|Home > LEO Computers > LEOPEDIA > Oral & Narrative Histories > Michael Jackson: Inte ... 25th July 2018, 63002
Michael Jackson: Interview, 25th July 2018, 63002
|Home > LEO Computers > LEOPEDIA > Oral & Narrative Histories > Michael Jackson: Inte ... 25th July 2018, 63002
Mike Jackson and LEO Computers Society
Digital audio of a recorded interview with Mike Jackson, who worked as a programmer on LEO IIs and in sales for LEO IIIs before moving to Freemans Mail Order on LEO 326 and then on to an established career in computing.
Interviewer: Elisabetta Mori
Transcript editor: Mike Jackson (P1) and Frank Land (P2)
Abstract: After University joined LEO as trainee programmer on LEO II. Later lectured on a course. Project leader for Eagle Star pension fund on LEO II/1. Project Manager for British Oxygen on LEO II/7.Wrote specification for Perkins Engines service job on LEO II/5. Rose to Senior Consultant (sales) writing proposals for sale of LEO III to many customers. Oversaw installation of LEO III to some of them. Moved to Freemans Mail Order as Computer Manager on LEO 326. Became Administration Manager, helping to transform the company with groundbreaking applications. Moved to Perkins Engines as Group Director Management Services. In 1980 moved to a company providing bespoke applications using micro-computers. Outside interests included dinghy sailing to international standards and designing what became an Olympic class dinghy. Also enjoyed flying and owned his own small aircraft. Married Helen Clark a LEO programming manager.Date : 25th July 2019
Physical Description : 1 digital file, audio
LEO COMPUTERS LIMITED - Oral History Project Interview with Mike Jackson by Elisabetta Mori Edited: Frank Land [Elisabetta Mori]: It's the 25th of July, 2018. It is eleven in the morning and I am Elisabetta Mori. I'm interviewing Mike Jackson to give us the story of his involvement with LEO Computers from the earliest days. Good morning Mike. We are recording this interview as a part of the LEO Computer Society Oral History Project. The audio version and the transcript will be lodged at the centre archive and made available for researchers and members of the public. We are recording this interview with a Tascam, DR-07 MKII. I am in Walton-on-Thames at Mike’s house. Welcome Mike. So, can you tell us a little bit about when you were born, where, and what was the occupation of your father, your mother and your early memories? [Mike Jackson]: I was born in 1931 in Hounslow, which is west of London. My father was a senior manager in a major London department store. My mother was a teacher, who, before I was born, had started a private preparatory school. So I was brought up in a preparatory school from birth. My mother continued to run that school until she retired very much later. After that I went to Latymer Upper School, a direct grant Grammar School, where I did quite well, principally in physics and maths, although I was not good at subjects such as languages and history. I went on to Queen Mary College, London University, and got a 2.1 honours degree in maths in 1953. On leaving university, I didn't do National Service but I was engaged in the building trade for a few years. From 1955, and most interestedly, was employed by John Laing Construction, a major UK national construction company, in their Research and Development Centre. I worked in the structures laboratory. I did both theoretical work and practical testing of novel designs for steel and reinforced concrete structures. We had a laboratory where we could test new construction designs by loading them up and seeing when they broke, which was a very interesting time. At that time, there were occasionally mentions of the LEO and Lyons Computing in the newspapers. One particular article in The Evening Standard in early 1957 influenced me to think ‘this must be for me’, so I wrote to LEO on the basis of that and got an interview for a trainee programmer. In those days they assessed the programming aptitude by putting you on a day’s course called an ‘appreciation course’, where you were taught a minimal bit of programming and then you had to write little bits of programme, all in that one day. As a result they offered me a job as trainee programmer. I gave my notice to Laing, left at the end of April 1957, and joined LEO the following Monday. I was put in the Lyons Statistical Office until the June programming course. [Elisabetta Mori]: So, was it the first time you heard about computers? [Mike Jackson]: Through the fifties, in the papers and scientific magazines etc., there were articles and news items on computers. But I didn’t take a particular interest until, as I said, I read that newspaper article about Lyons around the start of 1957. [Elisabetta Mori]: So what happened when you joined Lyons? [Mike Jackson]: Well I did my programming course and was immediately given one of the programmes on a technical job for Stewarts & Lloyds, the steel makers. This job calculated stresses in industrial pipe work when they heated up, such pipework in power stations, big industrial plants. This took up my first couple of months. I must have impressed my bosses because not only did they get me to lecture on the next trainee programming course that September, but I was given project leadership on a new bureau service job for Eagle Star Insurance. This was the annual renewal for the company pension funds that Eagle Star Insurance administered. For example, I remember particularly Heinz, the tinned food people, had about seven thousand employees on their pension fund administered by Eagle Star. And every year each company’s pension scheme has to be revalued the scheme, taking out leavers, bring in the joiners, update increases in salary, etc. Then they calculate for each person the amount of paid up pension they've already bought, the surrender value of what they've paid. And overall to calculate the total amount the employer has to pay for his contributions to meet the liabilities of the pension funds. In a ll I think we had something around two hundred thousand people with all their records on a punched cards that were updated once a year for each company scheme at a time. That was a very successful job. They had an excellent senior actuary, Tony Ratcliffe, that I worked opposite who went on to be the general manager of Eagle Star. We got on very well and the whole job was put into operation in less than three months. It might have been the first job for an insurance company, in the world, dealing with individual people. Up to that time, particularly in America, there had certainly been a number of applications of computers to things like actuarial tables and those sort of calculations. But we were dealing with individual people and their personal contributions, etc., and the present value of their own pension fund. [Elisabetta Mori]: This was done on a LEO II? [Mike Jackson]: It was done on LEO II/1, the first one, at Elms House. (Editor: Part of the Lyons Cadby Hall complex of buildings) Each company had a pack of punched cards, one for each of the employee in the pension fund. [Elisabetta Mori]: This was a service LEO provided? [Mike Jackson]: Yes [Elisabetta Mori]: Do you remember any particular episodes? [Mike Jackson]: Not particularly. As compared with some others, for example, it seemed to go very smoothly. They nominated a company with about a hundred and twenty employees to do as a pilot scheme to make sure it all worked. Each job was complicated in the sense that every scheme you had was different in some of its factors or criteria. Each scheme had its own method of calculating the contribution by the staff, sometimes they're non-contributory schemes, sometimes they are contributory schemes. Men, of course, are treated different to women in pension schemes. And there were a lot of other different factors according to the details of each scheme, as agreed between Eagle Star and their customer company. But I never had any repercussions after it was put fully live. After that, in early ’58, I was appointed by LEO to be project manager for the implementation of LEO II/7 computer which had been ordered by British Oxygen Equipment Division in Edmonton, London. This was for a complete production and stock management system. This Division manufactured equipment for use with British Oxygen gases, both for industrial and medical purposes. I had not been involved with the preparation of the Leo proposals for this application, and British Oxygen had already placed the order for the computer when I was given the task of managing the implementation of the proposed applications. This was less than eight months after I joined LEO. We immediately started planning everything, of course. They built a special building to house the computer and as soon as it was available I moved in with a small team of programmers from LEO to prepare the required programmes. The first job was decided to be sales invoicing, and then to work back through production scheduling, production control, and materials scheduling. But, in fact, just as we were about to do parallel running of sales invoicing, British Oxygen management decided they didn't want to go ahead at all, cancelled the project and returned the machine. This was an absolutely shattering blow to me. What had happened was that the commercial director, who was very imaginative, had made the case to their board for this whole scheme. I remember particularly that the case included an estimate that they would save a hundred and two people in the company on the basis of the computer doing all this work, particularly in production control, stock control, and provisioning of materials and suchlike. But he, himself, somehow lost the confidence of the other directors, and was dismissed. For me with this massive job, given to me only eight months into LEO employment, as I said it was absolutely shattering. After all these years it is easier for me to see the scenario now than I understood at the time. Computers were a complete mystery to almost everybody in management. There were fears in management about being proven inadequate in the job they were in because they didn't understand how the computer could work for them. They