Dan Bricklin invented the spreadsheet amongst other things. In this talk he shares the knowledge he has gained from producing software that people buy for over 40 years. A brilliant talk from someone with absolutely nothing left to prove in the software business.
My notes from the conference talk are here: http://thebln.com/2010/10/dan-bricklin-business-of-software/
BUSINESS OF SOFTWARE – FOR PEOPLE BUILDING GREAT SOFTWARE BUSINESSES.
This year will be the 7th Business of Software Conference – 15-17th September 2014, Boston.
A three day conference for founders who want to build sustainable, profitable software businesses. BoS has always been a special conference for our delegates and we want to keep it special. Attendance is restricted to just 400 attendees in 2014. Registration open now.
This year we are also running Business of Software Conference Europe, 25-26th June. For more details, visit the dedicated BoS Europe site. Two days in the birthplace of computers – Cambridge, England.
Join our mailing list to be the first to know when this year’s talks go live and to hear the latest news from the conference.
Joel Spolsky: Alright, our next speaker needs no introduction so I’m not going to introduce him. Please welcome Dan Bricklin. [applause]
Dan Bricklin: Thanks, thanks very much. Um, I’ve been follow-, I’ve been out of town until late last night, but I’ve been following this on Twitter and I realize that you guys have already seen these great, great speakers, so if I mess up it’s okay because you’ll already seen great stuff, which takes some of the pressure off. So, I’m gonna cover a lot that I’ve learned over the years. Let’s see, so I wanna talk about all, there’s always new technology. This is stuff that you can, that should be able to serve you for a long time. When should we care about new technology and how do we take advantage of it, because that’s often how we’re gonna be making money and stuff and which products to do and stuff of that sort. So I’m gonna try to touch on a lot of that. And I’m gonna try to apply some sort of a methodology.
What have I learned over 40 years? Um, it’s been over 40 years of developing products that other people use, ay-yay-yay. How many people have been developing for 40 years or so? [laughter] Yay! I saw one or two, yay! Okay, well, so I went from, on the Apple II I was one of the most popular productivity products on the Apple II and I have one of the most popular productivity products on the iPad now. Okay? I wrote a little bit of code for this and I wrote all the code for that one over there. So, and in between I own the one laptop per child, I have an open source spreadsheet that I did and Social Tech sells it. So, I’ve been on, whatever the technology is I might have touched it. Um, so I might be able to comment on it with a little bit of knowledge, which is of course dangerous. So, when there is new technology or some new opportunities, you have to ask yourself some questions, such as: “What is this new thing for?” Ya know, what about it is better and will it do a needed job, which I’m gonna talk about and how can I make sure whether this is the right thing? Um, some of the concepts I’m gonna be covering are these jobs to do 2:24
So, why do people buy products? Okay, why do they actually? There’s Clayton Christiansen who wrote the “Innovator’s Dilemma” and the “Innovator’s Solution.” Have people, anybody read those? They’re good books, especially the “Innovator’s Solution.” You read it and it’s very slow reading because as you’re reading it you’re always thinking, “Hmm, how does this apply to me? How does this apply to me?” so it makes for very slow reading, but it’s a wonderful book. Um, Clayton Christiansen wrote, um, that “Customers, people and companies have jobs that arise regularly that need to get done. When customers become aware of a job they need to get done in their lives they look around for a product or service that they can ‘hire’ to get the job done.” So, what we need to do is identify the job and see how to make a product that meets the job’s requirements. Otherwise they’ll hire or buy something else. So, I’m gonna, I’m here to do some history so one of the histories I’ll do, is I’ll do VisiCalc. So what was the “job” that I was building VisiCalc to do? What was I thinking about? It’s not what most people think, okay? Because what a lot of people think is that what I was doing with VisiCalc is I was building a thing that had to do with columns and row of numbers, right? Because like which is what we’ve had for many years before it’s like a database printout. It was all about columns and rows of numbers and stuff. That’s not what it was. Um, let’s switch over to this and this, voila! This is a spreadsheet, okay, historically. This is actually my father’s spreadsheet from his printing business, um, from back in 1960. And you’ll see all, how this was all done. This was all done by hand. This is how it used to be done, believe it or not. Of course as one of my daughters said “Daddy, did they teach spreadsheeting when you were in school?” Then she went “Oh, no!” [laughter]
So, I wanna talk about my friend Bob Frankstein and me. That’s what we looked like in the early 1980’s. That’s a publicity shot, yeah. [laughter] So I was yeah, a little different, yeah, I know it’s kinda fun to look at this and say. Of course the hair is actually darker in the photographs because they don’t have the same dynamic range. It was actually dark brown. [coughing] So, one of things that I think you heard about is that it’s good to have two, two people. We were lucky we had two people and we’re still very close friends. Um, last time I talked to Bob was yesterday. Um, we met at the Massachusetts Institute of Technology in 1970 and, uh, we both worked on the Multix project. Multix is, uh, the precursor to Unix, uh, which is the precursor to other stuff we know about. We worked in this building up here at 545 Tech Square, the artificial intelligence lab was on the top floor. Uh, Edwin Land ran Polaroid out of this building over here. And that was a major time sharing system, um, but it wasn’t very commercially successful. I learned about one of things that marketing matters. Uh, it doesn’t matter just, it’s good to have a good product and the concepts of Multix have continued to this day. Uh, but, um, you have to look at the other stuff. Um, after going to MIT, um, when I graduated in 1973 I went to work for Digital Equipment Corporation, may it rest in peace, and, uh, how many people here worked at DEC? Yay! What’s your badge number? [laughter] But, whatever. That was the thing, remembering that if you had a low badge number, wow, I was with Gordon Bell this weekend and he had a really low badge number. He’s still working, he’s at Microsoft now doing some really interesting work. Gordon Bell was the head technology person at DEC and invented what we call the mini-computer today.
I worked on their computerized typesetting business which was kinda cool cause I came from a printing family, a family of printers. And I got exposed to this machine. This is the Harris 2200,in the early 1970s. The Harris 2200 was kinda like a page maker machine except with hardware. It was, uh, you had a keyboard and you had some arrows keys or something and you could choose slugs of type and then you could adjust the point size and its position and stuff. So I was exposed to this way of being able to lay out classified advertisement because classified ads were important.
