Here's our first real problem of dividing costs. We have a situation where I was in New York, and I had a client in Houston, and a client in San Francisco. And I needed to travel to both destinations. And I could have done two roundtrips. But we're able to work things out, so I was able to do a triangle route. That is, I flew from New York to Houston, Houston to San Francisco and San Francisco back home. And if you add up the cost of each of those legs, the total trip was $2818. And Houston and San Francisco agreed that they were collectively responsible for the 2818. The question is how much should Houston pay and how much should San Francisco pay? So some ideas folks, what do you think? Percentage. So what you would do is you divide it up based on the $1243 to the $666 ratio. In the sense that the round trip between New York and Houston is $2,486, this is $1,332. And so, that's the ration which is about 65, 35, basically. I understand that, yeah, the proportional division. Other choices? >> 50-50. >> 50-50, so 50-50 is $1,409, $1,409. And I think the response to that is, Houston, we have a problem because- >> [LAUGH] >> $1409 is bigger than $1332. So, Houston would say, why should I pay half of something when it's more than what I would have paid all on my own? And then, in terms of proportional, so, the proportional division, oh, we have the, we could split the 909, we could just split this last leg half and half, so we have the split everything, so Houston is responsible for this leg. There is no question about that. San Francisco is responsible for that leg, there's no question here. So we have this intermediate leg. We'll split that evenly, or you could split it proportionately. If you split that proportionately, by the way everything is split in proportion, because the first two legs are already in that proportion. Okay. So, I will note that I have really failed as a professor. And the reason which I fail this because we've had four or five suggestions now and no one has applied this notion of the pie. That's not the pie. So the question is- does somebody want to give a shot at what the pie is? I'm not sure that I can get it exactly but it seems to me don't you have to look at the difference between taking the two round trips [INAUDIBLE] And then which client do you like better? >> No, no, no. Don't stop there. >> [LAUGH] >> It's not a question of which client you like better and no, we're not gonna bill both of them for the total and pocket the difference. So if you think about this, the best that you could ever do is $1332. So that's the most you would ever pay, in which case, if you were San Francisco, you say, "Okay, look you pay $1332. I'll pay the rest." If San Francisco says that to Houston, Houston should respond, "now, wait a second. Let's go the other way. Let's have you, San Fransisco, pay $2486, and I'll pay, The difference. And if we go in between those two things what we could really say is the following. If you add up the two round trips, they add up to $3818. The triangle route was $2818. So the difference is $1000. That's the savings by doing the triangle route compared to the the two round trips. Who is more responsible for that $1,000 savings, Houston or San Francisco? >> Equal. >> Equally, right, because if Houston says Monday doesn't work for me I gotta move it to Friday, what happens to the savings? >> Goes away, if San Francisco says Tuesday doesn't work I gotta move it to Friday, >> [INAUDIBLE] Well it does have something to do with it's not just at my convenience which is we save them a thousand dollars. [INAUDIBLE] But it depends on how we divide it up. And so I'm claiming is a thousand dollar savings by doing the triangle route compared to the two round trips. Who should get the thousand? My view is they should each get five hundred. And so therefore Houston pays $832 and San Francisco pays $1986. And I'll note that isn't what any of you proposed when we started out. And so the natural thing is to divide it in two, divide it proportionally, split one of the legs. We're trying to grapple and come up with some idea of something that's fair. >> And what proportional division does is treat each dollar of airfare the same way. But I don't know what we think fairness should apply to dollars of airfare, it should apply to people, because we need the two people to come to an agreement. And so once you see things in terms of the pie, then we can understand how to treat people fairly. And, we'll see the numbers are actually different, in the sense that if we split the Houston leg, Houston would pay $1,120. If we split proportionately, Houston would pay $983. Whereas if we split the $1,000 savings, Houston pays $832. Now, of course, if you're negotiating with somebody who hasn't been in this session, then, I would propose splitting the Houston leg evenly, if I were San Francisco. I'm not against you trying to get more than half the pie, but, when you're negotiating with somebody who is equally well understanding of the situation, then I think that's where you'll