Home > Articles > Software Development & Management

An Interview with Watts Humphrey, Part 7: The Director of Programming Role at IBM

  • Print
  • + Share This
In this transcript of an oral history, Grady Booch interviews SEI Fellow Watts Humphrey. In part 7, Humphrey discusses his role as Director of Programming at IBM, including how he got a handle on the volume of mail he received (a problem even in the 1960s), standing up to his boss, and introducing the concept of planning to the entire programming division at IBM.
From the author of

IBM Director of Programming

Somewhere in the fall of '65, IBM was just starting to ship the Model 50, and the head of research, Gardner Tucker, got a hold of me one day and said, "Hey Watts I'd like you to write a white paper on how you think we ought to manage programming." So I said, "Oh okay." And so I put one together, and it was a tirade. I was really very concerned about the management of programming and how it was done. It wasn’t business-like at all, and we ought to have plans and go do this and that. We had plans for the TSS system, the timesharing system, but we hadn't done the real planning for the modifications, which was a failure on my part because of the market pressure. The market pressure was realistic but I could have pushed them off to make a better plan, and we did not do that. And that was a real failure on my part. But in any event, I wrote this tirade and gave it to Gardner Tucker, and a little bit later I got this call again from Learson. And I had been promoted to Director of Programming. It turned out they had fired the previous director.

Booch: Ominous sign.

Humphrey: Well, actually they fired the guy for a reason that had nothing to do with management. He was running around with his secretary. Nice guy, I knew him. He lived in Chappaqua, he had 10 kids, and he was running around with his secretary. IBM had something called an open door policy, and when somebody would write a letter to Tom Watson, anybody in the company could contact Tom Watson or the Chairman. I don't think they can still do it, but they certainly could then. And this young manufacturing manager had written a letter to Tom Watson saying, "My wife is the secretary working for this big executive who's running programming, and she's running around with him and I can't compete with your executives when they're running around with my wife. I'd like you to fix that."

So Tom Watson had an open door investigation and discovered that the guy was right on so they fired the Director of Programming, bam. And he left and took the secretary with him, divorced his wife and married his secretary. So he went off and was doing other stuff. I won't tell you who he was. But I remember going and meeting with the wife and that sort of thing. It was sad. But so I got the job and it was I'm sure based on the white paper I wrote.

Booch: Do you remember what you said in that paper? What kinds of suggestions, predictions?

Humphrey: Well, I would think I wrote about the crucial need for planning and for an orderly structured process to manage commitments and the whole nine yards. I did not have quality in the list at the time. We didn't have enough perception then to do that, but I did really push on the planning cycle and what you ought to do. So I was called in.  I'm a little fuzzy on the order in which these things happened. But I think the first one was a meeting with Vin Learson. I remember I arrived at my office in White Plains, the program director's office, and I had two secretaries and an assistant and they were spending all their time sorting my mail.

Booch: Physical mail of course back then.

Humphrey: Physical mail. They would make a daily mail summary. And the daily mail summary was a brief paragraph on every document that I got from wherever I got it, and it ran to about three or four pages. And I had a stack of about three feet of mail I got every day. And they told me the previous director had taken all this mail home every night and that's what he'd spend his time doing, going through the mail. I said, "Well, let's do the following. I don't want any mail unless it's from my boss, his boss all the way up to Tom Watson. Anybody in that group writes directly to me, I don't want copies. If anybody writes directly to me I want those letters right away, and any letters they write that copy me, I want summaries of that stuff but nothing else."

And so all of sudden I started to get no mail. So I had time to do stuff and I was able to deal with the problems. I mean, the previous director was paralyzed with mail. I didn't have time to screw around with that stuff and most of it was totally uninteresting. And it was a lot of people writing to me and saying, "What do I do next?" And so I said, "If you've really got problems like that give me a ring." And so I would get those calls occasionally, "What do I do about this or how do that?" I'd been in the job, like, days. "That's your job you figure out what you ought to do next and then come tell us if we need to know or tell your boss. I’m not going to tell you what to do."

The Learson Commitment Meeting

Humphrey: And so I basically got out of that loop right away. And Learson, I think this was at the very beginning, called me in. It may have been the second week, because I think the first day I did was go out and visit several labs the first week. And I remember I met in a couple of the labs, first thing I'd ask is, "What are your commitments?" And they didn't know. Honest to goodness they knew that they had to deliver a bunch of things, they had a list of stuff they had to deliver and they had dates and they were striving to do it. But they were all basically blue letter stuff, what was announced. And the marketing announcements were what was driving everything they were all doing. No one had any plans to do anything. And I said, I said, "Suppose you were going to do the job the right way. How would you do it?" And so they'd describe a very logical approach. How they'd do it if they had plans and requirements and they'd have schedules and they'd do it all. I said, "Well, why don't you do it that way?" They said, "We don't have time." I said, "That's insane. You're saying you don't have the time to do it right, you’ve got more time to do it wrong?"