I then worked on DEC’s first word processing system later sold as the DECmate. This is the early 70s. This is before Wang was doing screen based word processing, let alone before, uh, MicroPro doing WordStar or today’s Word or any of those things. [cough] And we used this as a display, character based display. We had to teach Xerox how to be able to make printers that printed both directions and I wrote the printing code and all sorts of stuff. I was project leader of that project when I was at DEC, but I really wanted to start my own business and when I was at MIT I learned all this cool stuff about development and it really helped.
So I figured, maybe I should learn all the cool stuff about business and maybe that will help. So I went off the Harvard Business School to learn the magic of business. This was I guess it was 77. So I’m always into history so we had, this is me sitting there, we were, ya know it’s like this except you’ve got 90 people in the room, okay? The guy over there is now the governor of New Hampshire and in one of my other classes the woman over there is Meg Whitman, might be another governor. Uh, we had some interesting, and Clayton Christiansen was also in some of the classes I was in. But, we were doing pocket calculators so I brought in an old pocket calculator that I borrowed from a friend of mine who happened to be at the time I didn’t realize how important, he was one of the people who helped develop the slash between TCP and IP, uh the, who wrote the paper, well that’s, he’s the main [laughter], this, if you know, if you listen to what Lessex says, one of the most important papers that was written, it was called the “End to End Argument” about how with the internet the ends decide things and you don’t optimize along the middle, okay? You do things at the end and then you can improve anything you want also long at the ends do. It’s a great paper called the “End to End Argument” and David Reed is one of the authors of that and I borrowed, that’s his calculator.
So, what we do at Harvard Business School is we do cases [cough]. We take courses: Business Policy, Corporate Financial Management, Management of, Advertising, all these cool, Consumer Marketing, Consumer Behavior. We learn all business and stuff and we do cases. A case is a business problem of some sort and um, ya know, there was Mr. So-and-so and he brought his people together and the CFO, she said this ya know, and we get to have the exhibits. So we have to figure out what should they do, what ya know, we have all sorts of problems. We do three of these a day. They run anywhere from 10 to about 40 pages each and you do three of them a day and have to present the case.
So, this is what the exhibits look like. They look like this. You start seeing a lot of numbers and letters , but they’re all different. This is the type of stuff that we were exposed to day in and day out. That’s what homework looks like. This was my homework, okay? Notice I cross out things very often. Um, so you get the idea. This is what, ya know, and you make mistakes and of course you come into class and they say “well, present the case” and I write it up on the board and if your numbers are wrong, ya know forget it.
But we had time sharing from DEC. That’s what a I looked like, hair down to here. These are the programs that we got to run. They had all these nice can programs. This was about upstate canning. And if you wanted to make a change of course you had to know how to program and of course almost no one knew how to program in the class. But, see “print spreadsheet” and it prints out the numbers. So, this was the type of stuff that we were doing. But if what you wanted to do wasn’t one of these options you were kinda stuck, okay? You would have to go back and ask somebody to program it. This is what business was like in those days.
This is some more of my homework and this is spreadsheet paper. This is what I took my, wrote our, would do homework on and stuff like that and do my doodling on. The name of the newsletter was called “The Spread Sheet,” so I was kinda into spreadsheets. We also had calculators. This is the TI Business Analyst calculator, this was my calculator and um, we got really close to your calculator. It was your best friend. So much so that for Halloween I dressed up as a calculator [laughter] okay? [applause]
However, I had a, one of these Core 63 degrees from MIT which is electrical engineering and computer science, so inside I was “Integrated Circuit Man.” [laughter] So, I had this, ya know and when you have this many people and you’re sitting sorta down, I was like sitting right over there and you can daydream a lot. So I would sorta day dream in class and ya know, ya know I got numbers wrong a lot and I’m from word processing. Ya know, imagine if my, if my calculator had a ball on the bottom and I’d seen a mouse, this was 1978, okay? So, um, the mouse was not in general use but I had seen one. I think I had seen an Alto also around that time. And imagine the mouse, my calculator actually was a mouse. It has enough keys to have all the numbers and the letters and imagine I had a head up display like a fighter plane, ya know? And I could, ya know, like wordprocessing like this and I could circle it and push the sum button, there’d be a sum, so I could look somebody straight in the face and I’d be able to do a negotiation and say yes, a 10 percent raise would be acceptable. Ten percent was always our answer in those days for some reason because it was easy calculate. The cost of capital was 10 percent, ya know those days, um, interest rates were always 10 percent.
So, I imagined instead of doing this, let’s do this. So, um, I went to my business school professors to see, was this a good idea to do? This was the Spring of 1978 and um, might as well get professional advice and I tried to go see my finance professor. He wasn’t available for a few days so I went to see my Production Management professor and he said, “Uh, I don’t fully understand what you’re talking about, but you’re talking about they could use a modem and work on this at a distance. Do you know that people take blackboards, two rooms full of blackboards to do production calculations and they fly in to the same city so they can do it together. This sounds like it could be a good idea, whatever it is.” Then I went to my, um, Accounting professor. The professor who was teaching me about cost accounting, Professor Jim Cash, what a great for an accounting professor. [laughter] Jim had a masters in computers and had actually programmed accounting systems and all and he said “Ya know, Dan, I can’t visualize what you’re talking about,” you’ll find that no one can really visualize your product that well, “but I understand it’s about human interface and human interface is real important. Ya know, this sounds like a good thing to do.” Jim went on the become a tenured professor and head of all stuff and he’s on the board of General Electric and Microsoft.
Then I finally got to see my Finance professor and looks up from his four tran listings and said, “But Dan, there already are financial forecasting systems. Why would anybody want yours?” Okay, so here I’m totally deflated, he said “Why, I have a second year student, you’re a first year student, I have a second year student who’s doing some work about personal computers and he’ll tell you why real estate agents won’t buy your things. His name is Dan Filester. He has a company called Personal Software and they’re selling.” So I called him up. Um, I decided, and I just sorta learned a little bit. He said “We’re doing this personal software. We have MicroChess and whatever you do, you should do it on the personal computer.” I said “That’s sounds kind of interesting, but I’ll worry about it later” and didn’t tell him what I was doing.
