August, 2019 Page 10 of 12 make predictions about when to upgrade, and then we've got a conversation about what is happening, reclaiming, and not giving me a form at the end of my experience. Also, endings are obviously going to change the sequence of events as a consumer goes through those events. There are roughly about seven consumer endings. You can map them over services, products, digital. And you can imagine the timeout, in terms of services, you go two weeks, holiday, off for two weeks, that is ended, isn't it? You have a warranty period that also sell-by dates so there is an ending in that. And in digital a one-year licence agreement. You can imagine somebody turning up at your house to fix the boiler, it is completed. We have experienced a lot of broken withdrawal endings when we have signed up to start-up apps, where they go, "That's a cool feature, I love it, comment, commit, download, engage, make friends in app. They are doing well. They have sold out? Yahoo has bought them?" We also end up having lingering relationships with products, like your phone in a drawer. It might be in your house, you don't use it, it is out of date, lingering around, you haven't concluded that relationship. There is also aspects of proximity in our endings. For example, in services, in the UK open get the service of the BBC, and what a proud service it is. Then I moved to Sweden so I had to go on VPN so I was out of proximity unless I became illegal. But when you move from Apple to Android, you move out of the jurisdiction of that thing. Let me give you an example of how to analyse that in terms of some of your products. Which I think is a really messed up. Freemium comes in off in three flavours. You have time out when after three months of usage, that is the end of your freemium experience, you get credit out, after five events and that is the end of freemium. Or you will get only a certain feature set and outside of that you can sign up to get the full, bountiful feature set. What is funny about how polished the endings are in freemium models, the same companies that do really good freemium do really bad endings, when actually they've got endings down in some of these aspects. What I suggest they should do in freemium is make them more disruptive and make the people who pay for the real service make a much better ending there, and it would be way better. People often ask about how we can get this into projects, because no-one will think about endings on Monday. So you will have to do sneaky things and engage in a long philosophical debate about the meaningfulness of endings and then have to go back to the Protestants to justify it, which I'm sure you don't want to do. No-one has time for that, so 5/5/5, nobody cares about 5% of your project I'm. So you say it is just 5% of your time. Because no-one can think five years out, then you think, 5% of project time, no-one cares about. But you will have to deal with endings, because probably nothing in your product pool will last five years if you are building digital.