effort	50 to 100	Fewer than 25% would still message the person separately

AI runs the interviews and groups the themes, and I read the transcripts. I'd use AI-generated "synthetic users" only to check survey wording, never as evidence, because they invent plausible details no real data supports.

After launch: built to experiment

Day one

Experiment-ready

Every key feature ships behind a feature flag in a tool like GrowthBook or PostHog, so it can be tested or rolled back at once. Connected to Claude Code, an AI agent drafts each experiment and I approve it before it goes live.

The bank's gain

A holdout

A share of receivers never sees the bank account offer. The difference between the groups shows how many accounts the gifts create, rather than accounts people would have opened anyway.

Reading results

Bold tests, read honestly

Early traffic can only detect big effects, so I'd test real changes and read them with sequential testing instead of stopping at the first good day. Small choices, like the order of occasion templates, go to a bandit that shifts traffic to what works on its own.

Regulated

Compliance first

Anything that touches fees, verification or payment limits gets a compliance review before it's tested.

My process

I built Little Something alongside my job search.

DiscoveryMarket scan, interviews

DefinitionPRD, stories

Design13 screens

BrandName, look

Business modelLaunch path

BuildNext, in Claude Code

Market and desk research ran through discovery, definition and design

The phases overlapped more than a timeline suggests. Research, assumptions and interviews ran side by side, each one sharpening the others. The PRD kept changing after design started: the home screen and draft saving only became must haves once I saw the screens. And I designed in grayscale first, adding the brand only once the flow and every feature were designed.

What I'd do differently

I left the business model until the product felt clear. Looking back, deciding the launch path earlier would have changed things. Knowing it would be bank owned, I wouldn't have gone as deep into payment vendors, I would have checked the rules for building inside a bank sooner, and I might have put digital wallets like Apple Pay and Google Pay, and cards, first in the design.

The lesson I'm taking with me: the business model is an input to design, not something to add at the end.

Working with AI

Claude was my main tool, from research and documents in chat to the screens in Claude Design, and next the build in Claude Code. I ran the interviews and made the product calls. Claude did much of the desk research and drafting, and its most useful find was the Wells Fargo/Ipsos survey, the first hard numbers behind the problem.

What's next

Next I'm building Little Something as a working web app in Claude Code, with the code on GitHub. Before that, I'll run a quick test of the designs with a few friends, mainly to see what amounts feel comfortable and how receivers feel about the giver seeing "Opened". The build is my MVP: a play-money version I'll test with about ten real gift pairs. It's built to answer one question: does a money gift made this way feel like a gift? Alongside the build, I'm writing the go-to-market plan for the bank-owned launch.

The documents behind it

Working documents, lightly tidied.

PRDGoals, flow, requirements, decisions and priorities User storiesWith acceptance criteria Business modelLaunch paths, precedents, revenue, unit economics InterviewsGuide and synthesis Metrics notesBenchmarks, cost math, similar cases