That summer, the summer of 1978, I was riding my bike on Martha’s Vineyard, going along I said “Ya know, this sounds like fun, I think I’ll start my own business and I’ll build whatever this thing is and sell it door to door. Ya know, cause I’ll probably have to sell in piece of hardware and sell the whole thing. But ya know, I had a DEC mentality of you had to sell, ya know, your business software with the hardware and all that type of stuff. That’s Martha’s Vineyard, by the way, I don’t know if you knew that.
So I decided to form a company called Software Arts. My friend Bob and I founded the company. I borrowed an Apple II from him in the fall of 1978 and prototyped in Basic what became the spreadsheet. Well, there was no mouse. So instead of the mouse I figured I use the paddle. Remember the, the game paddles? It was for Pong and Pong was very, Space, Space Invaders, that was the reason you bought the machine.
VisiCalc was an excuse, but the real reason was, we used Space Invaders so much that the fire button broke and we had to put a heavy duty button in its place. So I used it, you could turn it and that would move the cursor this way and then you would hit the fire button and it would move it this way and that was a poor man’s mouse, but it was our C-circuit which sort of had drift and it didn’t work very well. So I switched to the arrow keys and that’s how we ended up with arrow keys. But, there were only two on the Apple II, luckily the spacebar acted as the fire button so you could switch it between this and this. If you remember the original VisiCalc did have just two [cough]. I had prototyped in Basic on the DEC system that they had a Harvard the idea and try to, got the idea of generally putting numbers around like this, which is what I wanted to do, into columns and rows so that I’d have an easy way of typing something that a human would know and the computer would know and that’s why I used columns and rows so you had a naming, an easy naming way. It also let you do things like ranges and stuff.
So I put all that type of stuff together. And we looked at the prototype, it looked good. We had a deal, so in those days it was author/publisher. So Filester ran this company called Personal Software. They would publish the product. Bob and I had our company called Software Arts and we would write the product and they would sell it and pay us a royalty or for accounting purposes we would call it something else. Accounting, it’s always a mess. So, Bob was working out of his house. This is Arlington, Massachusetts, about 10 miles from here and he worked up in the attic. We used time sharing to do our development because they didn’t have good compliers and stuff on the Apple II. We actually used Multix. We actually used it at night when it was cheap. So Bob worked late at night, he would actually be up all night, sleep during the day and he’d wake up in the afternoon, I would show up from class and I would do debugging and I wrote the documentation, what it should do, and he worked from my prototype, etc. This is the attic where VisiCalc was done. So it probably looks familiar for some people’s start ups and stuff like that. You can see the acoustic coupler over here. You remember those old days you had to be able to connect to time sharing and that’s the Apple II that we used, etc. There’s me talking to Bob’s father with an Apple II. This was Personal Software, the largest publisher of personal computing software in the world at the time, of games and stuff. They were the top three rooms up there in Allston, Massachusetts. They used the money from MicroChess to pay us to be able to do development.
Marketing is hard. You have to come up with names. The name VisiCalc was, um, the people who came up with it were Bob and Dan Filester came up with it while eating at Vick’s Egg 11. Both of them think that they came up the name. I said, “Well you could have looked at the menu and saw Vick’s , Vic, VisiCalc, who knows?” The Lord works in wondrous, mysterious ways, no? When I was a child I was into photography. My dream was to be in Popular Photography magazine. This photo was in Popular Photography magazine because I gave this talk many years ago and one of the people there wrote for some electronics magazine but also for Popular Photography was doing an article about people who do their own slides. [laughter] [cough] So, I ended up in Popular Photography. I also ended up in Sports Illustrated and the Men’s section of the New York Times, okay? [laughter] The Lord works in wondrous, mysterious ways, okay?
So the first ad for VisiCalc was in Byte Magazine. That was in May of 1979.This was the teaser ad that our publisher put in. Now, being an engineer I was not this aggressive. He was very, well even though he was an engineer he was a very aggressive marketer and “How did you ever do without it?” This was the teaser ad. They eventually, and of course it turned out to be true if you were an accountant. In those days, TI was about to enter the hardware business. Everybody was shaking cause Texas Industries had messed up other industries coming in and they were afraid that they were gonna be the big one that would knock everybody out. Radio Shack, fifty percent of these machines. Apple shipped 20,000 systems valued at 30 million. iPads alone during this conference, Apple shipped more than that, but you know. Now, we were still in development. This is the Spring of 1979 we’re still in development . Our product is secret, but I want to try it out because you’ve gotta test, test, test, test.
So, I was doing this case, Pepsi Cola case, which just so happened to be about Pepsi Cola, a case about John Scully, who later on went to head Apple and about something that he had done. So I went over to Bob’s and I ran the numbers, but we didn’t have “print” working yet, so I had to write everything down. We didn’t have “save” working yet so when it crashed I had to keep retyping all this stuff, but [laughter] look at the precision that I got, ya know, it was incredible precision. I ran five year projections and all that and I went to class the next day, raised my hand and the professor called on me and it’s Marketing class and said “Okay, Dan, present the case,” and I said this and that and we do five year projection and you find this and we try it at this percentage and that percentage and here’s the answer! And wow! Dan really aced the case there. But wow, that’s cool, how’d you do it?
Can’t tell them about my secret product, so “Well, I added this and subtracted that and multiplied by that and subtracted this and then that was the answer.” “Ooh, well why didn’t you just use a ratio?” “Oh, well that wouldn’t have been as exact.” What I didn’t tell him was “divide” wasn’t working yet. [laughter]
It’s the real world! We moved out of Bob’s and we moved into commercial quarters in Central Square near MIT. Down in the basement. [laughter] Now, when you get inexpensive space there could be problems. We’re right next to the subway so when the T goes by it rattles. But one of the worst things is that we were low, below sewer level so, okay, so when it rained, if you didn’t remember to close the toilet it would back up. So we bought our, we mortgaged our souls, borrowed money and bought our own time sharing system, a prime minicomputer to do development on. So here’s printout, here’s projections of what we’re gonna do, ya know how many units we’re gonna sell, when we’re gonna ship, of course we didn’t ship the month we said, we really shipped there. But it turned out that these things were, were relatively accurate, even though it was off quite a bit and of course it said we can’t pay ourselves salaries which is always the case. That’s what happens when the water backs up. That’s our life savings backing up the computer over there and water over here and you have to do things, ya know, sucking up all the water. Okay, we hired a programmer to fix the divide.