were carrying their current responsibilities successfully, but this was pushing them into the unknown. It now seems to me that when they got the machine there, filling a very large room, and us just about to implement the first job, they just reacted ‘no, no, no, no, we don't want this’. I repeat that having been preparing the project for some months, this was absolutely shattering for me, especially including the fact that this very, I thought, far-seeing commercial director lost his job. After this, for much of the time that I was in LEO I worked on production control applications. I was asked to specify a service job for Perkins Engines, well-known diesel engine makers. At that time they were employing about nine thousand people in Peterborough, and they wanted a service job for what's called ‘the explosion of the build schedule’. They would have a build schedule for several months ahead for each type of engine, and they needed to know the parts required for each month. And there was a bit of complexity because the schedule would be changed every month, even possibly for the next month’s build. That is they would have a forward long-term one, but they would be changing it as they went through. And they would be modifying the engines so that for one reason or another they would change parts in the engines to be built, which would come in at a specified future month. So you had a different list of components for each engine type, which could be amended from some future month. There are a lot of parts in a diesel engine, and a lot of different engine types. They were using an IBM punched card system which was actually logically unsound. I specified the job to handle all this properly, which was prepared by the Leo Bureau Programming Department, and run as a monthly service bureau job on LEO II/5 at Hartree House for some years very successfully. (Editor: the LEO site for its programming and service bureau located on the one floor of the well known Whiteley’s Department Store in Queensway, London). From that time until I left, I was engaged, virtually the whole time, in preparing detailed proposals for applications for potential Leo customers. Of course, the cost of computers in those times meant only very big companies could afford them. I regarded it as an enormous privilege at that age that I was dealing directly at senior director level of very large companies. [Elisabetta Mori]: So you were not a programmer anymore? [Mike Jackson]: No, the last time that I actually wrote a programme for LEO was for the Eagle Star job. After that, apart from one minor job, I never personally wrote a programme for a LEO Computer, although I did manage the programming team for a number of customers. I was dealing with customers, mainly preparing detailed submissions for the sale of these. They are all large well known national companies, because only such companies could afford the hundreds of thousands of pounds that a computer cost in those days. [Elisabetta Mori]: This change of activity, like from programming to marketing, was it voluntary? [Mike Jackson]: Well, I wasn't marketing in that sense at all. What I was doing was designing computer systems. That is, I was sent to companies that were interested in buying a large computer who were saying, ‘this is what we think we should be using a computer for’. I would design how the computer could do that work. I'd design the contents of the files, the processing, the job requirements to be programmed and working out the specification of the computer installation needed so we could price our submission. I wasn't marketing, I was designing systems for potential customers. [Elisabetta Mori]: Can you name some of the companies you did this work with? [Mike Jackson]: Well, yes. I made a submission to Renold Chains for a computer, and that's a long story in itself, we’ll come back to. But also the Board of Trade, which was successful and I oversaw its installation. Stewarts & Lloyds, a big steel company, who had already done some work on LEO II, but this was for a major production application. John Summers was another big steel company up in the northwest. Singer Sewing Machines, which everybody knows of, Woolworths, Marks and Spencers, but the biggest, by far, was the Post Office. Now you must have come across, in other work you've done, of Leo people involved with the Post Office. I was given the job as LEO contract manager. They had asked about eleven different companies to submit proposals and I dealt with the preparation of the tender documents. At that time the LEO 326 was new, the LEO III came out around 1960 and all my submissions that I've talked about were all done for LEO III’s. But the 326 then came out which was massively more powerful, it was much, much faster. LEO was successful in its bid, and the Post Office initially ordered five of these. The original tender had been for two at £1.3 million at 1963 prices, it's probably over twenty times that in present money value. I was then put in charge of a team of about a dozen programmers to work on preparing initial applications with some Post Office people involved. But the one most noted was that we did get telephone billing working in about three months, with only four programmers on that job. And that was the forerunner for the whole sweeping of the country of LEO 326 telephone billing, but I was only involved up to testing the application at just one exchange, which was completed early in 1964. [Elisabetta Mori]: Have you got any memories of your colleagues working with you at LEO? [Mike Jackson]: On joining, I worked for Jim Smith for the initial Stewarts and Lloyds and Eagle Star applications. Thereafter, from British Oxygen onwards, I always worked for Frank Land, Chief Consultant at LEO. I had various people working for me who generally moved on in various ways. I mean I have other names that you've probably come across but passing through were John Aris and Ninian Eadie both of whom worked for me at some stage. Mike Finlay, you may not have heard of, Ray Hennessey, you might have heard of, also both worked for me. [Elisabetta Mori]: Do you remember David Caminer? Like have you got any episodes of... [Mike Jackson]: In the later stages I worked with David Caminer a lot. Particularly on these very large submissions for sales, and the Post Office in particular, because this really was by far the largest sale contract LEO ever had. It helped that I was very naturally what I'd call a ‘LEO person’. That is that as part of my initial programming course we were told about some Lyons applications, and the tea shops job impressed me enormously as the way a computer should be used to operate a business, not just do sums for it, not just do payrolls for it, not just serve an accountant. But how it's used at the centre of a company’s operations, and I kept that in mind all along. I always worked very naturally and easily with Frank Land and David Caminer, because we sort of shared exactly the same vision of a position of a computer in a company. [Elisabetta Mori]: Do you remember any female co-workers at the time? [Mike Jackson]: Well the most obvious one was my wife. [Elisabetta Mori]: How did you meet your wife? [Mike Jackson]: Well she was on the next programming course after mine; they were held every three months. After my June 1957 course, I was asked to lecture on one of the subjects which was the arithmetic and sequence changing instructions for programming a computer, and she joined LEO for that course. She had recently graduated in maths at Manchester University. [Elisabetta Mori]: What's her name? [Mike Jackson]: Helen. [Elisabetta Mori]: And what happened? [Mike Jackson]: Well, we got married eventually, but it was a difficult time because I was already in a very unsuccessful marriage when I joined LEO. I was in a broken marriage that had virtually ended before I joined LEO in 1957. [Elisabetta Mori]: And what year did you meet Helen? [Mike Jackson]: At the time she came on the course on which I was lecturing, but we started meeting up together in the autumn of 1958. [Elisabetta Mori]: Why did you, what did you like of her? [Mike Jackson]: Oh, well, I mean a personal thing to say. I mean, I think programmers naturally get on together because they think in the same disciplined way. She was a very able programmer, and went on to become manager of the Hartree House Bureau Programming Department. She came to be responsible for fifty three programmers by the time she left LEO in 1965, when we started our family. We first started meeting up in ’58, we got married in ’61 and she left in ’65 when we had our first child. [Elisabetta Mori]: And did she ever go back to work after? [Mike Jackson]: Not at LEO. She got a job, well this is another bit of the start of Compustar, we’ll come on to that a bit later. What happened at LEO, was that what I had done with the Post Office project, contract and all the rest of it was deemed a tremendous success. I'd been doing a lot of submissions for large companies, most of which did not result in a sale. But for the one or two that succeeded I came to realise that the people there weren't going to be using this computer as I personally knew, and people like Frank and David Caminer knew, should be used in a business. And I came to the conclusion that I couldn't get any satisfaction unless I personally achieved that, made it happen, in a business. [Elisabetta Mori]: How long for each company did you spend preparing a plan? [Mike Jackson]: Oh, anything from six weeks to at least six months for the Post Office because it was so large and complicated, and their own excellent staff were very penetrating in their examination of the proposals. Qverall I worked in this way 1959 to 1964. It was five years working on all these projects. One notable one that we did eventually sell was the proposal for Renold Chains. This was an engineering company making a wide variety of chains, almost with a monopoly. It made chains for almost all types of equipment that used them. From tiny chains for instruments, through bicycle chains to very large chains for enormous cranes. They were a nice company, I enjoyed working with them. But they, like British Oxygen, had a very