How I found out
An AI told me I think in systems. I'd never heard that phrase.
I looked it up. I was amazed. There was a whole field of study — systems thinking — about the exact way my mind had always worked. I'd spent years making decisions that other people couldn't always follow. Seeing connections they didn't see. Feeling frustrated when something "almost worked" because I could see the whole chain and they could only see the part right in front of them.
Systems thinking named it. And once it was named, I could think about my thinking — not just do it. Now I notice when I'm tracing a chain forward. I notice when two threads that seem unrelated suddenly intersect. I notice when I'm asking the second question when most people stopped at the first.
I'm sharing this because the framework is learnable. It wasn't born in me — it was always there, unnamed. Once I named it, I could use it deliberately. You can too.
Thinking simultaneously
Several threads run at once. The connections between them are the insight.
I don't finish one thought before the next begins. Three or four run at once. The good stuff is where they touch. If I write you a long message, the end is where I figured it out — read the end first.
In EKKO
Every decision in EKKO arrived together, not in sequence. Non-custodial reinforces non-gambling reinforces non-coercive reinforces people-first. Pull any one thread and the machine breaks.
First principles, always
Strip every convention back to its reason. Then ask if that reason is still valid.
Every convention is a question, not an answer. Why do they do it this way? Is that reason still valid? Usually it isn't. The gap between what's assumed and what's actually true is where the new thing lives.
In EKKO
When the AI built my budget builder it put the category on the left and the amount on the right. Standard.
Then I asked: "wouldn't that feel like housing is taking $500 from me? What if I swapped them? — wouldn't it feel like I'm giving $500 to live there?"
Same numbers. Two completely different feelings.
So I swapped them. The amount on the left, the category on the right. You're not being charged. You're choosing.
Everything interlocks
Legal, design, code, and culture are components of the same machine.
Most products are built in separate rooms. Legal says one thing, design says another, engineering ships a third. I build one machine. Every decision locks into every other one. If the UI contradicts the legal position, one of them is broken. Find which. Fix it. Don't ship a product where the parts argue with each other.
In EKKO
EKKO retains zero from the prize pool. That's a legal decision, a financial one, a trust one, a design one — all at once. Same decision, four angles. The machine holds because all four point at the same thing.
"What else?"
The second and third question is where the real answer lives.
I never stop at the first answer. I ask: what else? Then again. The first answer is the obvious one — that's why everyone arrives at it. The third is where it gets interesting. The question is free. Most people just don't keep asking.
In EKKO
Nutrition tool outputs macros. What else? The user needs to know what to eat. What else? Shop for it. What else? Make it a habit. One question. A whole category of tools.
Curiosity that pays out
Every screen should have one question the user desperately wants answered.
Curiosity is the strongest pull there is. But it only works if the answer's worth finding. Slot machines use curiosity to pull you down. I use it to pull you forward. Question → honest answer → user's life is better. Not better for the business. Better for them. That's the only pull worth building.
In EKKO
"What's left each week?" is the question EKKO is built around. Answer it honestly and the next step is a budget. Then a plan. Then a game. Then a habit. Every step makes the user's life materially better.
Cut the noise
Every extra thing on a screen costs the user energy they didn't agree to spend.
Cognitive load is what kills products. Every button, every duplicate input, every clever feature "someone might use" — it's all weight the user carries through your interface. So before I add anything I ask: does it earn its place? What breaks if I cut it? Usually nothing. So it goes. The test is simple — remove it. If nothing breaks, it shouldn't have been there.
In EKKO
The budget builder had a slider on every row — identical to the text input right next to it. Two ways to do the same thing. That's not flexibility. It's noise. The slider had to go. Five minutes after I removed it, the product was better.
Think at system level. Act at specific level.
Never touch a part without first holding the whole.
Before I touch anything I walk the whole system first. What does this affect? What breaks upstream? What gets created downstream? The specific change is the LAST thing I think about, not the first. Skip the system level and you solve one problem while creating three you can't see yet.
In EKKO
Adding the meal planner meant updating the nutrition handoff, adding it to the discovery grid, and thinking about what comes next — shopping list, habit check-in, progress tracker. It's not a tool. It's a node in a chain that runs four more steps.
Build for the people who haven't been served
Three quarters of the world thinks collectively. Finance tools serve the other quarter.
Every personal finance app since the 1920s assumes money is individual. But most of the world — Māori, Pasifika, collective cultures globally — decides money as a household. That's billions of people no tool actually fits. So I build for them, not the saturated market.
In EKKO
Kotahitanga isn't a feature. It's the structural premise. Solo tools exist to feed the group decision. Five people reach a house deposit in 18 months instead of 8 years alone. That's the whole point.
I innovate. I don't replicate.
Convention is always a question. Never an answer.
I don't look left or right when I build. What others built tells me what's been tried — not what's right. I read Rich Dad Poor Dad. Went searching for income assets. Rentals: $100k upfront. Investing: 30 years of risk. Private lending: still needed savings first. Bank savings at 2% OCR: meaningless. The system was built against ordinary people. So I went my own way. Most founders add to what's there. I'm not interested in increments. I'm interested in what's true.