This was me announcing the product privately at the West Coast Computer Fair in 1979 to the computer press. Be real careful when you announce something that you’re not done with. One of the things that I, we mentioned that it was going to have all the transcendental functions, like sine cosine, stuff like that, but they weren’t done yet. Unfortunately, Carl Hummers, who’s a friend of our publisher Dan Filester who was the head of Byte Magazine, half of his review, which he wrote in advance, wrote about sine and cosine stuff so we had to finish it and I ended up, that’s the code that I ended up putting into VisiCalc was the transcendentals. We announced it publically in June of 1979 at the West Coast Computer Fair, I’m sorry not at the West, at the National Computer Conference, which was a big computer conference for big computers and off to the side, that’s in Madison Square Garden and in the Hilton or something, in the basement is where the personal computer conference was.
Francis X. Clines, who writes now for the editorial part of the New York Times, wrote this “A Layman’s Trip into the Mega-Mega Land of Computers” Ha, ha, ha “laymans” sounds like the gag reflex of the brain ha, ha, ha, sounds very funny. But he happened to be at the place where they were making signs to put next to our displays. “Even as the believers gather, the painters in the Coliseum sign room are adding to the pantheon, carefully lettering ‘VISICALC’ in giant black on yellow. All hail VISICALC.” The New York Times, all hail VisiCalc. That was the last mention of VisiCalc in a major business publication for a few years. But, Ben Rosen who worked at Morgan Stanley had this conference that was later given to Esther Dyson wrote about it “so who knows VisiCalc could someday become the software tail the wags and sells the personal computer dog. He then went off to become a venture capitalist, funded Compaq and was the head of Compaq for many years and funded Lotus which did 1, 2, 3. He knew what he was doing.
When we shipped, Apple actually had us in an advertisement on TV. Dick Cabot who had never done an ad before, did an ad for Apple. He’d push a button and up would come VisiCalc, he didn’t know what he was doing about it, but it was kinda cool to see our product on TV. That’s what it looked like. And that’s what it looked like on the screen. That’s what the good ol’ days were like. And here we go, that’s, so that’s how, um, now when you go to Harvard Business School, that’s the classroom where I thought up the idea and there is a plaque to say “You too might come up with a good idea.” Apparently people go over to it before tests and touch it. [laughter] This is one of my scribbles I did that ended up being the reference card on the back of some of the handouts.
Here I am back at Martha’s Vineyard with a book that I have that they didn’t give out to you . It’s called “Bricklin on Technology.” Okay, that’s it for this. Let’s go back. Here we go. So, what was the spreadsheet job? It’s a tool for doing a wide range of things you do with numbers on a piece of paper. It’s a general tool for a business person, like a word processor was a general tool for the write or the typist. It replaced a pencil, paper and a calculator. It wasn’t just for financial forecasting, it wasn’t just for reporting, it wasn’t just for mortgage calculating or checkbook balancing. All of which we were building it for. It was most anything that you might want to do. Why was it widely used? Whatever you wanted to do, it could do, within a certain range, okay?
Promoted unanticipated uses. This is real, this is one of the reasons it became wide, these type of tools. And I have a thing called the “Tailoring at the Ends Rule.” Applications in the productivity world are mainly custom, like all sorts of stuff even in the entertainment world and stuff like that. If you look at the variety of even standard things like an income statement, look at all the variations of it, let alone production planning, which is so specific to each. It seems to the lay person like they’re all the same, but they’re not, there are all the little differences. So general purpose tools win. If there are lots of things like nails, we want to invent a hammer. Now you do that by allowing people to tailor at the ends to meet their diverse needs. You build something so you can deal with unanticipated uses. This is what fosters innovation on the part of the user. You want a tool not an automaton. Instead of choosing out of anticipated cases, you want to do customated, customizing out of unanticipated cases. You mold something, so instead of building a model airplane kit you build a logo kit. Instead of bagels, you think of flour. Now, the trick is to provide an understandable framework.
That was one of the things, with VisiCalc what you did is, we had a, what we had was columns and rows and when you have this framework the user can figure out for themselves how to get the job done. VisiCalc’s framework with rows and columns imposed a slight order and restricted you a bit, so you couldn’t do everything you wanted, but it opened up a whole lot of things like it allowed you to do summation and stuff like that. But the freeform positioning of arbitrary cell references kept the framework from getting in your way. VisiCalc took care of the tedious parts of the job but didn’t define the problem being solved. It had calculating and recalculating of texts, numbers and lines, and numeric formatting which is a pain in the neck. It got rid of the tedium, but didn’t get in the way. Now, sometimes the person tailoring is a specialist, but often that’s just to get the user into the ballpark and they can find their own seat, okay? Which is why you have to have templates and stuff like that. But there’s a problem with things like wizards that sort of are like a cabbie that drops you off in the middle of nowhere and if that hotel, if that restaurant is closed you don’t know what to do, you don’t have map, okay?
Now, what did we learn from selling it? Well, it was as a really easy sale. It did what they wanted, it was, um, if you were doing time sharing, which in those days was what you did to do financial forecasting, you were paying $5000 a month. Well, you could buy an Apple II with a printer and our application for that and it was a two week pay back. [cough] People were using it for business deals. I found people at EDS which is this huge company that mainframes all over the place, they were using VisiCalc to do their deals. Hours of works went to minutes. People would push credit cards in your face when they saw it they wanted to buy it so much. So I had this thing called a Two-Week Payback Rule, something’s like a no-brainer to buy it, that helps cause there’s very little risk on the part and the first time they use it, it pays for itself. That really helps. Desktop publishing was that like that the first time you used it.