sceptical management. They had a finance director whose vision it was and in that respect it was like British Oxygen. The way we dealt that with the scepticism was that after we had made the overall proposal, which they were very impressed with, they commissioned a pilot scheme. I recommended that we do a pilot scheme as a means which both enables one to get the job defined properly, and demonstrates to management that it would produce the results promised. The more imaginative and novel the job the more you need to learn in actually trying to do it, rather than a job which might be a simple invoicing job or a payroll where the customer can fully specify the requirements. For jobs the user can specify, including all service bureau jobs, we would define exactly what you're going to do at the start, get the customer to sign it off and do it, but the more imaginative it is the more you've got to learn about what's involved in actually doing it. So we prepared this pilot scheme for the production control department, and that was run as a service bureau job before they confirmed the order for the computer. I'd made the proposal to buy the computer, the finance director was very keen on it; he was the one who initiated it. We got on very well with the head of O&M there. But production management were very sceptical people, and the term for the pilot scheme was, ‘the proof of the pudding’, and that would be said almost every meeting we went up there. The production control manager would be saying, ‘but how do I know it's going to work?’ And we said, ‘that’s what the pilot scheme is for - the proof of the pudding...’ That job was in the course of preparation at the time I left LEO. I never heard anything more other than they did order a LEO III, until a few months ago, when I saw in Maurice Bonney’s history that Renold Chains, like British Oxygen, discarded the project even after it had been developed and demonstrated to meet the requirements. (Editor: Maurice Bonney was hired by Renold Chains as their chief programmer. His fascinating account can be found in the LEO dropbox archive). [Elisabetta Mori]: Why did you leave LEO? [Mike Jackson]: Ah, well, I've just been explaining that in general I came to realise that our customers weren't going to be using computers in the operation of their business in the way we at LEO knew they could, to achieve benefits not possible with existing methods. I realised that such results could not be obtained by a consultant working from outside the business, which was extremely frustrating. I was absolutely confident that if I was working within the business, with proper authority and resources, I could do this. It's got to be a multi-year project. [Elisabetta Mori]: So did when did you leave LEO? [Mike Jackson]: This was when we had got the Post Office order, when I'd been in charge of getting some initial programmes prepared. That basic phase was complete. So about March 1064 I went to David Caminer and said ‘I've recently been overseeing the installation of the LEO III at CAV, I'm convinced there's nobody there that's actually going to make it work. I really would like to join a customer and show what can be done properly with a computer. As David says in his article in the book, (Editor: LEO; The Incredible story of the World’s First Business Computer, see also chapter 23 for Jackson’s own account of his work at Freeman’s) just after the Post Office sale, Freemans, one of the very large agency mail order companies, had bought the most powerful computer purchased by a commercial company at that time, 1964. The LEO 326 was head and shoulders above its competition, with its processor 5 times faster than the standard LEOIII. David had just been asked by Freemans to release a consultant to join them to take responsibility for managing the implementation. As a result of my coincidental approach to him, David recommended me and I joined Freemans as Computer Manager. [Elisabetta Mori]: But it was mainly about your personal satisfaction in your job? It was not a matter of like something like a difference in salary or anything else? It was really because you did... [Mike Jackson]: Oh, it wasn't to do with payment/salary or conditions or anything like that. It was simply, I was filled with conviction of the proper position of a computer in a company, i.e., it's got to be central, it's not got to be department oriented. And, the responsibility must be carried within the business and not by an external consultant. I didn't see anybody in the companies that I had been installing computers in who had people capable of achieving this. I'd already been hit very hard, mentally, by the British Oxygen failure. [Elisabetta Mori]: But was Freemans in London? [Mike Jackson]: Yes, in South London. So it meant that I didn't have to move, it was not much different to commuting to LEO. [Elisabetta Mori]: Where did you live at the time? [Mike Jackson]: Oh, well I lived still in Hounslow at the time I joined LEO, and only moved away from there when Helen and I got married in 1961, but we remained in outer West London suburbia. I went to join Freemans as Computer Manager. The big difference between the problems of places like British Oxygen and Renold Chains was that Freemans was, until the year before I joined, a private company. It was run in a ‘hands on’ way by a few family directors, immensely wealthy owning the shares of a company that had at that time a near thirty million a year turnover and employed about four and a half thousand people. It had a very flat management structure, and the volumes of business were very high. In fact when, at the time I left, in 1970, we were processing five hundred thousand, half a million, documents a week. But at the time I joined they were only then starting to get senior management into the company. There were more than sixty people who thought that they reported direct to the Managing Director. I was brought in at a level senior to everybody below the directors. And this meant that I held the final responsibility for everything the computer did. I never asked for approval of any job specification. We never had any relationship problems with anyone at any level, as they were never expected to carry the responsibility for anything in the computer systems we were developing. There was, though, at senior levels extreme scepticism, quite reasonably so, that we could ever apply it to agency accounting, but I’ll come to that in a minute. The directors were justifying the very powerful computer on the modernisation of their despatch goods handling. Even at the beginning we could be getting thirty thousand orders a day in, all of which had to be invoiced with a name and address, generally, an average of only about three items on it but it could easily contain a lot more, or sometimes just one. It was then done on a whole bank of thirty plus punched card tabulators, and it was really state of the art of how it could be done without a computer. I don't know if you know addressograph plates of those days, which was a plate embossed with names and addresses, but you had to take an order form round to collect a punched card for one item then take it to another place to get a punched card for the next item, and so on until you had a punched card for each of the items on the order. Then take the collection to the tabulators to get the despatch note printed from the cards. Then the despatch notes had to be taken to the bank of 200,000 agents name and address plates to pick the correct one for each despatch note. They weren't the first big agency mail order business to use a computer for despatch note printing and stock records. They bought the computer believing, quite confidently and reasonably, that this was a job a computer could do if it's done properly, and they ordered the most powerful computer that they could reasonably get. It was planned from the outset to be sized to handle agency accounting but that was a subject they were never committed to and nobody in the company really thought that it could or should be done. So, we initially got on with the job of despatch note printing. Now there's an enormous difference between agency accounting in those days and mail order as we know it now. The catalogue was about two inches thick, it was far too expensive to send out to individual customers on spec. We recruited agents who would sell from the catalogue and have 10% commission of all their takings and be reimbursed for their expenses. So they were running little businesses, and at the time I joined we had two hundred thousand agents, when I left six years later we had over three hundred thousand of them. But our agents ordered goods to be sent on approval, each item of which they had to account for by notifying the sale to a customer or returning it. So much more complex than present day mail order where the customer actually makes the computer entry on his own PC at home, and pays for it by card at the time, and the sale is completed. We normally had about a million pounds worth of goods out on approval, which remained company property until sold or returned. However, a major additional factor was that the agent sold items on twenty weeks rolling credit. If say, an agent sold a customer something for ten pounds, they would pay ten shillings a week, fifty ‘p’ a week, for twenty weeks. But if that customer bought something else later it would be added to the balance outstanding at that time and the total would then be extended over a new twenty weeks, and the new weekly payment for that customer might actually be lower than it had been before the purchase. It goes almost without saying that business success depends absolutely on constant routine monitoring of every agent to check that approval stock was accounted for and cash collected properly. Appropriate reminders had to be sent, and if necessary, agents stopped from receiving further goods. Without that, debtors would soon spiral out of control and break the company. Also, every six months, each agent had to be assessed as to whether enough business was being done to justify sending her the new catalogue. All of this was considered to be essentially human judgement work. It was being done at a very large number of clerical work stations, at each of which one full time clerk and assistant would supervise about 500 agents, getting round the lot each month. In addition to the copy despatch notes, there were about 350 people opening the post each day extracting not just the new orders, but also the sales forms and cash slips. Also the returns notes came up from the warehouse. All this meant the hundreds of thousands of randomly arriving documents each week had to be sorted and routed to the appropriate work station in the correct office. The first, daunting, task was actually to get all this paperwork into the computer. This must be done to a very high level of accuracy. If you're processing, as we were by 1970, half a million documents a week it doesn't make much to go wrong for queries and complaints from agents to reach totally unmanageable levels. For example queries such as ‘where are my goods?’ Or, ‘you've sent the wrong goods’. Or, ‘I sent in a payslip and you're telling me I haven't’, all have to be investigated. You'd break the company if you didn't keep on top of a) accuracy in sending goods out, and b) accuracy of keeping account of what happens later. At that time, there was no capability for direct keyboard entry into the computer records, and data entry had to be via punched cards or paper tape. To keep error rates low, it was normal for entries to be first punched by one operator and then verified by a second whos e machine read the first punching. This more than doubled the punching workload as the second operator would have to stop and investigate each time there was a verification difference. With the very high data load we had to find a better method. We found this by putting a check digit into all agent numbers and stock item numbers, and using accounting machines to prepare the punched paper tape. This checked the check digits time of entry and enabled us to avoid the second verification. Just as important, the accounting machine was also capable of checking that agents’ payment slips totalled correctly. We researched this deeply, and it all worked out very well. You know we were using Olivetti counting machines. [Elisabetta Mori]: Yes. But you, can you remember the name the Olivetti model? [Mike Jackson]: No, I can't remember that. For each batch of documents, which we batched and numbered and counted, the number of documents in a batch, batch totals. These numbers were also checked by the Olivetti machines in addition to the checks the machines did on agent and Item number check digits and cash totals on payment slips. It was a very big order, of course even for Olivetti. The initial order just for the agents’ goods orders was for ninety machines. We were dealing with very, very large numbers. As I have said, we started with handing agents’ orders and printing despatch notes. We started off very gently, to learn the lessons needed to handle all this stuff, from training entry operators to the quirks in the details of merchandise items, while handling low volumes. As well as the large transaction volumes, there were about fifty thousand different items in each catalogue, e.g. every shoe size is a different item. And because a new catalogue is issued every 6 months, at any one time you've still got items in stock from the previous catalogue, and which also may be coming back from agents’ approval stock. In addition, items for the next catalogue have stock coming in. We actually fished up with stock records for about a hundred and fifty thousand items. To start with though we chose a simple merchandise group with about three or four hundred items, and only entered the orders and printed the despatch notes for those items. Obviously this also enabled us to take on by merchandise group all the details of the record needed for each item. These details had to include the description, selling and cost price, etc., and the quantity of stock available, and update that continuously. We also kept it as simple as possible by not, as yet, dealing with agents’ names and addresses. The computer printed despatch notes for that little initial stock group to be sent to the addressograph machines to put on the agent’s name and address. So we implemented the despatch note printing one merchandise group at a time, sometimes needing small programme modification in order to cope with a particular type of merchandise. So we built up quite quickly, and we got about twenty percent on before end of November ’65. We stopped there, because pre-Christmas volumes were terrific and the whole place was up to its eyes in this. But immediately after Christmas we got going again and got the lot on in by about a couple of months after that. We straight away followed this by taking on the basic agents’ details such as name and address, and indicators to show whether an agent was stopped from receiving goods and whether they were to receive the next catalogue. With this we could now print the complete despatch notes to go straight to the warehouse, eliminating the addressograph plate process. I said earlier that I would occasionally be reminded by a director that they were not committed to ‘doing’ agency accounting on the computer, and I remember once responding ‘to me, it’s not a question of whether we do it, but what we should be doing for it’. The next step we took was pivotal in this. We now had a basic record for each agent, and details of all items of stock sent to them on approval. I thought ‘why not post the sales and returns forms to the system and thereby keep records for each agent of outstanding goods on approval, to enable us to generate the first stage reminder notices to be sent for overdue items?’ For the computer system this was a relatively small development step. So in accordance with our small step approach to development, we got permission to take two agency sections, i.e. about 1000 agents, and post the sales and returns for just them to the computer. We were extremely fortunate to have allocated to us a senior agency office manager, Mrs Thompson, who was just the right person to oversee this work. It was not possible with computers in the 1960s to have direct access to computer records, and we gave the clerks a simple enquiry slip which enabled them to have printed on the next run of the computer file details of approval stock movements and items outstanding for any agent. When this new job had settled, we then withheld the copy despatch notes and the sales and returns forms, and made them administer agents’ approval stock entirely from the computer records. This meant that the only agent completed forms to be filed were the payment slips, about 30% of the documents. This test group was soon deemed successful, and we got permission to extend it to about 10,000 agents. The immediate effect was unexpected. This was large enough number for it to become obvious to many at all levels in the agency offices the attraction of not having to distribute and file some hundreds of thousands of documents each week. There was a clamour for us to extend the job to the whole company. There was of course a lead time for this, as it more than doubled the volume of agency documents to be handled by the computer, needing a large number of extra Olivetti accounting machines and trained staff to use them. This was the time in the whole project when we most nearly overreached ourselves, at peak we had a few days of unposted sales forms and some worried faces in management, but we did pull through. There were three overall consequences of this. Firstly, for the first time they had decent management information on the content and age of approval stock out there. Secondly, whilst I had been given line responsibility for post opening and order processing when we started the despatch note printing, some 700 people, I was now given responsibility for the whole agency accounting office, which meant that I had line responsibility for a total of about 2800 staff. The new Agency Administration Manager, recruited soon after me, had been found unsatisfactory and dismissed, although I never knew what the reasons were. Fortunately, Peter Smith, who I had engaged as Programming Manager soon after I joined in 1964, was excellent as my right hand man to manage this new empire. Thirdly, at senior levels, they began to think that we might be able to do something useful for agency cash accounting We had well over a thousand clerks, monitoring each individual agent, determining whether they needed being sent a reminder, whether the debt or approval stock was so bad you wanted to stop sending anything to them, or they're not doing enough to get another catalogue. All these are regarded as human type decisions, and this is why there was such scepticism about a computer doing it. From the beginning, agency accounting was in the plan, despite the reservations. Behind the scenes I had been designing with Peter Smith a computer job for it. We now had the details of the sales to the individual customers of each agent, and posting the cash received from each would be a straightforward job. This meant records for about 1.5 million customers at that time. The problem was to be able to check whether the agent was sending us proper amounts, and print when appropriate chasing letters or even stop further despatches of goods. I have told you about the sales being by rolling credit, which meant that the cash to be collected by an agent each week depended on the outstanding balance of the particular customer she sold an item to. This in turn meant that we couldn’t assess an agents payments simply using agency total sales, and had to work at customer level. We used a technique known as ‘exponential smoothing’ (Editor: a statistical forecasting system) to determine a ‘payment factor’ for each agent, being the smoothed average of payments received as a percentage of the correct amount. When we had specified it, I asked Helen, my wife, to write this agency accounts programme, whilst she was at home looking after our first child, in 1966. [Elisabetta Mori]: She was working with you at Freemans.... [Mike Jackson]: But only from home, she only visited the place about two or three times. But she actually wrote the agents’ accounting