Bob said, "I want to add one new requirement to your project that reads something like this: 'KillerWattSoftware shall deliver the most important features to MegaCorp in a salable state three months after the start of the project. Then, they shall deliver the remaining features in three more similarly-sized, prioritized mini-projects-one every three months."'
"Whoa there, Killer," said Peter. "That's quite a mouthful, dude. I don't know what you mean."
"Sorry, Mega," said Bob. "Let me translate from formal requirements language into normal English. What I mean is this: Instead of delivering one big project that takes a full year, I want to deliver four smaller projects, each of which takes around three months. The first mini-project delivers FBU 1.0, which you guys sell and earn money from. It will be built to the same standard we always produce—MegaCorp worthy as you say—but it will only have the most valuable features. The second mini-project delivers FBU 1.1. It has more features, attracts even more customers, and brings in even more cash. Then we do the third and the forth project."
Peter screwed up his nose. "That doesn't make sense, dude! Who'd buy a quarter of a product?"
Bob said, "According to Charlie, our usability expert, you could turn off most of the features on most of the software we've shipped at KillerWattSoftware and no one would notice or care, because no one ever uses them. He's done studies. It's called the 80/20 principle: 80 percent of the value in the software comes from 20 percent of the product." Bob pointed at the spreadsheet and said, "We think the same principle applies to FBU, too."
Peter took another slurp from his beer. "Nuts!" he said.
Bob frowned. "Why's it nuts?"
"It's not nuts. I need nuts. This beer is making me crave salt."
He got up and went to the bar. Bob figured that Peter didn't really want nuts; he just wanted time to absorb this crazy new idea. A minute later, Peter returned with a bowl of nuts. He sat down and said, "OK, I buy your argument. It makes sense technically. I mean, the first iPod didn't have half as many features as its competition, but it succeeded because it did what it did very well. They added features over time, and more and more people bought it. And Google went live with Gmail when it was still a beta. They gave their customers a good product, loads of people started using it, they kept adding features and improving the service, and, all that time, they were earning advertising revenue. I'm not qualified or authorized to judge your suggestion from a commercial viewpoint, but if your relationship manager proposes it to our business folk, I'll support you from the technical point of view."
"Great!" said Bob. He wished he had ordered two beekeeping books. "Look, how about if we run some numbers on my laptop? I don't know, and I don't wanna know what your internal costs are, nor what your revenue predictions are, so we can just make some numbers up that feel good enough."
Bob opened up the spreadsheet, and within ten minutes they had prepared a simple model of the FBU project's "bank account." Bob suggested they assume that the costs were a static $1 million each month and that FBU 1.0 earned only three-quarters of a million a month, FBU 1.1 earned a million a month, FBU 1.2 earned $1.2 million each month, and the final version earned $1.5 million each month. They had no idea if the numbers were valid, but the approach needed a whole lot less of MegaCorp's cash than the original plan, and it painted a far prettier picture than cancelling FBU. The previous plan, with just one release at the end of twelve months, meant that MegaCorp built up a $12 million "debt." With the new plan, the most the project ever owed MegaCorp was $3.75 million. That was still a lot of money as far as Bob was concerned, but it was a lot less than $12 million. Who knew how real these numbers were? Only time would tell.
Peter selected a few columns on the spreadsheet, clicked a button, and a graph appeared.
"Wow!" said Peter. "They say a picture is worth a 1,000 words. Well, that picture right there might just be worth a few million dollars, and it might just save both our jobs."
Bob nodded. "Your granny will be pleased."
Peter said, "Yeah, she will. But look there. The FBU project pays for itself by the fifteenth month. If we'd done it the old way, with just one release, then MegaCorp wouldn't have gotten all of its money back until the twentieth month."
Bob pointed at the screen and said, "You see that gap between the two lines. It seems to be constant. Can you figure out how much it is, Pete?"