How do companies actually develop and test hypotheses? We will focus on different innovation-related examples to illustrate how hypotheses can be developed in very different contexts. The first example relates to a completely new business idea, Uber. Travis Kalanick came up with the idea behind Uber in 2008. He was in Paris, and he wasn't able to find a taxi anywhere. That is when he had the lightbulb moment and he realized: If people use taxis for everything, why wouldn't they use an app for taxis? He went back to San Francisco and kept thinking about creating an app to connect passengers with private drivers, since taxis operate in a highly-regulated industry. He quickly understood that for his idea to work he had to have both passengers, who were facing the same problem he had, so they had to be unable to easily find cabs or rides to their destination, and drivers, who had time to fill in between pre-booked rides. Let's focus on the drivers’ side. Travis Kalanick started with the following two guesses or hypotheses. Drivers need to fill in time between pre-booked rides. Drivers will be willing to fill this time with other rides. In order to test his very simple hypotheses, he googled ‘San Francisco chauffeur’ and cold-called the first 10 numbers that appeared in the search. While three drivers hung up right away, and four listened for a while before ending the conversation with Travis, three were willing to meet with him and hear more about the opportunity. For Travis, this was enough to confirm he was up to something. While this is not necessarily the most rigorous example of data testing, it showed that Travis Kalanick had a very clear idea of what he needed for his idea to work. And this guided his search for clues about the feasibility of his business idea. The second example is about how to decide to invest in a new business idea. There is growing evidence that shows how hard it is to decide whether to invest in new business ideas. Research suggests that even experienced entrepreneurs do not make better investment decisions than other investors. And this is likely because when ideas are early staged, their quality is hard to assess. This is a case where the use of a scientific approach can be most valuable. And in this example, we show what Jason Stoffer, Managing Partner at Maveron, has done. Maveron is a 20-year-old venture capital fund investing in startups selling consumer only products or services. Maveron backs firms operating in shopping, food, learning, retail, lifestyle, and related sectors. Jason Stoffer was recently interviewed by Harry Stebbings, the mind behind The Twenty Minute VC podcast, when he talked about one of his recent investments: Dolls Kill. Dolls Kill, based in San Francisco, is an online apparel boutique the targets edgy fashionistas: from punk rock, to boho, to goth, and glam. When Jason Stoffer came across this investment opportunity, he was wondering if a similar brand could actually work. This seemed like a very niche market, but as he started to analyze the market, he understood that there was a path to significant revenue. Alongside the founders of Dolls Kill, he started to collect existing evidence to formulate some hypotheses. His hypotheses then were: There is a niche market for online “edgy fast fashion”. This market is sizeable. This market can be better served with a dedicated online boutique. Dolls Kill, in particular, was targeting women aged 15 to 30. When you look at the size, this may appear like a small offering, but in reality, it is a giant market. Jason and Dolls Kill’s founders noticed that almost 30 percent of millennials attended music festivals in the last 12 months, compared to half of that in 2014. And this provided some evidence of the growing interest for "edgy fast fashion". Their reasoning was that a company like Zara, that specialized in fast fashion for a broad type of consumer, was worth 100 billion. Among the products sold by Zara, some were classified as edgy fashion, that feed the target Dolls Kill was going to address. If Zara is worth a 100 billion, can they create a brand focused on 15-20 percent of women who are more fashion-forward? Jason Stoffer and Dolls Kill engaged in a testing process that proved his hypotheses were right. According to Crunchbase, Dolls Kill has received more than $100 million in funding, and employs more than 100 people.