programme, which was the largest and most complex single programme we had. This was because it had to do virtually everything required in a single pass of the very large magnetic tape file. [Elisabetta Mori]: Was it something common for Freemans to have someone working from home? [Mike Jackson]: From home? I don’t think so. It was done so as to keep this background exploratory work away from the pressures on the systems/programming team whilst implementing the large volume work I’ve been describing. We took one section, 500 agents, to test on, with over 3,000 customers. Because agents on appointment were distributed randomly to office sections, 500 was a good statistical sample with which to discover all the snags, but absolutely negligible among the 200,000 plus for possibly damaging the business. We continued to have the excellent Mrs Thompson to work on this next stage. Whilst taking on the approval stock records for each agent required a more than doubling of the amount of agents’ paperwork we processed, it was a relatively simple logical step, albeit with very large benefits to cost of company operations and increase of valuable management information. Taking in and monitoring agents cash collections was altogether more complex, requiring 3 times as much information to be held on the computer agents record, and much more likely to cause significant disruption to the business. We used this small test number to tailor the computer rules for triggering reminder letters and then notifying the agency office clerk when an agent cash payment reached a level of arrears requiring action. When Peter Smith, Mrs Thompson and myself were satisfied we were ready, we took on a batch of 5,000 agents, to further test the system with a s olid volume, and used this to design the new procedures for the office clerks and the new pack of stationary to be sent out to agents. Also we developed considerably enhanced management information on agents performance compared with the very limited amount that could be got when all the details for agents was scattered in individual manual files. The task of taking on this work was much greater that for approval stock records, and we worked relatively gently, taking over a year to transfer all the agent s, completing it in 1969. Since I joined in 1964 the business had steadily expanded, the number of agents increasing from about 200,000 to over 300,000 by the time of completion of the agents cash accounting work, with corresponding increase in the volume of goods despatches. Whilst this was going on, in 1968 I was appointed to be a director of the company, together with a new Accounts Manager brought in a few years earlier. Both of us were still in our 30s, and the first non-family directors the company had had. One thing worth mentioning was that the director’s dining room was a vast dark oak panelled room, with a baronial dining table. With their new young directors, they decided to get a table tennis table in it, which was used every lunch time by new and old directors alike. We had now completed all the tasks considered in the original proposal made in 1963 for the LEO 326, whilst absorbing a 50% increase in business. The benefits were huge. The number of staff in administration and order processing had been halved in relation to volume handled; one clerk could handle 2000 agents as compared with two handling 500 agents. The impartial, systematic, continuous review of agents stock on approval and cash outstanding, taking action when appropriate, resulted in a reduction of approval stock and agents debts which more than paid for the whole computer and data prep installation and the software development. Noteworthy, is the fact that the volume of queries and complaints from agents dropped also, reflecting greater accuracy and timeliness in handling these large volumes. However, in addition to this work proposed when ordering the computer, we took on a major extension. Soon after I joined, they took on a warehouse manager directly reporting to the board, who in addition to managing the despatch operation, had the task of preparing for a proposed new modern warehouse with mechanised handling of the goods. The proposal came to be firmed up for a new warehouse in Peterborough, over 350 yards across its diagonal, with direct conveyor to the Post Office centre just across the railway line. With it came to be the need for the computer to analyse the orders and schedule the picking and packing of the items required for each order. It was necessary for all the items in one order to be picked from stock near simultaneously from anywhere in the vast warehouse and routed mechanically to the same packer to go into the same parcel. There were 75 packing stations, with the despatch notes sorted to be in packing station order and distributed to the appropriate packing stations. Goods to be picked for stock had picking slips printed in by stock location in timed batches so that goods arrived at the packing station at the correct time to be put together in the parcel. The required computer system was a necessary part of the integrated system for the warehouse to work at all. When all a day’s orders had been processed and sorted to print the despatch notes by packing station, and the items required sorted by warehouse location, we would print for the warehouse management the analysed volume of work for despatch the next day, and notify them of the number of packing stations required to be manned. We even sorted the work to meet the schedule of Post Office mail trains to the various parts of the country. All this additional work was done with the same computer ordered in 1964, completed in 1969. Is this a world record? The LEO 326 really was a magnificently powerful machine. IBM proposed a 7010, about as powerful as a LEO III, which couldn’t have looked at it. So, by 1970, as compared with 1964 when the computer was ordered, we were handling 50% more business, amounting to 500,000 documents a week, from over 300,000 agents with 2 million customer records. On pro-rata basis, we had half the office and order processing staff, and half the warehouse staff due to the new mechanised warehouse, which also provided agents with a faster and more reliable delivery service. There was a significant reduction in queries and complaints from agents. The systematic, frequent, monitoring of agents for unsold approval stock, late payments and bad debt reduced the approval stock plus agents debt on a like for like basis by about £800,000 (about £20million in today’s money). In the subsequent decade, the ‘70s, Freemans was the fastest expanding large mail order business in the country. We never had a computer hardware or systems fault which had a noticeable effect on company operations. No other major mail order company had achieved all this at that time. One such, Grattans, with whom we were friendly at director level, were already using a computer for order processing when I joined Freemans. I visited their computer department early in 1970, and learned that they had no plans for agency administration, which had delivered most of the major benefits to Freemans. But not only that, they were employing 60 systems and programming staff, whereas we never had more than 15. As compared with a significant number computer installations, which, even if not failures like at British Oxygen Equipment, failed to come close to meet their original objectives, I believe the success at Freemans was due to a combination of unusual circumstances. In most companies, especially at that time, a computer manager would be junior to top departmental managers, and expected to provide a service to them. But, as I have said, I was brought in senior to all below the directors. This meant that I alone was responsible for everything the computer was doing for the business. I never asked or received from any in management the approval for the systems we designed and implemented. I was never queried on requests for money or staff resources. We studied existing methods very carefully, and for each stage of development I decided on the details of what the computer should do for the business, and ‘carried the buck’. There was no way I could meaningfully pass this responsibility to the directors above me. They just had to, and did, accept what I did. But we were very careful. Because of the novelty of much of the work, each step in implementation was started with very low test volume, under close scrutiny of my systems team, and computer programmes were tailored to smooth the wrinkles. We couldn’t afford to, nor did we, put a step wrong. It may also be noteworthy that neither I nor anyone else in the computer department ever had significant disagreements or any relationship problems with anyone at any level in the company. [Elisabetta Mori]: So why did you leave Freemans? [Mike Jackson]: Yes, now, I, why did I leave Freemans? I, well we were at an end of an era but in more than one way. Before I had joined the board, there were three executive directors, one dealt with catalogue and marketing, one dealt with warehouse and goods buying and one dealt with all agents’ accounting and order handling, and who had been my principal contact while I was dealing with the computer. Now they were all extremely wealthy having had substantial proportions of shares of a family owned private company that had just gone public for tens of millions of pounds. The Manging Director really felt it was time for him to retire. We were at that stage in 1970 when we had effectively reached the level that was possible with equipment available at that time, including the mechanised warehouse. We had really completed modernisation of the business. He decided he wanted to retire and their choice to be the new managing director was the new finance director, who had been appointed director at the same time as I was. But I just didn't respect him well enough. We had always worked very well together, and he always went along with everything we did, but I just didn't feel he was somebody I wanted to work for. This was the end of an era and the start of a new one, we weren't talking about another two years or so. The whole company was working properly, I'd got