In EKKO
EKKO has no AI in the product. Everyone else is adding AI everywhere. I built without it. The game engine is deterministic and fair. AI is neither. The trust this product requires can't be probabilistic.
Worth saying out loud
I'm not a behavioural scientist.
Not a psychologist either. Never studied either field. But every decision in EKKO starts with the same question I asked about the budget builder: what does this feel like to read? What does it make me want to do?
That's not a method. It's paying attention to my own reaction. Catching the moment a screen makes me feel passive when it should make me feel in control — and asking why.
A behavioural OS doesn't need a behavioural scientist. It needs someone willing to notice. You can do that.
The decisions
Thinking in motion.
These are real decisions I made building EKKO. Not the conclusions — the full chain of questions that got there.
Why the budget builder has the amount on the LEFT and the slider moves TOWARDS the category.
Every budget app I'd seen put the category on the left and a number on the right. The user types in what they spend. Each row feels like the system taking money away from them.
Why does the typical layout feel bad? → Because the visual flow goes from "your money" to "what gets taken." The category is the destination. You're losing.
What if I flipped it? → Put the amount input on the LEFT, the category on the right. The slider moves left-to-right — from you, toward the thing it serves.
Why does direction matter that much? → Western reading flow is left-to-right. The eye treats the left side as the actor and the right side as the recipient. Whoever is on the left is doing something.
So if the amount is on the left and the category is on the right... → The user is the one giving. The category is what they're giving to. Housing isn't taking my money. I am giving money to my living.
What about the slider? → It moves TOWARDS the category. The motion itself is giving, not taking. And dragging it back to zero is instant — a one-second reset, not a punishment.
Same numbers, same categories, same math. Different direction. Completely different psychological reality.
The user should never feel defeated. The design has to make sure of it before any copy even tries.
The conclusion
The budget builder is built around the agency frame: I am giving money to my living. Not: my living is taking my money. Same data, same outcome — but one frame makes you feel like the author of your life and the other makes you feel like its victim. The frame is the product.
Why EKKO has no AI in the product when everyone else is adding it.
It's 2025. Every product is adding AI. The obvious move is to personalise savings targets, give AI coaching, predict outcomes. I chose not to.
What would AI do inside EKKO's game? → Personalise weekly targets. Advise on savings. Adjust difficulty.
What does the game require to earn trust? → That the rules are equal, fixed, and verifiable. Everyone plays the same game.
Is AI output deterministic? → No. Same input, different outputs. It varies.
Can you verify that AI is fair? → No. It's a black box. You can test it, but you can't prove it.
If the game uses AI, can I tell users it's fair? → I can tell them it's trying to be fair. That's different from fair.
What happens to trust when fairness is probabilistic? → It erodes. Users start wondering. The doubt compounds.
What's the alternative? → Pure functions. Same input, same output, always. Auditable. Provable.
The conclusion
No AI in the game engine. Not because AI is bad — because the trust this product requires can't be probabilistic. Determinism isn't a limitation. It's the design conviction that makes the whole thing hold.
Why the savings target escalates 2% per week — and why it resets.
I needed a number for the weekly target. Most savings apps just give you a fixed amount. I wanted something that genuinely teaches.
2% per week is fair and easy. Tiny increment. Anyone can do it.
But if you escalate forever, a $50 weekly starter becomes $1,000. Impossible. The user breaks. They quit.
So I capped it. The target escalates until it hits a percentage of current — say 80% above where they started — then resets. They start at $50 again.
Why a storm-like reset? Because the Arena is like a fortnight closing in. It's bearable. Everyone hits the breaking point at the same relative scale, not their absolute one.
The escalation forces a deeper behavior: "what if I can't make next week?" That question is gold. It teaches the user to think ahead.
That's why I built the stash savings account. Bag money for the escalated weeks or the rainy days. The mechanic is the lesson — you don't need a course on emergency funds, you just need to play the game.
The conclusion
The 2% escalation isn't just a difficulty curve. It primes users to think ahead, and the stash account gives them the tool to act on it. The game teaches forward planning through structure, not through lecture. That's the conviction: behavior change through mechanics, not advice.
Why lifelines exist (I call them echoes), and why exactly 3.
In a survival-style game, the obvious move is: miss the target, you're out. Make it brutal, make it dramatic. I didn't want that.
What's the actual win condition? → It's not the prize pool. It's the savings habit forming.
What kills habit formation? → Quitting before it sticks. Churn before week 4.
If I make MISS = elimination, what happens? → Most users get knocked out before the habit forms. The game wins but the user loses.
3 lifelines = an echo. Miss a week, you're still in. Miss another, still in. You've got room to be human while the habit takes hold.
Why 3 specifically? → It gets users to week 4 effortlessly. Week 4 is the habit-forming threshold. After that, they're saving because they save — not because the game makes them.