likely end up. Alright, now that we've done this a second time, let's keep trying other examples to see if we can really lock it in. We have a situation here of two airlines that have the ability to share a runway. And so, airline A needs a runway of length 1 and airline B needs a runway of length 2. The cost of each length of the runway is 1. So how should the two airlines split the cost? How much should airline A pay and how much should airline B pay. I'm sorry? >> Depends on how many planes are going to land there. >> They each have one plane, if you like. They each have one plane. Oh, they have to split it. >> So one and one is what I heard so I think Herb would have Airline A pay for the first length and Airline B would pay for the second? >> 75, 25 is another solution. >> A dollar 50, 50. >> I'm sorry? >> A dollar 50, 50. >> A dollar 50, 50, which is 75, 25. >> Yep. So actually so far, nobody has done the two and one strategy. The sort of proportional to the lengths. And I think that's great, because in fact, to me the argument is, airline B, is clearly responsible for the second leg. They're the only one who's using it. They can share the first leg. So that means Airline A could pay a half for that. Airline B pays a half for that and Airline B pays the full amount for the second leg. So it's one and a half or half 75, 25. If we thought about this in terms of the pie. What is the combined cost if they work together? Two, if they don't reach an agreement and have to go on their own. Three. One plus two, which is three. So the pie here is one. It's three versus two. Therefore, we split the pie. Each person gets a savings of a half. So that means A pays one minus a half, which is a half. B pays 2 minus a half, which is 1 and a half. Great. If somebody says to you, if they say to you as airline A, look, you need to build A the runway of length 1 anyway, so you pay that, and I'll just pay for the second part. What's your response? You're using the first one. >> My response is, you need to build a runway of length too. Why don't you pay for the whole think and I'll just use the first half? >> [LAUGH] >> And so once you see things in terms of the pie, any argument that is made can be flipped and turned the other way. Let's now try this for an Uber version. So we have two passengers who are going to an airport. A gets in first, and then B joins. And the first leg of the trip is $6, for convenience, the second leg is also $6. And the question is, how should A and B split up the $12 fare? That's a good question. Yes, I'd be 11. What do we think in terms of how to split up this? 9 and 3? So what this gentleman wants to say is, look A will pay all of 6 and A and B will split the second leg. >> Same problem. >> Same problem. Not necessarily. So, in particular, if it looked like this, you probably wouldn't split it six and three. And so somehow the angle should matter. And this brings us to your point here, which is we have to think about what's the PIE? And in order to understand the PIE we need to understand how much it would've cost to go for A directly to the airport. So in this circumstance what is the PIE? It would cost A 11 to go directly, B 6, so collectively what would it cost them to go together? 17, if they go together, it is 12, so therefore, the savings is 5. You split the savings. So what that means is that Alice should pay 8.5. And bob should pay 6- 2.50, or 3.50. Okay, so not that hard, once you decide to take things in the perspective of what is the pie? >> Except Bob should pay Alice for the extra time involved in picking him up. >> Yeah, so you're absolutely that I've ignore the time cost, and that's a great complication that we should include in this. Absolutely. I agree 100%. So let's think about this in terns of diversions for a moment. It is the case, you might argue, that Bob is a little out of the way. Because if Alice went straight, it'd be 11, but Bob is the detour. And the question is, who should pay for that detour? Should Bob pay for that detour, or Alice? Who is out of the way? Okay, how many of you are saying Bob is out of the way? The reason you're saying that is if only Bob lived right there, everything would be great. There would be no reason for detour. But my response to that, if I were Bob is, Alice, if only you lived over there. >> [LAUGH] >> Right, there'd be no reason for a detour. And so, there isn't any sense in which Alice is out of the way anymore that Bob's out of the way. The two of them are equally out of the way. If they're gonna share a car, then they need to create a detour of one. Who is more responsible for that detour? Neither. They're collectively responsible, so they should split the cost of the detour. And so, therefore, I can take your runway solution, if you'd like. The 3 and 6. I'm sorry, the 6 and 9. I'll say it right this time. 3 and 9 but, deal with the detour. Because when you had Alice paying 9, Alice at that point, or A, is paying the full price of the detour. If Alice gets half of that back, then Alice is paying 8.5, which is the