Lesson: Customer Understanding with Steve Blank

Step #4 Two-sided Market: A two-sided market simply has users and payers

Remember we've been talking about past failed signals and experiments. So as you get out of the building and start talking to these potential customers, you need to set up some experiments. How do I know these are the right customers or the right buyers? So you want to set up a series of experiments to test interest and figure out where do you want to test it and what kind of experiments can you run and how many do you test? By the way the "how many" is a classic question and the answer is it depends. In a business-to-business enterprise sale, you might get away with talking to less than 20 customers and that's a great number, but if you have a mobile or web app you want to talk to hundreds if not thousands of customers. Obviously you can't personally interview all those like you could in the enterprise and so you want to get out and reach them virtually but the number you want to get to really is an order of magnitude or too different than if you were doing B2B direct sales.

Just to caveat and a warning, when you're out of the building trying to talk to hundreds or thousands of potential web or mobile app users, you want to be careful that you're not completely dependent on online surveys. People simply lie on surveys and so what you want to have is some sample where you actually bring those people into your building or you go see them and you watch their pupils dilate as you actually ask them the question and see if that kind of diverges from the data you're getting when they're hitting your website or you're sending them a questionnaires. Then you'll have a correlation to be able to figure out is that data is valid or not? So just keep that in mind when you're trying to reach hundreds or thousands of people for customer discovery.

One other thing about the business model canvas in general, now that we've talked about customer segments is understanding that sometimes you have multiple customers, multiple segments and it's kind of interesting to take a look at Google as an example. Let's just take a look at Google Search, let's just think about who's the customer for Google Search, it's anyone searching on the web or mobile. And the product we have is a free search bar, a free mobile app. What are the key resources? They need software and software developers and installed user base but the key idea is, this product is free. If you think about the cost structure for this, it requires massive data centers and software developments and key activities or data center scale and speed but the key idea is, the revenue model is free. But Google and Google searches one of most profitable businesses on the face of the earth. How can Google do that?

Think about it, what we've just been doing is Google has been acquiring users but someone else is paying for this product and that someone else is a different customer segment called Advertisers. So we had users here but we now have payers here. What's really interesting is, for every customer segment, we have a value proposition. We had a free search bar for the users, but for the advertisers we have a completely different product. We have Google AdWords and our customer relationships are different. For users it was kind of an automated process; you just basically went to the Google website or you downloaded the Google Toolbar but for payers it might be automated but they also have telemarketing and direct sales to go after big advertisers. But their distribution channel in addition to being automated is also self service and direct sales. The key thing to note is that not only do they have different segments and different value props and different customer relationships, more importantly they have a different revenue model. It's a big idea.

Multiple customer segments require multiple value props, multiple customer relationships, sometimes different channels but always different revenue models. These are the payers. So Google Search is what we call a two-sided market, there are users and there are payers. So just to summarize, for Google Search, each had its own value propositions but most importantly each had its own revenue stream and one segment cannot exist without the other, that is if you had users but you had no revenue, you'd be out of business over time. But if you had a revenue model but no users, there would be no reason for those advertisers to show up. This was an example of a two-sided market. Let's see what happens when we have multiple sides.