NZ savings rate is real bad. People don't fail because they're lazy. They fail because the system was never built to let them succeed.
The conclusion
Echoes exist because the prize pool is not the point. The savings habit is the point. Lifelines protect the habit from the game — and the game protects the user from the broken system. I built EKKO because reading Rich Dad Poor Dad sent me searching for income assets. Rentals needed $100k. Investing meant 30 years of risk. Private lending still meant saving first. And the 2% OCR on bank savings was meaningless. So I built the start — for everyone.
Why Kotahitanga is the structural premise, not a feature add-on.
Every personal finance app treats money as an individual problem. I asked a different question first.
I sat down with my own whānau. 4 of us. $250 a week each.
$250 × 4 = $1,000/week. $4,000/month. $48,000/year.
What would take me 4 years alone, we hit in 12 months together.
So why isn't every family doing this? → Because nobody's set up to. The tools don't exist. KiwiSaver locks you in individually. Banks won't let four people contribute to one savings account easily. The whole system assumes solo.
It's not that families can't. It's that the system was designed against it.
We could all have houses easily. Nice cars. Stability. But everyone is out for themselves — and it's not intentional, it's by design.
I am changing that. Not as a feature. As the structural premise. Solo tools exist to feed the group decision, not the other way around.
The conclusion
Kotahitanga is the architecture, not the feature. The math models the relationship. The collective number is always the hero. The chart shows what we already knew but never had the structure to act on: together is always faster.
Why the free tier is genuinely complete — not a funnel.
Most products gate the good stuff behind a paywall. The free tier is a teaser, a trial, a tease. I built the opposite.
A friend asked me once after I explained EKKO: "So you're actually doing this for the people?" There was a small wtf in her voice. Like the idea was unheard of.
I am. I genuinely am. Because I'm part of the people too.
It's us, not me. If they won't do it, I do it myself.
If a user never upgrades to Gold and still gets a better savings, budget, debt, and planning experience than anything they'd pay for elsewhere — I've won.
I haven't failed at conversion. I've succeeded at the kaupapa.
The upgrade earns its way through demonstrated value. Not through artificial limitation.
Some founders would ship right now and charge a 14-day free trial. Cake. That's extraction. That's the system I'm replacing.
The conclusion
Free means free. The free tier IS the product. Gold adds automation — bank connection, auto-transfers, the Arena game — but the tools work completely without it. Patience is a design conviction. I'm building the reinforced boat, not patching holes later.
Why there are no leaderboards or rankings in Kotahitanga groups.
Every social product I've ever seen adds leaderboards. Gamification 101. I removed them entirely.
A leaderboard says: compare yourself to others. That's the wrong frame.
Kotahitanga isn't about competing. It's about getting ahead together.
What if instead of showing each person's number, I show the collective number — what the group has saved?
In the mobile app, games are private matches. Members invited each other. They see the collective savings progress bar — read-only, no individual balances.
Why no individual balances? → Data privacy. But there's a deeper reason: their game, their conversation. They can ask each other directly. They don't need the app to tell them.
What does the collective number do psychologically? → It sparks the dinner conversation. "We have $10,000!! What should we plan for?"
That hits completely different from seeing "I have $1,000." One creates excitement and a shared decision. The other creates loneliness and self-judgment.
The conclusion
No leaderboards because the structure they create is wrong. The collective number is always the hero. The visibility goes through conversation, not through dashboard. The product is designed to make the family dinner the most important screen — not the app itself.
Why the prize pool pays out fully and EKKO keeps nothing from it.
I needed a revenue model for the game. The obvious structure: take a house cut from the prize pool, like every competition platform.
How does a house cut work? → Players compete. A percentage stays with the house. The rest pays out.
Does that align my interests with the players'? → No. I profit when more people play and when the pool is bigger — regardless of whether they win or lose.
What's the actual relationship I want? → I profit when players genuinely build savings habits. Not when they play. When they succeed.
What model creates that alignment? → Subscription fee for the service. I get paid the same whether they win the game or not. My revenue comes from the subscription, not the outcome.
If subscription is the revenue, where does the prize pool come from? → I fund it. $10 of every $20 subscription goes into the pool. Out of my own revenue.
Does this make me profit when users lose? → No. It makes me profit when users value the product enough to subscribe. They value it when it actually helps them save. The alignment is structural.
Is this legal? → A subscriber paying for a service, the provider choosing to fund a competition from their own revenue, is not gambling. It's a loyalty model. Legal opinion confirms.
The conclusion
EKKO retains zero from the prize pool. Always. Every dollar goes to the winner. My revenue is the subscription — and that subscription is only worth paying if users actually build the habit. So I only profit when the product genuinely works. That's the only business model that can't eventually become dishonest.
The invitation
This way of thinking is learnable.
You don't have to be a genius to think in webs. You just have to slow down enough to ask the second question — and then the third.
Every tool in EKKO is built from this. Start anywhere — each one will show you something true about your situation that you didn't know before.
Built by Tyson Harding · Immiscible Tech Limited · Auckland, Aotearoa