same story as we get with the PIE. Okay, so it's connecting, if you want, the runway situation to the case where it's not perfectly straight. And we can do this in much more complicated situations. If you have Alice get in first, then Bob, then Alice gets out, then Bob gets out. It doesn't really matter. Here, the route costs 18. If they didn't work together, it would be 11 and nine or 20. So, therefore, they're saving a combination of two. So the pie is 2, so you split it each saves 1. In that case A would pay 11- 1 = 10 and Bob would pay 9- 1 = 8. So all we have to do is say, how much would it cost if they worked together, compared to what it would cost if they don't reach an agreement. By the way, as you may know, Uber is doing Uber Pool, where you need to share the rides. Right now they're just making it a flat fee of $5, in Los Angeles anyway. But I think you may see something like this happen in the future. And one of the reasons is that they can tell you what it would have cost to go on your own. They can predict, based on traffic and their GPS software, how much the individual route would take, so they say, here's the savings that you're gonna get. Let me now apply the same approach to something perhaps more practical or a story from one of my colleagues. After a concert in Lincoln Center, there's a limo waiting, a town car, to take us back from New York City to New Haven. And seeing as I'm about to get into the car I see a colleague and I realize this person is also going back to New Haven. And so there's room in the car. I like him. He likes me. It's all fine. The car service costs $150. It's worth $200 to me because it's basically it was late at night, the trains aren't all that frequent, so I was happy to have the car service. He was a little cheaper, he was going to take the train. And so the question is, should this colleague join me, and if so, how much should he pay? Drink is one answer. What other? >> [LAUGH] >> The train fare. >> Nothing. >> Nothing. The difference between the train fair and the limo. 75. >> [LAUGH] >> Is it man or woman? 75, so I think $75 can't be the right answer, because we don't know that this person values the car service at $75. And what we propose should be something that always makes sense. The rule that we're using should always lead us to have an agreement. >> What does the train cost? >> Let's say the train plus the taxi to the train station and the taxi home is $40. Now what is your answer sir? $40, so I would say if we thought if that was all that mattered then the PIE is $40. In which case what should the person pay? 20, and if the person pays 20, how much is he ahead? 20. And how much am I ahead? 20. Okay? But actually, the car is nicer than the train. And so let's say the car is $60 nicer than the train. And he volunteers that there's no strategy here, so now, the savings is 40 plus a 60 comfort, so what now is the PIE? >> $100 >> $100. In which case how much should he pay me? >> Pay you $50. >> $50 How much is he ahead? >> $50. >> He's paid $50 for something worth $100. How much am I ahead? >> $50. >> $50. Great. Let's say he gets a little car sick. Not quite- >> [LAUGH] >> Not to the point where he's gonna go further. And he thinks the train is $10 better than the car. >> [LAUGH] >> Now how much should he pay me? >> He should just go on the train. >> He should just go on the train? >> [LAUGH] >> Well, not quite. Because basically, what's the pie? >> $30. >> 30, in which case he should pay me? >> 15. >> 15, if he pays me 15, how much am I ahead? >> 15. >> 15, how much is he ahead? >> 15. Well, he saved 40, lost ten from being slightly nauseous, lost 15 that he paid me, so I net, he's also up 15. So I think one reason why we're hesitant to think about these payments is because we don't have this great framework for what's the fair outcome. Now, some of you said that, forget about it, what a jerk. I mean just let the person come in for free. >> [LAUGH] >> And if you think it' s incumbent on me to say come in for free, how should my colleague respond? >> [CROSSTALK] >> Thank you. >> [LAUGH] >> But what else should the person say? Can I give you, let's do a case where the person values at 100. I don't think the answer is he should volunteer to say, 50. I think he should volunteer and say, let me give you 100. Because if you think it's incumbent on me to say, here, you can have the whole pie, then what should he say in response? No, you can have the whole pie. So I offer free. He says no, let me pay a hundred. And where do we settle? We settle at 50. Right? >> You know that it costs 150? >> That's true. >> He doesn't know that it's worth 200 to you. >> The fact that it's worth 200 to me is not relevant in this particular case. Because that's not part of his pie. I'm gonna get the 50 surplus. The difference between the cost at $150 and the value, whether or not he comes on board with me or he doesn't. So that 50 is not available to him. So the only thing we're really arguing about is how to split up the 100 that he brings by joining me in the