Will a cofounder steal my idea? How often it happens, and how to share yours in stages

You have an idea you believe in, and you know you can't build it alone. To find a cofounder you'll have to explain it to people you barely know. Somewhere in the back of your mind a question keeps coming back: what if one of them likes it so much they go and build it without you?

It's one of the most common worries among first-time founders, and it stops some of them from looking at all. This post sets out what is actually known about idea theft between cofounders, why the idea matters less than it feels, and a practical way to share it in stages so that the detail you hand over grows with the trust you've built.

The short answer

• It happens, but rarely. There's no survey that measures how often a potential cofounder takes an idea. The cases that get written about are a handful of anecdotes, and most involve people who already had the full plan, not strangers at a first coffee.
• The idea is the cheap part. Building a company takes years of work, the right skills and constant changes of direction. Someone who hears your idea once still has to do all of that.
• Staying quiet costs more. Nobody commits to a mystery. If you won't say what you're working on, the people you most want will move on.
• Share in stages. Start with the problem. Add your insight once they've shown real interest, and the specifics once they've put in time on a trial project.
• NDAs have a narrow use. They make sense for genuinely confidential material later on, not for a first conversation about a problem.

How often do cofounders actually steal ideas?

Nobody knows the rate, and anyone who quotes one is guessing. We couldn't find a study that counts how often an idea shared with a potential cofounder ends up in someone else's company. What exists are first-hand stories, and they're worth reading closely because of who the "thief" turns out to be.

Steve Blank, a Silicon Valley entrepreneur and author, wrote in 2009 that in 21 years his ideas had been "stolen" twice (Steve Blank). In the serious case, a marketing executive he had pitched during customer interviews asked for a printed copy of the slides. About nine months later the same person was running a competing company with Blank's material in its deck (Steve Blank).

Brett Fox, a former startup CEO who writes about founder problems, describes two cofounders who had worked with him before, planned the company with him for about three months, and then quit the night before the company was due to be funded. They started a competing company with his plan and slides. His company found new cofounders and went on to raise money; theirs, he writes, "never went anywhere" (Brett Fox).

The best-known story is the ConnectU founders' claim that Mark Zuckerberg took their idea for Facebook. Facebook denied it, the two sides agreed a settlement, and in 2011 a US appeals court ruled that the Winklevoss twins had to accept it (Al Jazeera).

Look at the pattern. In each case the person who took the idea had the whole thing: the slides, the plan, months of conversations. None of them heard a two-minute description of a problem and ran off with it. Fox draws the same conclusion, writing that the biggest risk of someone stealing your ideas comes from inside your own company (Brett Fox).

There's one piece of research worth knowing. Lillien Ellis of the University of Virginia's Darden School and Brian Lucas of Cornell ran three studies with 965 adults, published in Social and Personality Psychology Compass. They found that people who create ideas expect thieves to go after polished, finished ideas, while the people asked to play the thief said they'd take ideas earlier than creators expected, partly because taking an early idea feels more like "borrowing or taking inspiration" (Cornell Chronicle). The practical reading: your earliest, most specific thinking is the part to share carefully, with people you have reason to trust. Hiding the whole idea would cost you more, as the next sections show.

Why the idea is the cheap part

Paul Graham made the point bluntly in 2005: "startup ideas are not million dollar ideas", and if you doubt it, "just try to sell one" (Paul Graham). In the same essay he argues that the main value of a first idea is that, while finding out what's wrong with it, you come up with the real one.

Derek Sivers put numbers on it the same year. In his scheme a brilliant idea is worth 20 and an awful one minus 1, while execution runs from $1 for none to $10 million for brilliant. The value of a business is the two multiplied, so a brilliant idea with no execution is worth $20 (Derek Sivers).

Blank's own story shows why. By the time the copied deck surfaced, his company had talked to customers, changed its plan and found a different, larger market. The competitor launched on the original idea in the market Blank had already left. In his words, they were "no longer the company in those stolen slides" (Steve Blank).

Marc Randolph, a cofounder of Netflix, calls the fear of theft "a fallacy, and a dangerous one at that", and points out that Netflix began as rental by mail, with due dates and late fees, a long way from what it became (Marc Randolph). The idea you have today is unlikely to be the one you end up building.

The computer pioneer Howard Aiken is credited with the classic line: "Don't worry about people stealing your idea. If it's original you will have to ram it down their throats." Quote Investigator traced it to a 1985 article about Aiken (Quote Investigator).

What you risk by not sharing

Secrecy has costs too, and they're easier to miss:

• Good people walk away. Someone deciding whether to give up evenings or a salary needs to know what they're signing up for. A vague pitch attracts people with nothing better to do.
• You can't test fit. You learn how someone thinks by watching them react to the problem: what they question, what excites them, where they push back.
• You lose feedback. Buffer's founders made sharing early a habit, on the view that telling people sooner tells you sooner whether anyone wants the thing (Buffer).
• You lose time. Every month spent guarding the idea is a month nobody is building it, including you.