And so they said, "Well, yes." I said, you know, "That makes no sense." And I left it at that, I had a couple of meetings like that at great big labs. I remember one in San Jose and a couple of others. And I went back thinking about what the heck to do. Well, along about that time I also got a call from Vin Learson to a meeting in my office in Poughkeepsie. I had an office in White Plains and another in Poughkeepsie. And Learson had called a meeting for 8:00 am in my office in Poughkeepsie, New York with me and all my senior management. And so I showed up. I showed up a few minutes late. He flew up by helicopter, the bum! So I drove there and I was stuck in traffic and I got there about 10 minutes late, and he was in the middle of a tirade. About 30 managers in this room sitting around the room and he was really ranting about how, "You're killing the company. You're going to destroy the business," and this sort of thing. And I walked in and was sort of sitting watching all this stuff and what had happened was the programming schedule had slipped about three times.

They were starting to deliver hardware. No one had any idea when the software would come and no one believed it. The marketing force was in an uproar. The customers were upset, and no one wanted to buy anything anymore. The market had kind of really frozen. What's interesting was we were selling the TSS Model 67, however, because people believed that. So, Learson finished his tirade and slammed his fist on the table. Learson, by the way, was six foot four, a great big guy, very powerful, one of the most intuitive, you know, really understood stuff. He was a brilliant man. But boy he really hammered on that table. And he then said, "God damn it, I've got to have a schedule in two weeks."

And so he looked around the room and everybody looked at me. I had had the job less than a week or maybe two weeks and I was supposed to make up a schedule. And so I looked at him and I said, "Vin," because I knew him pretty well at this point, "I said I can give you a schedule today if you want it but then I'll give you another one tomorrow." I said, "If you want a schedule we're going to meet, it will take 60 days." Then he started around the room one by one asking the managers, "What do you think?" And he went to every manager around the whole room and everyone of them said, "Yep, 60 days." And then he turned to me and he said, "Okay God damn it," and he slammed his fist on the table and he stormed out the door. I was the hero.

Booch: That's great. You were saying the things that no one had the guts to say.

Humphrey: No one had ever said it before. I knew Vin well enough to be able to do that because I knew it was crazy. I mean, some of these labs were in France and England and there was no way in the world I was going to be able to even speak to them all in two weeks. So getting schedules was a hell of a job. And so all of a sudden from being a bum I became a hero. And I'm convinced Learson did it on purpose.

It was a shoot out. I was the enemy. Now remember the TSS Model 67 was the enemy to the 360. So I was in charge of the timesharing group and that was the competitor to the 360. And so I was the enemy brought in to run 360 programming, and Learson had to make me a hero. And he knew how to do it. When I was Director of Systems and Applications Engineering, Learson used to ask me to meetings in his office. And I'd go listen to the meetings and I had no idea what was going on at the meeting or why I was even there.

The meeting would end and the guys would walk out and Learson would look at me and say, "Well what do they want?" And I'd say, "Well, I don't know, here's what they said they wanted." And he said, "That isn't what they want." He said, "Here's what they want." So he could see through all this stuff, just extraordinary. I remember once he was really beating us up because we were trying to get more systems engineering in the field, system engineers (I was Director of Systems Engineering as well as Programming). And I'd go in there with a story on why we needed more, and every time I did he'd have some facts and data -- I don't know where the hell he got it -- but he knew more about it than I did. And so he was always ahead of me. So I talked to my guys about this and so they were really searching around, and one of them one day they came to me and said, "Boss I got it. Here's a guy who used to work for Learson. He's in a basement office for a group in marketing, and he handles all the records and all the data and stuff, and he goes way back knows all this stuff. He's an old friend of Learson's." He said, "I think he's feeding Learson all this stuff."

Booch: He's your mole.

Humphrey: Yeah. I said, "Oh, okay. Let's go see him.” So we called up and went over and chatted with him. I told him our proposal and why and went through the whole thing with him and explained to him what we thought, and he bought it. He thought it made a lot of sense. I said, "Okay, thank you." So I went back and next time I went to Learson, I said, "By the way I chatted with so and so and here's what he said." Learson just smiled, and when we finished the discussion he bought it. So, he was real sharp. That's what he wanted to know, could I get there, and so we were able to do that. But he was an amazing guy.

So, I was Director of Programming and what I figured out was the problem the guys were having with not doing it right was that they didn't have any choice. And it wasn't really their problem, it was mine. The reason they weren't doing the job right was that when all they had to do to deliver programs they only had to do two things and what would those be? Code and test -- right?

Booch: Right.

Just Coding and Testing

Humphrey: That's all they were doing. And they had this enormous list of stuff they had to do. But there were only two items that were essential to ship product right now, and they had one problem which was to ship product. And as a consequence, if I let them ship product by just coding and testing, then that's all they had time to do. There wasn't time for anything else. So I realized that my problem was I didn't have an appropriate set of requirements on what they had to do to deliver products. So I went to Frank Cary, who was then Group Executive over all the development. And I said, "Frank, here's the problem we've got, and if we're going to fix it we have got to plan. And so I was going to stop everything. And to make sure we get plans for everything before we re-announce. I want to open all the schedules and tell the marketing organization we will get back with schedules in 60 days." And so he bought it.