car. Yes and that's where the 60 came in. The value of time, is why it went up from $40 up to $100. >> But you still have to figure out- >> [CROSSTALK] Yes, in fact, so he valued that at $60. >> Yes. >> Yes, more and more comfortable, and my company is worth something. >> [LAUGH] >> Now, what actually happened in the situation is that my colleague offered the ride for free. The passenger sent her a $100 bottle of wine, but she doesn't like wine- >> [LAUGH] >> and the whole thing would have been much better and simpler if they just agreed that he gave her $50. >> And which tastes the same as the $10 bottle of wine. >> It tastes the same as the $10.00 bottle of wine. >> [LAUGH] >> Thank you. So here is a problem that I've adapted from the Talmud, and in the Talmud they give you some examples of how to divide up a pie. They don't necessarily give you the rule, they just give you the numbers. And here's a situation that they present. We have a project, which, if is done, will benefit Partner A 100, and Partner B 200. In those days it was Zuz, but it doesn't matter what the units are. If you like, you can think about this project as a software program that will save you a dollar per employee in terms of payroll costs. And division A has 100 employees. Division B has 200 employees. So division A will get half the benefit as division B. And in this circumstance where this project costs $50, it should be split up 25, 25, in terms of the cost. So equal. In the case where the project cost 150, then the division should be 50 and 100, which looks like proportional. And in the circumstance where the cost is 250, then the claim is it should be split up 75 and 175, which looks like God knows what. And the question is, is there a unifying principle for all three of these rows, all three of these examples? And of course the answer is yes, and what do you think it is? It's the pie. And so let's try and calculate the pie in the first row. So what do we think the pie is here? >> There are 300 aggregates. >> 300 aggregrates. 300 is the aggregate benefit to the two parties if they do this. 250 would be the number once you take into account the cost, okay? But that's still not the pie because 250 is the value they get if they work together. But what else do we have to consider? >> What happens if they don't reach an agreement? And if they don't reach an agreement, what are they gonna get? >> Zero. >> Not quite zero. What will A do? >> A will do the project. >> A will do the project on its own and get 50 and B will the project on its own and get- >> 150. >> 150, so therefore, collectively, they will get 200 if they don't reach an agreement. 50 and 150. If they do reach an agreement, they will collectively get 250. The 300 minus the 50. So therefore, working together gets them an extra 50. What should they do with that 50? >> Split it up. Split it, right. Who's needed more for that 50, A or B? >> Equally. >> Equally. Okay. So therefore we split it 50/50. That means A gets as a net benefit the 50 from itself from doing it on its own, plus the 25 from splitting the pie, or 75. In order to get that $75 how much should they pay? 25 cuz if A pays 25 and has a value of 100, A walks away with a net of 75. How much should B get? Well B's getting 175 in total benefits, the 150 on its own plus the 25, that means B is gonna pay $25. So therefore, it's perfect, 25-25 is that division of the pie. Okay? Can I have somebody try the middle row for me? What is the pie in the middle row? >> Should be net cost in proportion- >> So it would be, if they do it separately, A has a net cost. >> Okay. Well let's do first, if they do it together, what do they get? >> 150. >> They get 300 minus 150, which is 150. If they don't reach an agreement, what will they get? What will A get? Minus 50? >> Yeah, negative. >> No. >> A won't do it. >> A won't do it. There's no gun to A's head saying you have to do this, so therefore if they don't reach an agreement A will get zero and B will get 50, therefore collectively they will get 50. Therefore, the gain from working together is 100 and 50 minus 50, which is 100. They split that evenly, which is 50/50. Therefore, A gets 0 plus 50. Right, 0 on its own, plus 50, which is 50. Which means A pays 50. B gets the 50 from working on it's own, plus 50 from working together, so that's 100, which means B pays 100. Where have you seen this problem before? Not quite the Uber ride. The town car back from Lincoln Center. The value was 200 to me, 100 to my friend. The cost of the care service is 150. If we don't reach an agreement, my colleague won't do it. They'll get zero. If we don't reach an agreement, I'll still take the car, and make 50. Working together, brings us an extra 100. So we're all good. And by the way, if I go back to that first line, you can think about it this way. What is really going on here, if we work together, we don't have to buy two copies of the software. Each copy of the software is $50, if we don't reach an agreement, we buy two packages. If we coordinate, we buy one. So therefore the