Now there’s another thing. You all know about the long tail, right? Anybody not know about the long tails? Okay. Value to some user, value to seller, percent of revenue, but you mainly have, sometimes people have this and sometimes have this. Amazon does this XXXX. Okay. So it’s useful to have many, many, many options, but there’s this thing called hyper-differentiation. I heard a speech about this from Professor Clemmens down at University of Pennsylvania. “Assumption: With many choices available, a customer’s willingness to pay for products and services falls off quickly when the fit between these offerings and his or her ideal product location decreases.” In other words, if it’s exactly what you want, you’ll pay a lot for it. There are people who specifically did research on beer. There are people who love lots of hops in beer. Most people think it’s yick, okay, but some people just love it and they’ll pay a lot extra for it. But if it’s not exactly what you want you’ll pay a lot less, okay? It’s a nice curve like this. And he had this, he said, if there’s uncertainly of getting exactly what you want, if you know that that hotel usually is perfect, but sometimes they mess up, you won’t pay as much, okay? The curve goes down. I looked at that and said, “Huh? Hmm, this graph is interesting. So you know, a country inn without photos makes you less comfortable to make reservations and pay high prices. Give more information about it.” But I what I realized is, if I have a particular need that can be fulfilled by a tool, the tool I’ll choose will be the one that I know will more likely fulfill that need. If I have to limit my tool kit in advance, because I’m only gonna buy so many apps or something like that, to just a few tools, I’ll try to include in my tool kit just those tools that will meet as wide a range of need as possible from the range of needs that I think I’ll have in the future. That’s why you have hammers, screwdrivers, pliers, word processor, personal automobiles, all of those things are very general purpose tools that we know will handle things we might need to do in the future. Okay?
Then there’s this thing called Must-Haves. This is Donna Dubinsky. I was interviewing her in a Podcast. She was one of the early people at Palm and Palm Computing, and then Handspring and did the Trio and that stuff, but mainly did the Palm Pilot what they’re best known for. Changed the world. So, said “Rather than adding 3 to 5 applications to the existing built-in ones,” the palm came with your calendar and, ya know, name and address and stuff, “Rather than adding 3 to 5 applications to the device, customers would find one that was compelling for them and that to them was a make-it or break-it thing. It might have been where the stars were, it have been world traveler applications, it might have been querying a detailed database at work, but there was always some additional compelling application for the Palm owner. So people bought the Palm Pilot for the built-in applications, plus one long tail one.
So, the product that is likely to do the common things, plus the special one that you care about will dominate over ones that aren’t as flexible. So, think of a telephone. You have a few must-haves that are currently out here in the long tail. Think of the telephone. Everybody needs to call an airline at some point. Fewer, but still many call L.L. Bean. Many in a particular city will need to call for concert ticket reservations. Okay, it’s getting smaller. Only those who live in the neighborhood will call a particular pizza restaurant. Fewer will call an auto service station and even fewer, a particular individual. But, there are some individuals, my relatives, my friends, etc. that I really, really want to talk to, but my car is making a funny noise and I’m about to take a long trip. I really wanna talk to my car mechanic. That’s why I pay for phone service. I don’t just pay for it so I can call L.L. Bean or receive calls from a Verizon representative. Okay? It’s that the whole thing with a cell phone is, I don’t want to only have the ones that they chose are important for me. I wanna be able to choose any of them. And of course you have add those that I anticipate I might need in the future. I want to be able to have friends to call, friends I haven’t met, future customers and perhaps I wanna be able to call a doctor for a disease I haven’t come down with yet. That’s some of your thinking when you’re buying. So therefore, special purpose systems, restricted ones that only let you have the top videos don’t make it verses the general purpose system that let’s you see YouTube.
When I wrote this I was thinking QualCom was having this thing “Oh, we’ll have the hundred most best important videos coming to your cell phone,” well we really wanna be able to go to YouTube and see one of the millions. We don’t know which one it’s going to be. So, I’ll put up with lousy sound on a cell phone to talk to my friends, rather than clear sound on professional audio. So, this is “The tail that wags the dog,” another reference to that where results are often customized and you can’t figure out in advance, open general purpose systems. Not always the case, but in these cases it is. What was special about 1978? This was the emergence of the personal computer. It was a new beast. So part of seeing what makes special times. How as it different. If we can power this up here. There we go! In those days, mainframe. In those days we had mainframes, By the way this is my program on iPad. I do this, cause this is something anybody can do you don’t need to have a graphic designer or something. The connection here is relatively high bandwidth. You can move a lot of data back and forth there. And the connection here is relatively low bandwidth. It’s actually like 10 characters a second eventually maybe up to 100 characters a second.
So what is this good for? It’s good for typing in transactions and storing many, many, many, many transactions up there or for asking simple questions like uh, “How many of the people who were paid had their salaries suddenly doubled who had a surname the ends in X?” Lots of data back and forth, lots of computations, out comes the answer. Great for General Motors to ask that, okay? So that was the way people think of and then of course there were the minicomputers and the minicomputers of the day were smaller CPU, smaller disk and maybe fewer terminals. Though for example digital was the, minicomputer manufacture was one of the main terminal produces. Same layout, now you could be Joe’s Autoparts and ask the same type of question. Early PC. [cough] Well, that’s being generous, okay? [laughter]. And so, they didn’t have a disk, they had a cassette drive. They had a floppy drive that handled, it had 100k. And just one screen, so just one screen. So, that’s an audio connection. So this is the way the mainframe people thought of it. You’d be lucky if you could do accounting for one person, do payroll for one person. This is a pretty silly thing, what a toy! How could I use this? But it turns that’s not what it was.
Games were important. Cheap was important. To make it cheaper and to make it good for games, the memory here in the CPU was shared with the memory of the display. That made it cheaper, at least on the Apple II was built that way. So was, some of the other personal computers. So really what you had was high bandwidth here. This is a different beast. I could change the screen like that. Blows away any of those except for the CAD systems where it was the same configuration in terms of that. This is great for word processing and spreadsheet. You have to think about that. What happened is you have what I call two orders of magnitude improvement over there. So we also added high bandwidth to the page with the printer. We could put dots anywhere we wanted when the laser printer came in, okay?
