The Best Startup Idea Might Already Exist. Here's How to Use It

Most founders believe the hard part of starting a business is having the idea. So they wait for one that nobody has ever had. It rarely shows up, and when it does, it usually turns out there was a reason nobody built it.

Here’s a less romantic way to look at it: the best idea for your next business might already be working for someone else.

Why are original ideas overrated?

An original idea comes with an expensive question attached: does anyone actually want this? You have to answer that question with your own time and money, and most of the time the answer is no.

A model that already works has answered that question for you. Somebody has proven that customers will pay, roughly how much they’ll pay, and what they expect in return. That isn’t a guarantee, but it removes the biggest risk most new businesses face.

Original ideas aren’t worthless. They’re just not the only path, and for a first business, they’re usually the slower one.

What did Rocket Internet actually do?

Rocket Internet, founded in Berlin in 2007 by the Samwer brothers, became famous (and a little notorious) for a simple approach. They looked for online businesses with real traction in the US and built versions of them for markets the originals hadn’t reached yet.

Zalando started with a model close to Zappos: sell shoes online with free shipping and free returns. Rocket and its founders also backed or built companies like Lazada in Southeast Asia and Jumia in Africa, which took the Amazon-style marketplace to regions where nobody was doing it well. Before Rocket, the Samwers sold CityDeal, a European take on group buying, to Groupon.

Their playbook went roughly like this:

1. Find a startup with real, measurable traction.
2. Rebuild the model quickly.
3. Localize everything: payments, logistics, language, marketing.
4. Out-execute on speed and hiring.
5. Sell or scale before the original company arrives.

It wasn’t glamorous, and plenty of people criticized it. But it built several companies worth billions.

Is adapting a model the same as copying?

No. Copying is taking someone’s product, brand, and content. Adapting is recognizing a pattern that works and fitting it to a context the original doesn’t serve.

The difference matters legally and practically. Legally, business models aren’t protected, but brands, code, and content are. Practically, a straight copy has no reason to exist. A good adaptation does: it serves a customer the original ignores.

What can you adapt besides geography?

Rocket Internet mostly worked across countries. Today, you have far more options. You can borrow:

• A pricing model. Per-seat pricing, usage-based billing, or a free tier with paid upgrades can work in industries that still sell annual contracts.
• A growth loop. Calendly spreads every time someone sends a booking link. What in your market could carry your product the same way?
• A content format. Duolingo turned language learning into short daily lessons with streaks. The same format works for coding, finance, or music.
• A delivery model. “Uber for X” became a joke because it was overused, but on-demand delivery of a service still transforms some industries.
• An audience. A tool built for large companies can often be simplified and repriced for freelancers, or the other way around.

How do you find a model worth adapting?

Start with what’s already making money, not with what’s trending. A few practical places to look:

• Mature markets ahead of yours. If you’re outside the US, look at what US startups were doing three to five years ago and ask whether your market has it yet.
• Products you pay for. If you already pay for a tool, someone like you would pay for a version built for a different niche.
• Review sites and forums. One-star reviews of popular products are lists of unmet needs.
• Job listings. When a company hires many salespeople for one product, that product is selling.

Then ask one question about each candidate: what is the original bad at, for a specific group of people? That gap is your business.

How do you win when someone else did it first?

You win on execution and fit. The original company has to serve everyone. You only have to serve your corner of the market, and serve it better.

That means local payment methods, a language and tone your customers recognize, pricing that fits their budgets, and support that answers quickly. These sound small, but they’re often the whole reason a customer picks you.

Speed matters too. AI tools have made it dramatically cheaper and faster to build a first version. A solo founder can now ship in weeks what used to take a team months. That makes the adapt-and-execute approach more accessible than it has ever been.

The bottom line

You don’t need to invent the future to build a good business. You need to spot what already works, understand who it doesn’t serve, and ship a better version for them faster than anyone else.

Look around at the products you use and the businesses you admire. Which one could you rebuild, reposition, or repackage for people you understand better than anyone?