savings is the avoiding the duplication. So avoiding $50, who's responsible more for avoiding the 50? Split it, so that's why it's 25/25. Okay, how about the last row? What's the pie in the last row? Yes, please. >> It's so expensive, that neither one of them would buy- >> Yeah, so therefore if we work together, what's the pie? >> The pie is 300 minus 250. >> 300 minus 250, which is 50. If we don't work together, what will A get? >> Zero. >> And what will B get? >> Zero. >> So 0 plus 0 is 0, so, therefore, the pie in this case is 50 minus 0, which is 50. We split it, each of us gets 25, which means A pays 75, and B pays 175. So by seeing things in the context of the pie, we're able to take what looks like a mess, and three different answers, and quickly come up with one sensible approach. >> But the hardest thing is to figure out what the pie actually is. >> I think there's two aspects to negotiation. One is figuring out what the pie is. The other is figuring out how to divide it. And we have time to do one activity here. Part of the challenge is people come into a negotiation, and they bring in different concepts of fairness. Which then gets them all confused cuz is the pie miles? Is it legs? And so once you have clarity about what it is that we're gonna divide, then it actually focuses everything in terms of your negotiation. So let me give you a couple more real world examples and see if that helps in this regard. I've got a merger here between two companies. One company is the Planet, the other is the Gazette. These are two newspapers. And prior to their merger, we have that the Gazette is more valuable than the Planet. The Gazette is twice as big as the Planet. The Gazette has 200,000 readers, the Planet has 100,000 readers. And the Gazette runs a tighter ship. And so their market capitalization, the value of all their stock is 22 million, the Planet's is 10 million. But if they come together, they can do some synergies in terms of joint purchasing. And we'll argue those synergies are $6 million. And so the plan is to divide that up in proportion to their readership, two to one. Now it's also the case that the Gazette has some know-how, and if they apply that know-how, they can help the Planet save $1 million a year. But since it's the Gazette's know-how, the Gazette says, I want that whole million. And then there's some overhead that they can reduce, and that's 1.2 million. They think about splitting that up in proportion to their market caps. And lastly, there's some new readers. The Planet will get 10,000 readers from the Gazette, the Gazette will get 5,000 readers from the Planet, because the Planet's half the size of the Gazette. And so what they say is look, the Gazette, I want the value of the readers I send to you, and you can have the value of the readers you bring to me. And when you add all that up, that's 2.925 for the Planet, 6.925 for the Gazette. What do you think about that as a division? >> You have to know the post-merger market cap. >> Well, actually I do. The post-merger market cap is 32 million plus 9.85. And I'm thinking about how we're gonna split up that 9.85. >> That's the pie. >> That's the pie? What do you think about these arguments I've just given you? >> They're [INAUDIBLE] >> They're the net?. >> No, they are [INAUDIBLE] >> Well is it really the case that we should split up their synergies from joint purchasing, two and four? What would you say as the Planet to that? >> I would say 50/50. >> 50/50, basically if this merger doesn't happen, those 6 million disappear. And so therefore you need me just as much as I need you to have it, so I'd argue it should be three and three. And how about the know-how, the million dollar know-how? It's the Gazettes know-how, shouldn't they get it? What is the argument you'd make as the Planet? >> It's more [CROSSTALK]. >> You're getting zero [CROSSTALK]. >> One view is you're getting zero if you don't work with me, if we don't do this deal. But I can go even further in terms of symmetry. What are you bringing to the table as the Planet? >> [INAUDIBLE] the know-how. >> Yeah, your inefficiencies, essentially. >> [LAUGH] >> The Gazette's know-how isn't worth anything without the Planet's inefficiencies. >> [LAUGH] >> And so, only by putting those two things together do you create the million. And so therefore, I think you should split it half and half. Or I could say, slightly less facetiously, it's really our operations, our readers. The fact that our paper exists is why your know-how is worth something. If I thought about the overhead, I'm gonna split that 50/50. I have no particular reason to in terms of market caps. If we do readers, what's the argument to the, it looks like this is fair. Each side is getting the value of readers they bring to the other. So if you are the Planet and the Gazette says, well, I want the value of the 10,000 readers I sent to you, you can have the value of the 5,000 readers you sent to me. How do you respond? >> If the deal didn't happen it wouldn't [CROSSTALK] >> Yeah, if the deal didn't