I’ve started more than twenty businesses, brands, and projects since I was a kid — lawn care, apparel, e-commerce, a marketing agency, consumer products, a long tail of hustles in between. Most of them worked, in the small sense. They made sales. They served customers. Some of them ran for years.
And almost none of them deserved the years I gave them.
So here’s the direct answer to how to choose a business idea, learned the long way: filter by ceiling, not by floor. Before committing, ask one question — if I execute this brilliantly for ten years, can it become something enormous? A nine-figure company — $100 million or more? If the honest answer is no, the idea can still make money. It just can’t have your decade. That’s the filter, and these days nothing I build gets past the whiteboard without clearing it.
Hustles vs. companies
The distinction the filter runs on: a hustle sells your effort; a company builds an asset.
A hustle is a transaction loop — flip the product, ride the trend, run the pop-up, collect the spread. Nothing wrong with any of it; money is money. But stop feeding a hustle and it dies the same week, because nothing accumulated. There’s no brand anyone would miss, no system that runs without you, no customer who identifies with the thing. You didn’t build equity. You rented yourself out — to yourself.
A company compounds. Brand equity accumulates. Customers return without being re-acquired, then start recruiting other customers for free. Systems outlive the founder’s attention. Culture forms — and culture and community are the moat a factory can never produce, because anyone can copy a product and nobody can copy what people feel like they belong to. A hustle and a company can look identical in month three. By year ten they aren’t even the same species.
The years are the cost
Nobody warns you that the expensive part of a business isn’t the money. It’s the years. Money comes back. The five years you spend scaling something with a low ceiling are gone, and they were probably your sharpest five.
Run the arithmetic nobody runs: you get maybe four or five serious, decade-scale swings in a working life. That’s the whole inventory. And here’s the brutal part — a small game costs exactly the same years as a big one. Same stress, same grind, same missed weekends. The widget shop and the hundred-million-dollar brand both take a decade to find out. Only one of them can ever pay for the decade.
I know this one from the inside. Twenty-plus ventures deep, I looked up and realized I’d been an absolute machine at starting things and a terrible allocator of my own years — paying full price in time for businesses that were capped from day one. The filter isn’t business theory to me. It’s the receipt.
Why aim that high
The obvious objection: most businesses never touch $100 million, so isn’t the filter delusional? No — because the filter isn’t a forecast. It’s a constraint on the category of thing you build. It doesn’t predict where you land. It dictates what game you’re playing.
An idea with an open ceiling forces certain inputs from day one: a real brand, a product genuinely better than what people already use, a market with room, distribution that can compound. An idea with a closed ceiling forces commodity inputs — price wars, margin scraps, trend timing. You become what the game demands. So even when a ceiling-open idea lands short of enormous, it lands as something real: an actual brand, actual customers, an asset someone would buy. When a ceiling-closed idea lands short, it lands as a garage full of inventory and a lesson.
Build things to sell even if you never sell — real companies, real brands. The filter is just that sentence, with a number on it, applied before the years get spent instead of after.
What passes the filter
The traits, concretely. A product better than what you yourself already use daily — if you wouldn’t switch, nobody else will. A brand people join rather than merely buy, because belonging is the only moat that deepens with time. Demand that repeats by its nature, so growth stacks instead of resetting to zero every month. A market big enough that a decade of compounding has somewhere to go. And an operation that could, eventually, run and be valued without your face in every room.
Miss on most of those and the ceiling is structural — no amount of hustle raises it. Hustle harder inside a capped game and the prize is a slightly bigger cap.
Running the filter: five questions
To make it usable instead of theoretical, here’s the trait list converted into the questions I actually ask before anything gets on my calendar:
1. Would I switch to this myself, today, at full price? Not “would I use it if it existed.” Would I abandon what I currently use and pay for this instead? If the founder wouldn’t switch, the pitch is fiction.
2. If it disappeared in a year, would anyone miss it? Missed things have gravity — community, identity, trust. Unmissed things were interchangeable the whole time, whatever the revenue said.
3. Does demand repeat without being re-earned from zero? Businesses that restart at zero every month aren’t compounding; they’re sprinting on a treadmill with better branding.
4. Could it eventually run — and be valued — without my face in every room? If the answer is permanently no, I don’t own a company. The company owns me.
5. Is there a decade of room? Market big enough, trend-proof enough, real enough that ten years of compounding has somewhere to go.
Five honest yeses is rare. That’s correct — the filter is supposed to say no almost every time. Its whole job is protecting the four or five decade-scale swings you get, and a filter that approves everything is a rubber stamp, not a filter. Every “no” is the system working.
What the filter kills
Trend-surfing. Widget flips. Pop-up concepts. Margin-thin resale. Anything whose entire pitch is “easy money right now.” I’ve done nearly all of these, which is exactly why they’re named here.
And killing them stings, because here’s the trap: the dangerous ideas aren’t the ones that fail fast. They’re the ones that work just well enough to keep you. A failing idea releases you in six months. A capped idea that pays okay can hold you for six years — comfortable, busy, and quietly costing you the only swing inventory you’ll ever have. The filter exists to kill the second kind, because your own results never will.
The exception: your first business
One honest carve-out. The filter governs where your years go — it is not a reason to delay starting. Your first business is tuition, and tuition doesn’t need a nine-figure ceiling. Mine was a lawn company run out of the back seat of a sedan. It would have failed the filter on every line, and it was the best education I’ve ever paid for — sales, service, pricing, payroll, all of it learned on a machine small enough to survive my mistakes. Start with whatever you’ve got; learn the game on the small table.
The filter clicks on at the commitment: the moment you’re about to give something years instead of months. That’s when the question stops being “could this work?” — almost anything works — and becomes “does this deserve the decade it’s about to take?”
Most ideas are auditioning for your decade. Very few should get the part.
Choose like the years are the money. They are.
Tyk Sullivan
Serial entrepreneur in Tampa, Florida. Founder of Plutus Capital and NUMEN. More about Tyk →