an excellent first line team that could easily carry it on without me, and I just didn't feel I wanted to embark on a new period with somebody that I thought was too limited for me to want to work for. So I left. [Elisabetta Mori]: And you joined.... [Mike Jackson]: What happened then was that Frank Land, with whom I had remained in contact, suggested me to Paul Dixon, of whom you may have heard. (Editor: Paul Dixon’s Oral History is held in the LEO dropbox archive) Back in 1957, I had taken leadership of the Eagle Star project in its earliest stage from Paul, then a LEO senior programmer, who was leaving to take a job with Massey Ferguson in Canada, farm equipment makers. Frank had remained in contact with him all those years. Now when I left Freemans it so happened that Paul Dixon was by then corporate head of world-wide computing for Massey Ferguson. Massey Ferguson had bought Perkins Engines, (Editor: a major manufacturer of diesel engines situated in Peterborough) and Paul had mentioned to Frank that they really wanted somebody to come into Perkins to get a grip of their computer systems. Now I had already done work for Perkins in the service job for production scheduling, and had already made a submission for the sale of a LEO III, way back in my early days, and I knew a few people there in production control. They regarded LEO well, particularly as we had already done a successful job for them. But Perkins had been told by Massey Ferguson you can have what computer you like as long as it's IBM and so they hadn't bought the LEO. I was taken on as Group Director Management Services. In addition to responsibility for the computer department at Peterborough, I had staff responsibility for the computer departments in Perkins engine manufacturing plants in Detroit, Säo Paulo and Paris. [Elisabetta Mori]: Which computer did you use? [Mike Jackson]: IBM, they already had two. I walked in, I went over it and said, ‘you haven't got enough work at the moment for two thirds of one, so we threw one out. It so happened IBM’s policy at that time was ‘well nobody who buys a computer’s is ever going to be able to throw it out, there's going to be too much needed’, etc. etc. So in fact they were on rental with no notice period. So there was no cost in disposing of it, well other than perhaps removal costs, so I threw one out. But, as compared with Freemans, Perkins had all the characteristics of a manufacturing engineering company that had caused the problems at British Oxygen and at Renold Chains. Management with no understanding of computers, and really not wanting to know. Unless, of course, they have in mind some job for their own department which the computer might be able to do usefully for them. In any case the whole IBM philosophy that had infiltrated was that a computer should be a sort of service bureau to the departments; until then I'd never heard the term ‘user department’. I'd always thought the computer must be central, the means by which the operations of the company were coordinated across the departments with an effectiveness not possible with previous methods, providing large benefits to the business as a whole. This had been originally planned for British Oxygen and Renold Chains. When I joined the computer department was held in low regard by management generally. I got down to the initial necessary tasks to get some proper standards into the jobs currently running, into the preparation of new work in hand, and into the disciplines of the computer operating department. One job I had to do was to completely rewrite the parts scheduling job which had been running on the LEO service bureau, and transferred to their own IBM machine when it became available. However, when I arrived I found that they had used the same unsound logic of the original punched card system. This was but one reason for the low esteem of the computer department. I suppose that had I stuck at it, in time I might have worked my way round bit by bit to getting the computer into its proper place in the operation of the business, but the other side of it was the whole North American corporate governance involvement and culture of Massey Ferguson. It got off to a bad start because Paul Dixon had arranged that we combine our computer hardware facilities with the Massey Ferguson tractor and farm machinery factory in Coventry. I had been told this when offered the job. I agreed to this in principle. Even by 1970 there were good line communications available, and it didn't really matter where the computer was. I did very genuinely start trying to make it work by selecting an initial job to be transferred and discussing with them the necessary arrangements for this. I was constructive in suggesting details for this. However, as soon as we started this first trial it became apparent that it could never work. Their computer operating disciplines were unacceptably slack, and there was no reliability of job schedules being run to time. There seemed to be problems with Massey jobs which would always take priority over Perkins work. Overall, their whole way and approach to using computers had little of the rigour instilled at the outset at LEO. It led to the most difficult relationship I ever had in the whole of my working life with a colleague. Paul Dixon had sold his hardware centralisation plan at Massey Ferguson’s Toronto. He was group director and all the rest of it and it just couldn’t be made to work in the way he had planned. So I got down to the task of bringing the Perkins computer department up to scratch. I did at least get the computer department held in good esteem in the company. But it just wasn't the place for me, in that sort of international corporate scene where top line managers are looking over shoulders as to how their own departments is performing rather than looking to see how total company performance can be improved. It was the opposite of the business environment that I wanted to work in. When I had been there nearly 3 years I went to the managing director for a discussion. I reminded him how very much better the computer was serving the business with the department costing much less than before. I told him that I felt I was waiting in the wings to be playing a much more central role in the company. However, it really came to a head soon after that when we had a three weeks strike by the factory workers. Never having anywhere near such a situation before, it was quite interesting to experience. The pickets physically barred all entrance gates, and even as a director they wouldn't allow you in unless you identified yourself. [Elisabetta Mori]: Which year was the strike? [Mike Jackson]: This was 1973. In the ‘70s strikes were rife in manufacturing businesses, but I had never been anywhere near one before. The positions being taken by both sides didn’t seem sensible to me. In my naivety, after about 2 weeks into the strike I said a few injudicious words about how I thought the company was handling it. After the strike was over, the managing director asked me in and said ‘you're not really for us are you?’ And I wasn't, there's no way I could work in a North American corporate environment, I just wasn't that sort of person. I couldn't play the game, I didn't want to play the game. [Elisabetta Mori]: So what happened? [Mike Jackson]: Well I left, and I didn't have anything to do. [Elisabetta Mori]: And what did you do afterwards? [Mike Jackson]: I started looking for another senior role in computing, but in my heart I was disenchanted with the prospect of working within the management structure of a large company after experience of British Oxygen and now Perkins Engines. At that time, 1974, it still needed a large business to afford a decently powerful computer, and in general senior management was still ignorant, sceptical, and very wary of using them. Helen and I had had very great difficulty in finding a house we liked in the Peterborough area, and I had been with Perkins well over a year before we found one. Already I was thinking that Perkins might not be a long term prospect for me, so when we moved, we let our Surrey house rather than sold it; one of the wisest things I ever did. We had bought a house near Peterborough with 1½ acres of land at what seemed a very low price compared with the London area, sold half of the land to a developer for new house building. The great inflation period around 1972/3 meant that the house with remaining land fetched nearly three times what we had paid, so that when we moved back to our Surrey house we had quite a decent amount of capital. I decided to give up computing, and in 1974 I put that money into a small manufacturing business far too small to use a decently powerful computer in. Originally we were buying our components from Italy and selling in our product in England. In accordance with being a loyal British man, we changed our component suppliers to this country and developed an export market to a number of countries in the Middle East, which was good, taking about half of our production. We were having to sell in dollars to the Middle East, at generally around $1.60 to the pound, and we had a good business. But what happened was that in the early Thatcher period of 1979/80 the pound went through the roof. We could just about live with $1.80 to the pound, but at two dollars we didn't have a business; we couldn't sell at a price which covered even the component cost. The pound actually went up to $2.40. So we wound the business up, and I was out on my ear again, end of that venture. [Elisabetta Mori]: And what happened about.... [Mike Jackson]: Well, I thought, what do I do? Our children were all at grammar school, and Helen had a part-time bookkeeping job with a very small local building and civil engineering company, John Hirn Construction. There was only one manager, the business proprietor. He was designing a new form of structural beam for building purposes, and that took me back to my John Laing Development Structural Engineering in a way. He had been getting Helen to go down to Atkins Computing, a service bureau in Epsom, to write programmes for structural calculations, because he knew that she had worked as a programmer for LEO. After