happen [CROSSTALK] >> The deal didn't happen, none of this would go through. But even more than that, what about the 10,000 readers that you sent to me? >> [INAUDIBLE] >> I'm sorry, you still have them? >> You would still have them [CROSSTALK] >> You still have them, but they're now buying a second paper, so what's the argument here? It's not just that, yes, you sent me the readers, but I gave them something to read. >> [INAUDIBLE] >> Okay, so it's not enough just to have your readers, it has to be my content. So only by putting those two things together do we create that value. And so I want to split the value of the readers that you brought to me within my content. And I'll also share the readers I sent to you, because they were my readers and your content. So we split everything. And so my point here is that you're thinking about this negotiation. The standard approach would be to argue over everything line by line, and come up with arguments that seem fair in terms of we'll split production cost savings in proportion to market caps. Sorry, overhead in proportion to market caps, volume discounts in proportion to readers, know-how based on where it's coming. But ultimately, instead what you should do it just ask, what's the total savings that are created? In short, what's the pie? Here it's 9.85 million. And divide it up 50/50. But that's not the way people normally look at this negotiation. They normally argue every bit by itself, and that creates great difficulties. It makes things just much harder to do. So yes, we have to challenge ourselves to figure out what the pie is. But boy, life is easier once you have a sense of what you're going to be doing with it. And let me take an example of this. So, one response here is that every argument can be turned around. And in particular, once you see things from the perspective of the pie, There's never an argument for why one side should get more than half. So let me take you through and example of this from my own life. Some of you may know that along with being a professor here at Yale I started a tea company with one of my students, Seth Goldman, class of 95. And it's called Honest Tea. Honest Tea, I guess, recently sold its billionth bottle, so things are looking good, and we have some product for you after the session here today. [APPLAUSE] >> Do we have to split it equally? >> Gotta split it evenly? [LAUGH] Oh boy, that's good. So at the time- >> You should be paying us to drink. >> Yes, come see me afterwards. >> [LAUGH] >> So at the time we were in negotiations with Coca-Cola, they came to us and said, you're paying $0.19 for each one of these plastic bottles. We can get them for $0.11. And at the time we were selling 100 million bottles a year. So what is the savings that Coca-Cola can make possible? $0.08 cents on 100 is, $8 million, okay? And now the question is how should that be divided up? So Coke says well let's see, our sales are 40 billion, your sales are 20 million. That's 2,000 to 1. Okay so, you can have 50,000. We'll have 7,950,000, which doesn't really seem fair. And so, one of the things I point out is that, what might look like it's fair in terms of dividing up the costs between the Gazette and the Planet, in terms of sales. When it's kind of 2 to 1, once you realize when it's 100 to 1, 2000 to 1, those arguments clearly break down. But a principle shouldn't really depend on the ratio. And so, what's my response to that? Should be 50/50. And they say, well, wait a second, you know. You can't get those savings without us. And what's the response? >> [CROSSTALK] [INAUDIBLE] >> It's your know-how, but it's all those consumers that like drinking that expensive organic delicious tea. And without those 100 million consumers out there, your cost savings don't really help. And so then they say, well, okay, look, we're Coca-Cola, you're Honest Tea. For us, $8 million isn't isn't even a rounding error on a rounding error. >> [LAUGH] >> Whereas for you a million dollars really matters. So, we'll take seven. You can have one. >> [LAUGH] >> And what's the response? You're told it's a rounding error of a rounding error, so give it all to us. Who will notice? [LAUGH] >> So you can't really make the argument that something doesn't matter to you, therefore, you should get it all. And my point again is once you see something in terms of the pie then everything really ends up being symmetric. Now that story turns out to be a little bit embellished because ultimately what happened, oh is that Coca-cola ended up buying the company. And so therefore what really mattered wasnt the cost savings just for the bottles but how much they ended up paying for the company as a whole. And here's a case where we really did apply this principle to the pie. Coca Cola wants to buy the company. We're saying we're not really quite ready to sell. We're having too much fun. But we're willing to sell to you in three years. In the meantime, Coke is going to help us with lower costs. They're going to give us their distribution. And they say, okay. But we don't