With the mouse, we add high bandwidth to the hand okay? Each of these things improved by a couple orders of magnitude, the ability to do something.
Okay. Let’s go back the slides.
Number 1. So major improvements in capacity are fertile conditions for invention and innovation. I call it the “Two Orders of Magnitude” rule or the “Two Week Payback.”
You have to think about what is it really better for and then what can you do with that thing, okay? That’s why if you thought of doing payroll, the betterness of the spreadsheet was huge.
The other thing, Clever Exploitation of New Power, think of the GPS system. Computer power is so cheap now we can measure billionths of a second difference in stuff coming down, in things coming down from up in the satellites, billionths of a second difference in timing. Take huge databases, putting them together to make real-time maps, so if you happen to glance it will show you where you are. That’s brute force using lots of CPU, being very clever about using memory and stuff like that. [cough]
There’s also new modes of distribution all the time. When I came into the PC business at that point I’d been in the mainframe and minicomputer business, it was direct sales. That was the big stuff around 1977 where you had a salesman call on someone and sell them maybe hardware but sell them software and stuff. Then around 1978-9, 80, computer stores, and then computer stores became very big and selling through distributors to computers stores was the big thing to do. When Lotus came out with 123 they did a special deal with Softsell the major distributor and through that special deal they were able to get into all the stores really quick and working through the stores helping them really helped, okay? Then we had selling through magazines. Borlan did that. They were selling, they bought the back of all these magazines and sold things cheaper, okay? Each time, then corporate software, being able to sell where there were people who sold software directly, packaged software directly to corporations. A company called Corporate Software is big in that. Then we were selling through the web which we’re still doing with a lot of prepackaged stuff. Then we have Cloud based stuff, the Apps store. Each time there’s an opportunity to be a new leader, so look for new types of distribution, obviously.
So, situations ripe for success, 100 times improvement in capabilities, two orders of magnitude, understand the jobs, think about tailoring at the ends, two week payback, think about what you can do with the new power that we have that you couldn’t do before, okay? We used to sort of just do an approximation. Now we can do a table look up and get the actual thing. Look for new distribution.
Let’s apply this. The early 1990s we had these tablets if you remember. First with Go from Microsoft and even from Apple with the Newton. You could point with them, but they weren’t as precise as a mouse. So it’s kinda, for word processing you couldn’t get exactly on the character you wanted as easy. They were great for standing operating. They were black and white because of various technology reasons like keeping the power consumption low and let the pen work. [cough] But then color came in and color and a back light and was electrically noisy and made the pens not work as well. So, they were black and white and the laptops were color. They were really lousy for text and numeric input unless you had keyboard connected to it, in which case it’s a laptop. They were heavy, short battery life, slow operation. There was no such wireless connectivity. You were using the pager system if you had an Artist modem which was this big. And PCs in those days were I mainly for data capturing and processing so what good does this tablet have to do with you. I mean, they were marketing for people who were scared of keyboards. That’s not a very big market. But they looked like today’s pad computers.
Now the Palm Pilot on the other hand, it was easy to carry around, very long battery life compared to anything else, I mean it would last for days and days and days. It was good response, they solve the handwriting recognition by doing graffiti. It was much better than paper for scheduling and names and addresses and stuff like that. And there was no problem with black and white for what you were using it for. And you had your PC anyway and used it for backup so it was symbiotic- it didn’t replace the PC.
Now the early 2000’s tablets, the early tablet PCs, ya know, they were basically laptops but you could flip them around and stuff like that. But they had WiFi and 3G data on the internet so you actually use them for reading so that was kinda useful. They were expensive though, but there wasn’t much tuned it, there weren’t many general purpose apps and it was not pen-centric. You know, now the thing is that pens were very successful with an external thing in the graphics world, so pens did come in. But look at the iPhone and iPad. It’s not just a phone, the iPhone. It has a camera, GPS, WiFi, accelerometer, CPU storage. To me one of the most important things is the graphics processor.
Games again. Because it has a graphics processor it can be very responsive and have this nice feel to it. Touch and screen size. We now have internet connectivity. You know, there’s more data going out there on 3, ya know, one cellular network than there is voice. And there was an easy and inexpensive channel for apps. And whenever you needed data, the data was out there and you could get it now. None of this stuff existed back then. Could have, but didn’t so that, that works. And since it’s a phone, you’re carrying one anyway, okay? Now, the direct manipulation that it does, which is helped by the graphic processor, people don’t seem to realize how important that built-in graphics processor and taking advantage of it in the whole OS, makes the customizing the UI more obvious. So people, it’s more obvious for people what’s going on.
Now, Apple spent a lot of money advertising “There’s an app for that,” right? They had all the stuff for a few years? So people realize that “Oh, this is a general purpose tool, whatever I might want to do, there’s an app for that.” That’s back to that long tail. This is a long tail machine, okay? That ad campaign really works with that stuff. iPad, they got bigger screen. Because they have a bigger screen you could have a much richer user interface for both input and output. You have lots more room to be able to do controls and stuff like that and you can have two or three people at a time. It was the first computer that really is much more social. How many people have sat there with GoogleMaps and sat there like this with other people around with an iPad, and “Let’s go here or do this?” You can, people are using in sales now, where sales people sit there with Powerpoint on Keynote and are sitting there and it’s, it’s a much more social thing and that really matters. You can do much deeper applications on the pad than you can . Because you have more space you can do just-in-time training and stuff like that. The iPad is a different beast than the iPhone. What’s magical about the Apple iPad? They call it magical. What’s magic? If we look at the definitions, one’s “the exercise of slight of hand or conjuring” let’s forget that, that’s not the iPad. Number 3, “a mysterious quality of enchantment, “okay, well yeah it’s kind of enchanting, but to me, look at the first one. “The art that purports to control or forecast natural events, effects, or forces, okay.
Control things. What’s magical about the iPad is that you feel in control.