How to share a startup idea with a potential cofounder, in stages

You don't have to choose between telling everyone everything and telling nobody anything. Share in layers, and move to the next one when the other person has done something that earns it.

Stage 1: the problem

Describe the problem and the people who have it, in a couple of sentences. "Small dental clinics lose a lot of revenue to missed appointments, and the reminder tools they use don't connect to their booking systems" tells a good candidate plenty. Problems aren't secret: anyone who works in that field already knows about them. Our guide on how to ask someone to be your cofounder has example first messages written this way.

Stage 2: your insight

Once they've shown interest and told you something about their own work, share what you know that others don't: what customers told you, why you think the timing is right, the rough shape of your approach. Keep this conversational. Talk it through on a call rather than sending a document.

Stage 3: the specifics, during a trial

Customer names, numbers, a technical method, a working prototype: share these when you're working on something together and they need them to do the work. A short cofounder trial project is the best test of trust there is, because you see how someone behaves with real information and real deadlines.

Stage 4: everything

When you've both decided to start the company, there's nothing left to hold back. Write down what you've agreed about roles and how decisions get made at this point.

What to keep back until later

Most of what's in your head can be shared early. A short list of things is worth holding back until stage 3 or later:

• Source code and access to anything you've built.
• Your list of customer contacts and where each conversation stands.
• Data you've collected, especially anything about customers.
• Deals or partnerships that aren't public yet.
• A technical method or process that is genuinely hard to work out.
• Logins to accounts, domains and tools.

Blank's rule is useful here: never put the real secrets on a slide, and assume anything you hand over in writing will be kept (Steve Blank).

Signs someone is more interested in your idea than in you

Most people you meet are looking for a cofounder for the same reasons you are. A few signals are worth noticing:

• They want detail before they've shared anything. They ask for your plan, numbers and customer list in the first conversation but say little about their own work.
• They care about the plan, not the problem. They're keen on how you'll make money but show no curiosity about the people you're building for.
• They push for documents. "Just send me the deck" after one call, when a conversation would do.
• They won't do a trial. They want to see everything but won't commit a few weeks to working on something together.
• They go quiet after a deep dive. Silence right after you've shared the details is worth a direct question.
• Their story has gaps. Past projects that ended with them keeping the work, or a job at a company in the same space they haven't mentioned.

Our cofounder red flags guide covers wider warning signs, including someone who plans to build on their employer's time.

Should you ask a potential cofounder to sign an NDA?

An NDA (non-disclosure agreement) is a contract in which one side, or both, agrees to keep information confidential. A one-way NDA protects one party's information; a mutual NDA protects both. The UK Intellectual Property Office publishes example one-way and mutual NDAs for people who want to discuss a new product or idea with a potential collaborator (GOV.UK).

When it's realistic. Later on, when you're about to hand over something genuinely confidential, such as code, customer data or a technical method, during a trial project. At that point a mutual NDA is a normal request between two people who are already working together.

When it isn't. Before a first conversation about a problem. Many experienced people will decline, and asking can signal distrust before you've met. Investors are a useful comparison: Fred Wilson of Union Square Ventures wrote in 2006 that he had never signed one in 20 years (AVC), and Wil Schroter of Startups.com puts the share of investors who won't sign at 99%, his own estimate (Startups.com).

What it can't do. An NDA covers information. It doesn't stop someone from having the same idea, and enforcing one costs time and money. This post isn't legal advice: if you have something that needs protecting, talk to a lawyer where you live.

Habits that protect you without paperwork

• Keep dated notes. Write down what you shared, with whom and when. Blank suggests putting your initials and the date on anything you draw.
• Talk rather than send. Walk people through your thinking on a call. Send documents once you're working together.
• Write up each conversation. A short message afterwards ("good to talk; next step is X") keeps a record and shows you're serious.
• Keep building. Every week of progress makes the original idea a smaller part of what you have.

If you do think someone has taken your idea, compare it against your notes before you react, and talk to them directly. Fox's advice is that suing early is rarely worth it and that competing in the market usually is (Brett Fox). For anything legal, speak to a lawyer.

Finding a cofounder without broadcasting your idea

The safest way to share an idea is with someone who is the right fit for you, and that is what Foundinity is built for. You state what you need in a cofounder, and answer an assessment about how you work: commitment, working style, ambition, personality, goals, values, alongside your skills and interests. A multi-trait algorithm built on established psychological compatibility models ranks the people you see. Among the things you state is whether you want to see people who already have a specific idea, people who aren't set on one, or both.

You don't have to post your idea anywhere to be found. Your profile is visible only to compatible members, and every match shows where the two of you align and where you differ. A 10-day clock and a limited number of messages before a call move the conversation towards a live call, which is where trust is easiest to judge, and you book that call in the app. It's free to get started; see how it works.