Every founder started somewhere.

Stuck between worksheets and Roblox

Existing options sat at two extremes. Worksheets and school programs teach real concepts but lose a kid's attention. Tycoon games hold attention for hours but teach nothing that transfers. Kid debit cards like Greenlight teach spending, not earning.

LilPreneur takes the pull of a game and ties every decision to one real money concept. The kid plays; the playing is the lesson.

42 people, four bets

Before the developer pitch on Aug 18, the PMs interviewed 20 parents and 22 kids. Kid sessions were run with a parent present, which mirrored the product's own safety stance. The team framed the work as four connected bets, each with its own test.

My research: boys 7–12, and their parents

Jennifer and I split the kid research by age band (7–9 and 10–12). I took boys: how they ask for help, react to failure and feedback, build confidence, and take risks. I also researched what parents of boys want, and what makes a parent trust a learning app. Peer-reviewed backing came up as the biggest trust factor, which shaped how the landing page talks about safety and real founders.

Numbers come from rules, not AI

A key safety call shaped every screen: no live AI ever talks to a child. All outcomes come from a fixed set of formulas, so the same choices always give the same result. AI only works after the session, server-side, to pick badges and write the parent report, and a separate model checks that text before anyone sees it.

For design, this meant every choice had to be a clear, finite option. No typing, no open questions, and no number a 7-year-old has to calculate.

Every choice is visible in the scene

The stand is built from stacked layers, and each decision swaps one of them. Pick the park and the background changes; add a sign and it appears on the cart; a heat event tints the whole scene. Kids see cause and effect before they read a single number.

The layering also kept the art manageable. Quantity × flavor × wrapper × location × weather would otherwise mean thousands of hand-made scenes.

How a badge is earned

Badges are never handed out mid-game. After the day ends, every choice and outcome is saved to a log, and that log is checked against a fixed rule for each badge. A child earns a badge for a pattern of decisions that work together, so a lucky day alone doesn't earn one.

Badges appear on the Level Complete screen before the money question, so even a day that loses money ends on something earned. They also show up in the parent's report, as evidence of what the child practiced.

From testing: show what's still locked

The first build only showed the badges a child earned. Testers suggested showing the locked ones too, with a line like “2 more badges waiting, replay to unlock?” That turns the badge row into a reason to play again. Use the toggle above to see the locked state.

More in the collection

Customer Whisperer, Resilience Champion and Inventory Ninja appear on the landing page, showing parents the kind of skills the badge collection grows into.

One contrast rule shaped every button

Navy text on marigold reaches 6.3:1 contrast. White on marigold drops to 2.1:1 and fails. So every orange button in the app uses navy text.

I set the type pairing too: PT Serif Bold for headings, Inter 400/600 for everything else. A colleague described the early UI kit as approachable and fun, without feeling dumbed down for kids.

Handoff without a click-through prototype

By mid-September the Figma file held 140 frames. Instead of wiring a prototype, the team mapped every scene to its live frame, art folder and open questions in a handoff tracker. Older versions are labeled “superseded” so developers build only the live one.

We also pushed to build one componentized scene with team and weather as variables, instead of separate copies of the run-your-stand sequence for each helper.

How we're testing

Before families play, each team member owns part of the app in user acceptance testing. My area is the first impression: the website, sign-up and setup, the avatar, the business story, and the first decision screens. The checks: does it look polished, is it friendly, and is it easy for a child's hands and eyes on a tablet.

We also watch children play with a parent's consent, without helping, and note where they hesitate, laugh or get stuck. Each playthrough is scored with a skills rubric covering cause and effect, trade-offs and managing risk, so we measure learning, not just clicks.

What the sign-ups told us

The sign-up form asked parents their child's stage (7–8, 9–10 or 11–12) and which skill to master first: pricing, bouncing back from failure, spotting opportunities, or weighing trade-offs. Most parents picked all four. Their answers to “what do you wish schools taught?” clustered around money management and real-life projects.