You’re in control of your book. You’re in control of the maps application in a way that you never felt before. Just that, you know how a magician just takes, how easy it is to be able to do things with a magic wand. Well you can just do the same things here is the way the interface is so it has that part of magic that you are the magician, you can control things. StarWalk, how many of you have seen StarWalk? Which is the one where you can hold it up, the iPad up to? Oh, oh you see you have to see it on the iPad. You can hold it up and it shows you the stars. As you move it the accelerometer knows where you are and the stars move to show you what star you’re looking at and you can zoom in and out and you can bring up dust, it’s so cool.
The books, it’s not just that the books are there, it’s that you can control the books. That’s the, there’s nothing like the picture application. You know the, through millions of pictures, just like that and choose one and make, ya know and show people and control. That’s an important thing I think about the iPad. [cough] Developing products. An important thing is prototyping. How many of you have been in Barcelona and seen this? Okay, yeah this is Gaudi. He designed this cathedral, they’re still building it. His name was Gaudi and we now use the word gaudy to mean something that has too many bells and whistles on it. It has all these nice arches and stuff like that in it. So what did he do? Prototype. He built, he, they had these drawings to show what it would look like. But this is complicated. How did they figure out how to build all those arches? This was the early 1900s. They didn’t have calculations like that. Well, he figured out something. The arch that he wants to do like this is mechanically the same as hanging weights on a string like this, except upside down. So he could model using string and a little bit of lead weights, he could model exactly for something that complicated and know exactly where things should be. The is actually one of Gaudi’s models and if you put a mirror underneath it you can see what it would look like standing up. Isn’t that clever?
This is like when IBM was testing out voice recognition by having someone in the other room listening and typing it. So when you’re prototyping, prototyping is important so you can learn about your area and learn what matters and what doesn’t matter, okay? Cause that’s real important. Cause after nearly 100 years this is being built as envisioned. I mean, what a great spec and stuff, but it still not finished. Ya know, hasn’t shipped, very patient investors. You should only have that, right? So, the hard part of feature design is what to leave out, okay? When we only had 32k and an Apple II, the help system went out. Well, it sort of at least would list the letters you could type, when you hit slash it would tell you the letters you could type. That happened to also be the look up table that we used to compare the thing that you typed was the same string. Remember they always talk about this. Here’s in sculpture. Here’s the David. I mean, this is one of the most magnificent things you’ll ever see if you go to Florence and did Michelangelo sit there and ya know, do it with clay, no, and envision, no. Michelangelo started with a block of rock and removed what wasn’t the David. Down the block, just a few feet away from the David is The Prisoner. Here you can see is a partially completed one. Yes I used my own pictures [laughter]. This is, you can see where it hasn’t been removed yet, knowing what to remove. Now he’s removing good things. He’s removing what would have been the hands on David which are incredible, but got rid of it because this is what he’s making.
Remember that, what to leave out is really, really hard. If you’re sure you must have that. That’s why prototyping is useful for testing to see whether it matters to people, what you should be adding, etc because you’re always, you have a vision for this and you have to ship something like that. Don’t forget to experiment, experiment’s a good thing to do cause you learn from failure. [cough] So, just because something failed or wasn’t good enough in the past, doesn’t mean that it won’t eventually have it’s time come.
So I’m gonna switch to a video, so this is one we’re not supposed to record. This is human flight. This is what people tried to do. That’s before they figured out the physics. Now look how determined, these people are so determined, okay? This is how people see you. [laughter] Yeah, you gotta do your testing. You’ve gotta be, you know you’ve gotta have that determination. [laughter] This one actually tries to use physics of some sort, but they were wrong, trying. [laughter] But eventually you get the physics, you get the materials and when you do that it can be much more beautiful than anything you could ever have imagined originally. This is what we’re dreaming about. This is what engineering’s about. Of course he’s just one person. This person’s a little bit of a show off. He goes into a stall and kicks off a ski. Now, personal flight isn’t the only thing where we’ve advanced like this.
Advancing is what’s always happening and you get to catch the time on that. Another example would be storage. That’s how we used to do storage, with the Rosetta Stone. Then we used to do it this way, put them in books and some of you will remember punch cards. Tape. I mean, that’s what it’s like, okay? Give you the credits here because that’s the least I can do. Yup, there we go. So, fast forward this a little bit. There we go. Now, today we have tools so many of us can be in the software business, okay? You can start out like many of us have started before. So now, there are a lot of us who are able to do things like this. You learn all you can. You come to conferences like this and you get ready, you check all your instruments, etc. and stuff and hopefully we have better tools then they had when they first started out, but if you’re about the jump off there you can feel all your family, friends, and the whole world watching you as you take that step off and launch your product and they think you’re gonna go down like this. But, let us hope that all of your cash flows will be updrafts, and that you’ll see that soaring as you go on.
Thanks a lot and do I have any time for questions? Yeah I have time for some questions. Yes?
Audience: I’m a like, a historian or some of your work from the era. Can you explain a little bit more about how you were using the Multix system as the complier and you brought back the ?. I actually don’t understand the relationship…
Dan Bricklin: There was a company called ECD that made a $8765.54 or something like that, they pulled it down like this, computer that was based on the 6502 and made into a, into it was like a Multix machine, it was a competitor to Apple, it was wiped out by Apple, but they actually used Multix and they actually wrote for Multix a cross complier that let you write in assembly language code on Multix, complie into 6502 assembly, uh, machine code and download it into your 6502 and run it. And that was available generally on the Multix system so we started with that and that gave us a very powerful macrocomplier, macroassembler, and Bob actually wrote, I think, three go tos. This is an assembler. Three go tos in the entire VisiCalc because he used a structured assembly language because you’re supposed to use if, then, else and stuff. We eventually, then, recreated on our own minicomputer. I wrote the accounting system and I wrote the editor that we used and Bob wrote the complier and linker and we used that because ya know, we’re talking about these little machines. You couldn’t have good environments. And we eventually then wrote, ya know, just like when you’re using the iPad and iPhone. You do all your development on X codes, it’s the same type of thing and you download it the same way. The world hasn’t changed that much, it’s just faster and bigger and more people. In the old days, that’s what the New York Times does. Now, if there’s a problem with Facebook or something, it’s one of the top things on the evening news! Who would have thought, ya know? It’s like, yes?
