I experienced failure first-hand multiple times throughout my career. Of course, I tried to learn from each failure, but it was like in a video game: As I progressed, the lessons became increasingly difficult and harder to put into practice. Ultimately, I applied them by helping to build more than 100 new business ventures, ensuring other leaders and their companies did not fall into the same traps. Allow me to share some of those lessons.

I started my first company while in high school in Switzerland in the mid-1990s, creating websites for small businesses. Afterwards, I joined the leading web agency in the nation at the time, and I could do some amazing things there, like creating the first online real estate platform for a bank—which I regard as my first big project. It failed miserably. No one wanted a real estate platform. I learned a valuable lesson there: You must build products that solve actual customer problems, as opposed to building what you want to offer as a business.

I joined a strategy consultancy, Roland Berger, in Zurich after graduating from the University of St. Gallen, and this was another big learning curve. What puzzled me as a young consultant, though, was how much of what we conceptualized never got executed properly. While cost-cutting mandates were straightforward and delivered results, all we really did was recommend how many people should get fired. But on the other side, and somewhat strangely to me at the time, most growth strategies failed to achieve impact.

The myth of business-model transformation

So, I decided to get into the trenches and become a manager at a large company. I wanted to learn how to get things done. I happened to join Kuoni, the largest and most prestigious travel company in Switzerland at the time—a client of my previous consulting firm, whom we had advised on a transformation program.

It was clear to me and the whole management that travel would change forever, with the emergence of online travel agencies and increasing shares of direct bookings at hotels and airlines eating away our market share and margins. Consequently, we were tasked by the board to create new online business for the company. The strategy consisted of two words: “Go online.” But we failed miserably. The company went down and eventually was sold off and ripped apart.

There was a plethora of learning in that experience, so much so that I wrote a case study on it with Insead. But the key point I learned was that business-model transformation is a falsified approach, nothing more than a myth. It would mean transforming a whole value chain. You simply cannot do that, as it would mean changing everything.

After leaving the established travel company in 2013, I joined Everyglobe, a Swiss startup that aimed to become the travel agency of the future. We raised 1.7 million CHF, a lot of money for a seed round back then, which we used to build the technology for a transformative customer experience. Our product was very simple: a search field and four sliders for travel preferences, such as relaxation or social activities. The interface would in real time customize travel recommendations on where to go and what to do, and you could immediately book the trip. The technology felt like magic. We were a truly digitized travel advisory.

On what effectively was a tiny budget, we built technology that was at least 10 times better than the kind used by my old travel company. We also achieved great traction in terms of user growth through viral marketing campaigns, at times having up to 10% of our whole target group engaging on our platform. The problem was that people were reluctant to book with us, even though we had the same prices as other online platforms. The reason was simple: We were an unknown startup. After all, a trip costs a lot of money, and people preferred to do the actual bookings with more established companies. We ran out of cash, and that was it.

What I took away: It must be possible to combine both worlds. To this day, I’m convinced that our travel startup would have taken off if it could have benefited from my old travel company’s customer reach and brand trust. I’m also convinced that my old company would still be around if it had managed to leverage its many valuable assets to create new business.

That’s when I came to believe that it’s in fact possible to combine both worlds, but it had to be done in a smart way to overcome large companies’ limitations, which I was subject to myself. That is why I founded Stryber, a strategic growth consultancy and a corporate venture builder.

Combining these two worlds is far easier said than done. An organization must overcome almost insurmountable internal problems, which I had encountered myself being part of the large corporate structure. You need to have a very strong strategic mandate, owned by the CEO and backed by the shareholders, to overcome the typical analysis paralysis and move to action. Problems need to be solved and things need to get done, without fear of failure. You need to cut through the noise, the latest trends, the promises of salvation, and be extremely careful with how you select your strategies for impact and whom you trust to execute on them.

To solve these problems, we have tried many different routes. We tried corporate venture capital, open innovation startup collaborations, internal ideation contests, and accelerator and intrapreneurship programs. My conclusion is as simple as it is sobering: None of this works for new business creation.

Build it or buy it

At Stryber, we have found only two effective ways to create new business: build it or buy it. It’s as simple as that. Our data suggests that a proper acquisition of a company at scale, or at least in a late growth stage, is the dominant option to create a new line of business, albeit at a hefty price. And yes, we are acutely aware of the pitfalls of M&A. But the risk of an M&A strategy is not the post-merger integration failures, which need to be considered, but that in 9 out of 10 cases there is simply not the right target to be acquired. This often leaves only one strategy available: Companies must build new business models themselves, despite all the internal pitfalls. That’s what I call proper “corporate venture building,” and it’s without alternative.

Over the years, my company and I have built well over 100 new business ventures on behalf of and in cooperation with large organizations. Below are the learnings we’ve uncovered through working with large banks, retailers, insurers, health groups, food companies, private equity firms, and sovereign wealth funds.

The root causes for both success and failure have nothing to do with innovation or the venturing process itself. The first key success factor is having a clear vision for the parent company’s future, not the new business initiative, based on a common understanding of the market dynamics at large. To support that vision, a clear financial ambition must exist. Building up a common view of the future among both senior management and the shareholders is key. This helps to create the equity story, which leadership must buy into and commit to, and shareholders literally must buy, by either injecting cash or forgoing dividends.

Equally important is setting up the proper governance. Many managers wildly underestimate this aspect, because in their normal business the existing corporate governance is working perfectly fine. But operating a new business in the same governance as your core business means stripping it of any chance of success. For new business, you must create many degrees of freedom—and often complete independence for the venture to find its way.

Another lesson: Ideas come a dime a dozen. Even after launching in the market, 89% of even the most promising startups never make it to scale, as our analysis has shown. From the earliest funnel stage of systematic market analysis, idea screening, and deal sourcing to a venture that scales, we see a total loss rate of 97% along the way. That’s why falling in love with your ideas, and overspending in early stages of development, is a terrible idea. Only after a venture has made it to product-market fit should you start to aggressively double down with follow-on investment, until the new business venture reaches scale. This is also when it gets really expensive and is why you need the strong shareholder mandate in the first place, because they will need to fund the new business through its growth stage, until it’s mature itself.

A large organization that seeks to create new business has only one choice: to act systematically and programmatically on both M&A and venture building. Had we known—and applied—what I know now, I am convinced that both the old travel company and my travel startup would still be around.

Jan Sedlacek is founder and managing partner of Stryber, a strategic growth consultancy headquartered in Switzerland.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.