a while he asked Helen ‘Can you write a production planning system for my new Hirn Beam for all the contracts I'm getting for it, using this Apple 2e computer I’ve just bought. In spite of the fact his was a small firm one, he was selling his construction technique for some sizeable contracts. One of these would be for a major extension to Cardiff Hospital. Helen said, ‘oh, I don't know that I'm good at this but I know someone who is’, and so I was asked to do it. This is 1980, I'd just left Perkins, his computer was about the size of a small portable typewriter, and I thought ‘yes, I’ll have a go at that’. Immediately on beginning to get to grips with it, there were two things I could hardly believe. One was that this little thing was certainly more powerful than a LEO II, even in those very first days of personal computers. But the other thing was what a superb language Basic was for a trained programmer who had only written in machine code. It just did everything you wanted. So I wrote this programme and it all worked to the satisfaction of John Hirn. Nevertheless, I was still basically out of work. I’ll be talking more later about my involvement in sailing, but it will suffice for now to record that by this time I was quite heavily involved in committees of both the Royal Yachting Association and the International Yacht Racing Union. I didn’t make a secret of the fact I was out of work, you don't want to try and pretend like some people do. I had reached the stage of having been on a number of committees and chairman of one for a long time. I knew a lot of people in business, some of them really quite wealthy. One of those was chairman of the RYA Yacht Racing Divisional Committee, of which I was a member. He was a partner in Tarbutts, a small investment house where with two other partners managed their own and extended families share portfolios. He invited me to look at their accounting operation and see if I could do something for them. The partners spent most of their time dealing on the Stock Exchange, just the three of them, but with a lot of money around. Their only employees were one secretary between them and three men, all who smoked pipes, in a back office keeping the accounts of all the deals they did on behalf of the clients. Obviously you've got to be very strict in keeping proper records for each client of all the stocks they own, all the cash they've paid in, and distribute the dividends, etc., etc. Investment management is much more regulated now than it was then but even so client accounting was something that really had to be done to very high standards when you're dealing with people’s fortunes. So I went in and studied their bookkeeping. They were keeping, as they had to, a proper set of manual double entry books. They told me that they balanced the books, that is prepared a trial balance, every three months, but it could take them up to six weeks to get it balanced, to eradicate errors in the manual entries. Now it so happened that early in my time at Freemans I had been asked to go on a course of evening classes entitled ‘Accounting for Non-Financial Mangers’. This proved to be by far the most valuable piece of postgraduate training I ever had. As a brash maths graduate I had thought that the very idea of ‘double entry’ was old fashioned mumbo jumbo that computers would eliminate. The course instructor told us at the start that double entry originated in the 14th century by a monk trying to keep the accounts for his monastery, and the principles were unchanged to this day. How wrong I had been! I was almost overwhelmed on discovering its elegance and power. What this learning gave me was the respect from accountants when I could immediately understanding their accounting needs. Some years later the Finance Director of quite a large company told me that I was the only computer person he had met that properly understood accounting. So, I bought a small personal computer for Tarbutts, and wrote a bookkeeping system, tailored to their particular operation. This was right at the start of using PCs in business; at that time there was virtually no decent package software for them. Only word processing packages and some very poor accounting packages. I had been trained as a programmer on LEO II, with the disciplines developed by LEO that had enabled the machines of the 1950s to do large commercial jobs. It was this that enabled me to do so much with early PCs. In a couple of weeks they had the basic system installed. The chief clerk, with whom I had agreed the details of the job, stopped all manual bookkeeping within a month of starting the new system, he was so confident. Obviously they had to post the stocks and shares bought and sold, and the cash movements. But they now had a trial balance printed at the end of each day, and 2 or 3 days after the end of each quarter they had a full set of company and individual client accounts, all done. Then he said to me, ‘can you print contract notes?’ So, I thought, ‘well, why not?’ They're sending us details of contracts for the sale or purchase of shares, so I said, ‘yes, why not?’ This was the first small step of developments which led in time to a comprehensive fund management system which by the time I retired in 1995 had £2 billion of client funds under management. All this expansion of volumes came following Tarbutt merging with Jupiter Fund Managers which was also quite small at the start, and quite rapidly expanded. And they kept my systems at its heart, which I had steadily enhanced, for example to provide valuation on demand of any client’s portfolio, by downl oading each morning the previous day Stock Exchange closing prices for all securities held within the system. By this time the power of PCs was steadily increasing as were good networking facilities. Using networked PCs rather than a single larger computer meant that each time an additional terminal was required it brought with it a further processor which did the processing for its user, so computing power available increased with the increasing number of office staff required for the increasing volume s. Also, of course, each year the power of each new individual the PC was greater than earlier ones. The biggest increase in volume came when Jupiter took over another finance house which was four times the size in value of funds under management, but less profitable, at least partly due to the much lower cost pro rata of the back office using my system. Because by that time my fund management system was so comprehensive that not one programme modification was needed to tale on all this extra work from a different company. One matter worth recounting followed from the ‘due diligence’ research that companies have to do when making a takeover bid. For this purpose, Jupiter ask me to study their computer operation which was costing them about £4M pa. This consisted of a large computer and systems team at their head office, and computer costs for a small bank which came with the business being taken over, which had a conventional ICL computer at each of its three branches, each in its own air conditioned room. After a detailed study of their systems and costs, I told Jupiter that I could save them £2M pa, by eliminating their head office computer department and replacing the bank ICL computers with networked PCs. Their principle business of fund management would be brought onto my more efficient and comprehensive systems at Jupiter, since we only needed some more PC terminals and an enhanced server for this. For the banks, the capital cost for the same number of terminals, with a second standby server at each location plus an additional one at head office, was significantly less than the annual maintenance cost of the ICLs. I could only do this as the banking software they were using was also available for Novell Netware which we were already using for our Jupiter networking. Not surprisingly, they were sceptical of my claim of £2M pa saving, but I had been supplying their computer systems for several years by then and earned a lot of confidence. We did a deal whereby I would be paid 20% of up to £500,000 first year savings, and 35% of savings above that. In fact, I achieved about £2M pa audited savings and got paid about £600,000, most of which I put into my then tiny personal pension pot. When I did the first job in 1980 for Tarbutt I knew that I had to work as a limited company, and formed Compustar Ltd. Whilst I was involved with Tarbutt/Jupiter throughout my time until retirement in 1995, Iprovided systems for several other firms, all from sailing contacts or recommendation by them. These included Piers Godfrey, another fund manager, Robert Dyas, ironmongers, Marina Developments with a chain of six marinas, Remane Bros, jewellery wholesalers, Marley Floors, Hunters & Frankau, cigar importers. I would note that some sites were some way out of London, particularly the marinas from Ipswich to Poole. While at Freemans I had learned to fly as a private pilot, and to service all my widespread client locations, in 1982 I bought a small helicopter as I could land directly at most of the sites. This was a real business value as otherwise a whole day would be needed for a one hour visit. Every line of code in all the systems provided by Compustar had been hand written by me. The big thing was that I was an adept programmer who could talk business objectives with directors. Never once did I write a specification of a job for the customer to approve. It was for me so much simpler and quicker to agree objectives, programme the basic job, take in a PC and demonstrate what they would get. And tailor it where necessary for acceptance. With Basic as a language, this was quicker than writing a specification, and much easier for the customer to understand. And know how it would actually work. [Elisabetta Mori]: When did you start sailing? [Mike Jackson]: Oh, 1942, in war time. I was only eleven years old. [Elisabetta Mori]: How did you end up sailing? Because you liked it? Because your parents did? [Mike Jackson]: My father bought a small sailing dinghy. One couldn't get to the coast in the war time, but on the Thames in west London, river, people were allowed to keep