wanna pay for sales that we end up making possible. So how should we structure the deal? >> Split the gains. >> Split the gains. So basically we say, okay, here's out proposal. What you do is you pay us the market multiple on sales up to x, and any sales above x, you should pay us half the market multiple. And what is x? X were the sales that we could have made possible without you. So if we could have got to 100 million without you, then you should pay us the market multiple on sales up to 100 million. And everything above that we could have done without your trucks. But you couldn't have done without our product. So only by putting our product and your trucks together can we now create this extra pie, so let's split it 50-50. And it turns out that we agreed on that framework pretty much in the first hour of negotiations. Then there was some dispute over what we could've done without them, so what was x, and what was the market multiple. The market multiple is really a piece of data. And they could show us what all the other deals were, and we could try and figure who we were most comparable to. So really the argument was about x. But that really I think help focus our discussions, and made it much less confrontational, and in terms of why there had to be a pie, well look that's why we're doing this deal. If they couldn't sell more than we could've, then there's no reason for them to have purchased us. And so they can't just go and claim the pie is going to be small, otherwise there is no reason for them to buy our company. If you don't know what the pie is, how do you solve that problem? So imagine they don't know how many bottles we're gonna sell next year. How do you resolve that? So it may not be 8 million. How do you handle the uncertain pie? Continuously. So basically if you said instead we'll sell you the bottles at $0.15 a bottle. Then we're saving $.04 a bottle, they are making $.04 a bottle. Whatever the number of bottles is, we're splitting the pie. So essentially we agree we'll measure the pie ex post and split it, or in subtends we'll split the pie 50/50 as every slice gets made. So that's another way of approaching this. >> [INAUDIBLE] >> This product is incredibly not sweet. And so, it was their first organic. We call it, just a tad sweet product. And so, we're really going after the Snapples and Sobes of this world, not going after the Coca Cola drinkers. So we like to think we're expanding the pie, but I'll let you be the judge. Another example you can think of is a merger between Rio Tinto and BHP. So these numbers, it's crazy to think how big they are, but Rio Tinto is worth about $160 billion. BHP is about $240 billion. They thought by coming together, they could create $30 billion of synergies in the Pilbara, the Australian Outback. And so the question is, how should they split up that $30 billion? And what would be the natural way of doing it? >> Size of their company. >> They split it up in proportion to their market caps. Which is 3 to 2, or 12 billion to Rio and 18 billion to BHP. And what do I think the answer should be? >> [INAUDIBLE] >> 15/15, so what's that worth? 15 billion compared to 12 billion, that's worth $3 billion. So, if only I could get a little piece of that. Unfortunately, this, or fortunately, this merger ultimately was blocked by the European Union. So, it ended up being irrelevant question, but this perspective of the pie can really make a difference when you're doing negotiations. So, when people traditionally go this way, they can yell, they can bully, they can do the five, four, three, two, one but that's not an argument. Now the folks you are going to be negotiating with are going to be receptive to principled arguments. And once you have this perspective of the pie I think you have the ability to persuade people. >> Your goal to get the best outcome for you. If you're BHP, >> I would not- >> Wouldn't you come to the table with the proportion. If you're Rio Tinto, wouldn't you come to the table with- >> I completely agree with you. As BHP, proportional division is the way I see the world. And I would not present this lecture to those folks. But if I'm Rio Tinto, I would make this argument and more than that having made it, I don't have a counter as BHP. Okay, so I think the BHP story, the proportional ultimately does not carry water. It is not, and you see that when you take it to the extremes as example with the 40 billion, 20 million, Coke and Honest Tea. So people think it's fair, it sounds fair. That doesn't mean it really is. And so, if you like, essentially, when people do negotiations, they try to make arguments about fairness, and that before today, I think you had one argument of fairness, which was proportional division. And I'm not against proportional division, it may suit you well, in some circumstances. But today, I've doubled your options because I've presented this idea of the pie and once you see things in that perspective, life gets much simpler. And it allow you to think about what is a truly fair division, which is 50-50.