Finora, a money app that tells you where you stand

• Every app opens on a dashboard. None of them opens on an answer.
• Advice arrives with no reasoning, so there is no way to judge whether to trust it.
• Progress is shown as a balance, not as the thing you are saving for.
• Colour is used for decoration, so it stops carrying meaning.
• Every screen offers ten options. The hard part was never the options.

This is a concept project. Nobody briefed me and nothing shipped, so nothing on this page is a result. What it is instead is a set of decisions, each one with the reasoning attached, and a plan for what I would test first.

1. 01

   Monarch Money

   The most complete tracker. It tells you what happened, not whether it was fine.

2. 02

   Copilot Money

   The best at sorting transactions. It tidies the data rather than interpreting it.

3. 03

   Empower

   Asks the right question, on track or not, but only for retirement and against a benchmark.

4. 04

   Cleo

   The only one that talks first. The teasing tone stops working when the number worries you.

Research

I wrote the jobs around the moment someone opens the app, not around the features they might use. Each one starts with a feeling, because that is what actually makes a person pick up their phone and check.

The home screen holds the three things people check: what they have, what they are saving for, and what they spent. The AI sits in two places. Small notes appear on the screens you are already looking at, and a full coach screen waits for when you want to ask something. Small help first, then bigger help when it has been earned.

I mapped the five steps a new user goes through to find where confidence drops. It drops in the middle, at the moment the app finishes reading your accounts and shows you the result. The numbers arrive, but nobody says whether they are good. That one moment is what the rest of the app is built to fix.

Identity

One purple does two jobs: it marks anything the AI said, and it marks anything you can tap. Green, red and amber mean money is going well, badly, or needs a look, and nothing else. There is a second, separate set of colours for chart categories like housing and transport. Those are labels, not judgements, and I keep them apart from the four above so that a colour never means two things at once.

One typeface, Manrope, everywhere. Its round shapes read as friendly rather than corporate, which matters on a screen showing someone their savings. Ten sizes, 34 down to 12, following the iOS scale so the app feels native. Nothing smaller than 12. Every figure uses tabular numerals, so a column of amounts lines up on the decimal point instead of wobbling.

Four columns on a 393 point screen. The margin is 16 and the gutter is 8, both from the 8 point scale, which leaves 361 points of content. The margin and gutter never change; the columns stretch to fill whatever is left. Every screen sits on the same four, whether a card runs the full width or splits across them.

Design

Eight screens. Each one is here for a decision rather than for the screen itself, so what sits above every image is the reason it went that way and what the choice gave up.

The home screen answers the question first and shows the numbers second. Net worth sits at the top, then one note from the coach in plain words, then goals. I cut the six cards I first drew down to four, because a home screen that lists everything is the same as a home screen that says nothing.

The coach answers questions in sentences, not charts. Every reply has the same three parts: what it looked at, what it thinks, and one thing you could do. It also shows how confident it is, which was the hardest decision here. A number that admits doubt is less impressive, and it is the only reason to believe the rest.

The planner shows what a change would be worth before it asks you to make it. Each recommendation carries its own number, so you can weigh the effort against the size of the gain. The projection draws two lines rather than one, your path now and your path with the plan applied, and shades the gap between them. That gap is the only thing being sold, so it is the only thing the chart emphasises.

Spending is the screen where an app usually makes you feel bad. So this one leads with what went right, spending down 8% on last month, before it breaks anything down. Categories are sorted by size, and subscriptions get their own card because that is where the money quietly leaks.

A goal is shown as a fraction, never as a balance. $15,600 means nothing; $15,600 of the $20,000 you need means you are nearly there. Each goal also carries a date and whether you are ahead or behind it, because that is the part people actually check.

The portfolio opens on one number and a split, not on a list of holdings. Everything else is one tap down. I built it for someone who owns six things and checks monthly, not for a day trader, which meant leaving out most of what investment apps put on this screen.

This is the one screen where Finora has to look like a real trading app, because that is what people expect when they tap a stock. The AI analysis sits under the chart rather than above it, so it reads as a second opinion on what you are already looking at, not as a recommendation to buy.

Insights is the app talking without being asked, which is the easiest way to become annoying. So it is limited to a monthly review and a short list, each item with a number attached to it. If a note cannot say what it is worth in dollars or in months, it does not appear.

1. 01

   Does the home screen actually answer the question?

   Show someone the home screen for five seconds, take it away, and ask them whether they are doing okay with their money. If they can answer, and their answer matches what the screen said, the main idea holds. If they start reciting numbers, it does not.

2. 02

   Does showing confidence build trust or damage it?

   The coach says how sure it is. My assumption is that admitting doubt earns trust. It might do the opposite and just make the advice feel weak. I would show two versions of the same reply, one with the confidence score and one without, and ask which they would act on.

3. 03

   Is one next step enough, or does it feel like being told what to do?

   Every screen ends with a single action. That is the principle I am least confident in, because the line between helpful and bossy is thin and probably personal. I would watch for the moment someone looks for the option that is not there.

4. 04

   Will anyone connect their bank?

   The whole app depends on step two, and step two is asking a stranger for access to every account you own. No amount of visual design fixes that on its own. I would test whether saying exactly what is read, and what is not, moves the number at all.

1. 01

   Writing the principles first made the screens faster to draw

   Four short rules settled most of the small arguments before they started. When I was stuck on a screen, it was almost always because two of them were pulling against each other, and knowing that was usually enough to unstick it.

2. 02

   Making the numbers add up changed the design

   I set the fake data so that it actually reconciles: investments plus savings equal net worth, income minus spending equals the savings rate. Doing that exposed screens where I had been laying out a shape rather than showing information.

3. 03

   A colour rule is easy to write and easy to break

   I wrote that green, red and amber only ever mean something about money, then used a category chart that spent them on housing and transport. Rules do not hold on their own. They hold when the palette makes the wrong choice hard to reach.

4. 04

   Designing my own brief is harder than being given one

   With nobody to say no, the scope kept growing and every idea looked reasonable. The jobs to be done ended up doing the job a client normally does, which is telling me what this app is not for.