And so I did. I sent out a wire to everybody to that effect. Now, the engineers could keep writing programs and testing all they wanted. But we weren't going to announce anything. We weren't going to ship anything. We weren't going to do anything, not going to start any programs. I wouldn't even fund anything unless I had a plan on my desk and they had to get back to me within 60 days, and they had to review it with me, all the important plans. I couldn't look at all the details. But every important plan they had to review with me personally, and I wanted to make sure it was signed off by everybody who had got to work on it. I didn't want them making plans for somebody else if they didn't agree.

And so I put that directive out and it made a hell of an impact, I'll tell you, worldwide. And so people started planning. And we had to get the manufacturing planning people -- I got agreement for them to go sit with the programming guys and show them what plans were. And they'd be consultants for them on planning. So the manufacturing community was extremely helpful in going in and showing these guys what a plan is. They didn’t know. And so we got damn good plans and I reviewed them all. They'd come in, and as I'd go through their review I'd say, "Look first of all I want a plan you're going to meet." Here all of a sudden I turned my hat completely around, because before we were pushing to get this thing going. But I told them, "Look, the prior director was fired." They didn't know why but I did. I said, "The prior director was fired and I don't want to be fired." I said, "If you miss a schedule you may be embarrassed, I could get canned." I'm not going to get canned, so I want schedules you're going to meet. And so they'd go through their plans and I'd poke at them. Get them to demonstrate that they could do it, that everybody agreed with this stuff. Have you got enough resources? What's your timing? I didn't cut a single schedule. I lengthened some of them. And I also put a 90 day cushion on every one.

Booch: Would it be fair to say that this was the most serious planning this group had ever really done? So this was novel to them.

Meeting the Plan

Humphrey: They had never made a plan. They didn't know what plans were. We put together a course. I got Al Pietrasonta to move up from the Federal Systems Division to put it together.  I'm not sure if it was a one week or a two week course on planning. And in the next several years we put 1,000 managers through planning training. We had people from marketing and other places, but we put everybody through it. It was a marvelous course. Al did a great job. He died early a number of years ago and I dedicated one of my books to him, I don't know if you knew him. But in any event, I talked to every class. I'd go talk on Friday afternoon. And so it was something. So these guys put these plans together and they met them. We didn’t miss a date for 2-1/2 years. Isn't that incredible?

Booch: That is amazing. How much of that 90-day cushion did you end up eating over time though?

Humphrey: Well let me come back to that. First of all, Frank Cary, when he agreed that I could stop everything and do this and open all the schedules and everything, he said, "You're going to have to go out and give a presentation on exactly what it is we're doing for the marketing Hundred Percent Clubs. There were four Hundred Percent Clubs. Those were all the top salesmen in the company. It'd be about 3 or 4,000 people in these club meetings. And so he said, "You're going to have to go give them a talk and tell them what we're doing." I said, "Okay." So I put together a one hour talk, and I walked through what we were doing. I said, "Here are the problems, here's exactly what we're doing." We had performance problems, schedule problems, all kinds of stuff. And so I laid that out, and then I laid out here's what we're doing about it. And so I went out and went to these Hundred Percent Club meetings, and told them,"One, we're opening all the schedules, we're going to have plans in 60 days, and here's what we're doing." So I went through, I described the performance problems and the measurement system we were putting in place for measuring performance and all the other stuff. It was a pretty damn good talk. I gave it four times, and I later had a number of marketing people, one was Mike Armstrong. You know who Mike Armstrong is?

Booch: The name is familiar to me, but I really know nothing about him. Tell me.

Humphrey: He was a neighbor when I lived in Darien [CT] and he was a top IBM executive. He left IBM in some reorganization. He didn't get moved up when somebody else did. So he left and became president of Hughes Aircraft for awhile, and then he moved from there to become CEO of AT&T, the old AT&T. And he was the guy who ran that, and actually they ran into some real problems, but nonetheless, he was a very good guy, marvelous boss. But he sort of got trapped by the technologists. But in any event, he told me when I met with him later at IBM he'd been one of the salesmen in the crowd, and he told me, "For the first time we had something we could sell." So my Hundred Percent Club talk basically turned the marketing team around. They said, "Now we got something we can believe." And so they went back, and they did turn the market around and it was that they believed me, there was confidence. And although I hadn't told them what the schedules were, we opened everything. The teams did come back and we got the plans within 60 days. The 90-day cushion I had put on, the teams all said, "No, we recognize your 90 days, but we'll meet our schedule and we'll stick with it." And the first couple of releases were 90 days early. And all of a sudden the whole marketing complexion changed, I mean all of a sudden these guys were delivering ahead of schedule, and what a difference it made. And so they could sell it. It also completely changed the dynamics of the debates I had with marketing, because people now believed me. I didn't have to fight anybody if I said it was going to take another year or six months or whatever. They didn't argue with me.

  • + Share This
  • 🔖 Save To Your Account