their own boats, though they had to be registered with a number on them [Elisabetta Mori]: You went with your father? Have you got any brothers or sisters? [Mike Jackson]: I was youngest of five children. When we first had the boat they tended to come but I was the only one that carried on with it. I was the youngest by five years from the next oldest. I went into racing about 1947 /48 being introduced by one of my school teachers to his sailing club. My parents bought me small twelve foot racing dinghy, and I was quite successful at club level. Sold it when I went to university, where I sailed with the University of London Sailing Team; I was a regular member of that. [Elisabetta Mori]: Which kind of sailing boats? [Mike Jackson]: That was a National Firefly, it was a dinghy that was used in the 1948 London Olympic Games when the sailing events were held in Torquay. From the start I had been interested in sailing boat designs. I came to think that I'm sure I can design a better racing dinghy and so I started designing boats, building my own. The National Twelve is a development class, it is only governed by a set of measurements, you can have what shape of boat you like within those measurements. [Elisabetta Mori]: What is this photograph? [Mike Jackson]: That is Helen and I winning the Principal Championship Race from two hundred and twenty starters. [Elisabetta Mori]: So it was with your wife? [Mike Jackson]: In 1964, in a boat of my own design, with my wife, and I had made the sails and the mast, and designed and built the boat from scratch. [Elisabetta Mori]: Did you use computers to design the boat? [Mike Jackson]: No. Boat design is much more like an art, well it was in those days anyway, though calculations have to be done for displacement. Now they've lots of computer graphics and things to do it, but not in those days. [Elisabetta Mori]: Not even for calculations? [Mike Jackson]: No. You need to take areas from shapes on drawings, and there were no computer facilities for that at that time. But I bought in a new fundamentally different way than had been used for over twenty years as the basis of the shape of a racing dinghy. I thought, ‘I can do better than that. I'm sure if you shape the mid-ship section differently it's a better boat’. More stable, it will plane faster, lower wetted area, etc. And so I did it in one. [Elisabetta Mori]: Which year was that one? [Mike Jackson]: Well, my first boat I built in 1956. I sold that when I needed to buy a house. I then built another boat in 1959 in which I was very successful, but by that time I'd done very little sea sailing and all the championships were at sea. With this next boat I sailed much more at sea. With that experience I knew I could design a championship winning boat. I built this in 1962 I won the National Championships in 1963 and again in 1964. Since then other designers copied my basis for design, and since then all championship winning boats in development classes in the UK have been built on the basis of design that I introduced. The mid-ship section is substantially different to boats built before time. I continued to race in the National Twelve class and designed a few more boats. I designed the Lark one design class of which over 2500 have been built. I stopped really winning at championship’s level when I got a bit older, but continued to race quite successfully in the class until about the mid 1990s. [Elisabetta Mori]: And then, how much are you still in boats, in sailing, do you know much about sailing boats now? [Mike Jackson]: Do you know the ‘49er’? The Olympic boat? [Elisabetta Mori]: Yes. [Mike Jackson]: Two trapezes, enormous gennaker. [Elisabetta Mori]: Can you explain a bit more about... [Mike Jackson]: I chaired the committee of the International Yacht Racing Union that was appointed to choose a new type of boat for the 2000 Olympics. This was done in 1966, when they were to select the classes for the following Olympics. We invited entrants to propose a boat to be considered for selection. We had a few months to examine them in detail before conducting sailing trials on Lake Garda in Italy in August. I think we had eleven designs competing for selection. Before the trials I had arranged to sail in every one so that I knew the sailing characteristics of each. For the past 30 or 40 years it had been normal to have trials for a proposed new Olympic class. This had been done by a selection committee observing the boats being sailed by the entrants own sailors, and making their own judgement on the most suitable boat. This had often led to accusations of political bias in the choice made. What I did for the first time was instead of the committee selecting the boat we invited national authorities to send impartial top sailors to sail all the boats and give their own preferred choice. This relieved us from all the political side resulting from pressure from boat builders and designers and countries with national pride in their own classes of boat. Sailors were sent by 20 countries from all around the world. Of these, sixteen plumped very firmly for the 49er to be chosen. The other four were all quite closely associated with the builders of the boats that they had been sailing at the event. And so there was no dissention whatever when the IYRU confirmed the 49er for the 2000 Olympics. Anyway, you were asking about my own sailing. The upshot of it was I so loved that 49er, I bought one myself and just sailed it for club racing. It's the only boat I've ever seen that I wished I'd been able to design and hadn't. It's a fabulous boat, totally unforgi ving of error, but equally rewarding when sailed well. You might have watched Olympic sailing in them, and seen that even the world’s best sailors sometimes capsize. I last sailed it when I was seventy one years old. That might be a world record. But I had a very amenable son to crew and look after me. When I sold that boat it was the end of my sailing [Elisabetta Mori]: And did you always sail with your wife, but, or also with other members of your.family?.. [Mike Jackson]: I sailed regularly with Helen from 1959. We won together the championships in ’63, ’64 and ’66, but she didn't really sail after that, and I never won championships without her, I was second a few times. We had our first child in 1965. Anyway, that's my sailing. [Elisabetta Mori]: And did you ever sail with your sons or daughters? How many did you have? [Mike Jackson]: Yes, we, we've got three. In addition, I told you that I had a previous unsuccessful marriage, and have three daughters from that. One is in Scotland, who still has this boat in the photograph, which I built in ’62. They've got it up there and they can sail it on a lake in, near Avimore. [Elisabetta Mori]: Do they all have the passion of sailing like you? [Mike Jackson]: Not really, not like I did, no. All of them sailed with me at some time or other and two of my first family were very good racing crews. My son, the youngest, crewed me regularly for a few years until he grew too large for a twelve foot boat. But in my old age I could not possibly have had such enjoyment with the 49er without him. [Elisabetta Mori]: And anyone of them has got the passion for computers? Or do they work in computers? [Mike Jackson]: Only my son did. He worked in Compustar with me for a couple of years just before I retired in 1995. After that he worked as a contract programmer for a number of companies before going into business on his own account in a totally different field. [Elisabetta Mori]: So finally, would you like to reflect on your working experience and maybe tell me what was your proudest achievement during your career? [Mike Jackson]: Well the biggest reflection on LEO is the disciplines that it embodied in handling data and thinking properly through the logic of jobs with flow charts. That really set the whole scene because if those are not attended to developments will cost more and greatly increase the likelihood of failure. Doubtless lack of these disciplines has contribute to some very high profile failures, particularly in the sixties and seventies of large computer systems. I think I very naturally, sort of, absorbed these disciplines; they fitted in with my own natural ways of thinking. But they were essential to get anything done effectively with computers. I was almost amazed when, despite joining in ’57 when computing was only a few years old, I saw the formality of the sort of things like the forms in which you write programmes and the detailed design of data forms. With direct entry into the computer these says, people seem unaware of the task of getting data into a computer sufficiently accurately, and the care with which you have to vet the data when it's in because there’ll always be some errors. In a job with large volumes like Freemans, with those volumes, what may seem a low error rate will result in chaos because of the amount of catching up and complaints from customers. Obviously my proudest achievement was at Freemans with such very large volumes and complexity in those early days, though investment fund management at Jupiter must come a close second as it was a one man personal job. The biggest thing that hit me at LEO was just the whole matter of discipline in getting computers applied. [Elisabetta Mori]: I would really thank Mike for his time and for sharing his memories with us. It's been a real pleasure talking to you. This interview with Mike Jackson has been recorded by the LEO Computer Society. The Society would like to thank him very much for his time and his reminiscences. The interview and the transcript form part of an Oral History Project to document the early use of electronic computers in business and other applications, but particularly in business. Any opinions expressed are those of the interviewee, that is Mike Jackson, and not of the Society. The copyright of this interview in recorded form and in transcript remains the property of the LEO Computer Society. [End]
Archive References : CMLEO/LS/AV/JACKSON-20190725 , DCMLEO20221230004
This exhibit has a reference ID of CH63002. Please quote this reference ID in any communication with the Centre for Computing History.
Click on the Images