Audience: It’s pretty awesome that you have like all those old photos and papers and things like that. Did you know, like are you just a, do you keep that stuff by default or did you somehow know that maybe this will have significance or something?
Dan Bricklin: I’m, I’m a pack rat as my wife will tell you and I get rewarded for it. But no, what happened is very early on around 1982 or something like that, I was invited to speak at a conference. I dunno, it was a big investment conference, we weren’t looking for money. So I said, I’ll teach you what the business is like cause you’re all big investors but you should see what it really is like as a software business. I took some of those slides, the early slides that I took. I went back and took more and did the boxes talk [cough] with an overhead projector, not on an iPad. And then afterwards I said, “oh, I should keep all this stuff.” So I’ve just been keeping things in a pile knowing that I, and each time I did a speech I added more stuff. For the, for my homework and stuff like that I actually went back. Somebody was doing some research. Adam Green was doing some research for a masters in history and I went down to the basement and found all my old stuff and said “that will be good for after having read Christiansen’s stuff, so that’s only a couple years old. And more stuff I found to put in my book in “Bricklin on Technology.” So the videos , yeah I’ve always sorta kept, I got, if we had time, I got videos to show you, but I decided to show more of the other stuff. I can show you Bill Gates as a little, as a young kid, ya know and stuff. [laughter] Yes? Yeah, two more questions, okay, that’s one.
Audience: So are you disappointed with how little things have changed? I mean we see VisiCalc for 30 years ago and you can still see Excel in there. And also does Microsoft pay you every time they sell [laughter]?
Dan Bricklin: No, they don’t pay me. No they don’t pay. No, nobody pays. The only payment I get is if you go out and buy Notetaker HD, [laughter] for $4.99 and Apple gives me $3.50. I’ll get my $3.50. A lot of you are buying it, so it’s in the top 20. It’s been in the Top 20 of all paid Apps.
Audience: So like, are you disappointed, I mean it just looks so similar, it’s just, just like you said right now. Things aren’t just that different. But more mobile.
Dan Bricklin: Yeah, well it’s if you look back before, probably it was the same. Ya know, go to Babbage’s engine probably he was doing some of the same stuff. So I mean, it’s, ya know, it’s I had a boss that was twice my age when I was working on the word processor. He was doing demos. He pulled all nighters to do a demo for TV, when Edward R. Murrow was interviewing Jay Forrester in 1951, the year I was born. And he was telling me about war stories about how he had to do this demo about stuff to put some visuals up. Which is the same thing you’re doing for your demos today. It’s just that’s the craft, ya know if you’re, if Michelangelo is doing that, if you’re in that business you’re doing that too except we have better tools today. The big difference is how many people you can effect and how many different parts of the world you can be, I don’t have to just do business. I have medical students who are taking all of their notes on this thing, ya know. Of course when VisiCalc came out they were using it for all anesthesiology calculations for open heart surgery, which we never told our insurance people about [laughter].
So, that’s what’s fun about this business but it’s also why you have to be willing to retool, but you don’t always catch that wave.
I mean I’ve had a few wins every decade or two, okay? So take advantage of it and realize that when you’re up here it may not gone on forever and you have to decide when, part of the whole advantage of timing is to know when you’ve hit the top. If you have the right investors they may want to take that chance that you’ll go all the way up and you can do that, but most of them don’t. You have to decide your, what you want versus what your goal is. Do you like the journey, ya know there’s all this stuff there, but that’s separate. Whatever you’ve learned before can be useful in the future, but there’s always a new twist. I didn’t have to do animation. I didn’t have to worry, but now I have to worry about animation, all this stuff that I didn’t have to worry about back then. So, it’s the same, but it’s different. And it feels great that Joel asked me to be here. Thirty years later they still care. There I was in high school, I was in college.[applause] Think about it, I was in college and I threw that Hail Mary pass that was caught and they still remember years later, right, ya know? That’s kind of cool, but on the other hand I’m one of the top apps, productivity apps on the iPad, so, yes?
Audience: I have a history question. On that drawing your drew. They were connected. PC to mini to mainframe. Was that what, was connectivity between those three relevant, was it 100% relevant?
Dan Brickin: Not, not early on. We had connectivity, didn’t, connectivity is relatively new. I mean in the last 20 years for within an organization and the last 10 years within the world and the last few years among you while you’re walking , okay? Just think, Lotus Notes was pretty popular among businesses, so much so that IBM bought Lotus to get Notes, okay? But what happened with LotusNotes is it sort of forced all these corporations to put computers on everybody’s desk that were networked together and had servers that it replicated from, it was real cool. And then when Netscape came in they knew why it was worthwhile and they already had the wires in, so they just put Netscape in and the web could get a kick in business because of that, because they had already figured cause now they went from local to around the world. So, there have been changes like that. I just concentrated on the one thing about high bandwidth, the one thing, but connectivity is high bandwidth to other machines. [cough] To databases, the fact in the old days we had to dial up to Compuserve if you wanted this and AOL if all of you wanted that, ya know? Now, imagine was a mash-up with Googlemaps is like? Googlemaps, Googlemaps, knows ya know, will tell you red and green, how fast people are going on roads and it updates every minute it seems and it seems to really know because by you watching it, it knows how fast your car is going and you’re sending it data as are a lot of other people, ya know? Brilliant. Ya know, I mean they don’t have to have cameras everywhere. In the old days, but now we’re ubiquitous enough we can use brute force that we could not use before. We can use the social networks that we did not have before. So all sorts, it’s all interconnected that each time we add one of these hundred times better, they combine to the benefit of all the others and that’s an opportunity to build something, okay? And there will be new opportunities constantly. If you don’t make it on this one, you’re learning something that will help you on the next one. So I guess that’s it, thank you very much. [Applause]
Business of Software, Boston, MA, October 24-26th 2011 http://businessofsoftware.org For people growing sustainable, profitable, software businesses.