Why the UK's IBM Failed
in 1968 the british government arranged a shotgun marriage between three computer manufacturers to create international computers limited or icl the government envisioned icl has a united kingdom's answer to ipm a national champion in computers but the company could not compete against american and east asian competition and eventually sold to fujitsu the united kingdom pioneered computing technology so why did icl fail in this video we're going to trace the history of the british computing giant from its punch card days to its final demise in 2000. sit back and relax this is going to be a long one but first i want to talk about the asian armature patreon if you like what this channel does you can support the work by joining the early access tier early access members get to see new videos and selected references for them before they are released to the public so head on over to the patreon page and take a look i deeply appreciate anything you'd be able to sign up for thank you and on with the show in the 1880s an american named hermann hollerath invented an electromechanical tabulating machine for punched cards he invented this tabulating machine for the 1890 u.s census the americans do a census every 10 years herm got the contract to do it but realized that it would take 20 years to count up all the cards the tabulating machine drew influences from cash registers and typewriters to do it faster than anything else before it was the first of its kind and revolutionized the sector hermann started the tabulating machine company of new york or tmc to sell this machine later on he sold tmc to an office machine conglomerate called computing tabulating recording company or ctr that name obviously had to go and so in 1924 they renamed themselves to international business machines or ibm let me give a little overview of punch card technology these machines use punch cards for data input output and storage the hollerith machines use cards strictly for reading and summarizing data counting the punched holes but by the 1920s the machines began performing basic mathematical operations like multiplication and division they were immensely useful to businesses for instance magazine subscriptions subscription punch cards had information about the shipping address amount due expiration date and subscription type subscribers filled them out and sent them back to the company the magazine company can then sort them by city state or subscriber name much faster than if it were to be done by hand punch cards had many advantages and disadvantages paper is relatively durable and whatnot but they did bend and break which led to input delays and their storage space was limited a two inch stack of ibm punch cards held about 40 640 bytes fundamentally however these were calculating machines they helped collate process and store data they did not run programs nor did they host applications but for a while that was fine across the pond in 1907 a few of hollera's british friends licenses technology for commercial and government use mr r p porter and sir raleigh philpotts a classic harry potter name if i have ever heard one founded a company called the british tabulating machine company or btm hollerith gave them an exclusive license to the punch card technology throughout the british empire except canada in exchange for 25 of revenue not profit but revenue an astonishing number that came about when porter and philpots could not raise the initial request of 20 000 pounds this license was inherited by ibm the first automatic tabulator became available in 1908 and processed about 150 cards per minute is soon sold to a variety of large organizations like the lancashire and yorkshire railway the great western railway and more punch card technology was on the rise in 1911 the registrar general used it to conduct a population census of england and wales scotland decided to use it too this majorly popularized the use of the technology and companies sprouted with their own variants it is the 1920s ibm and btm aligned together in the face of a common rival their major punch card competitor was a company called power samas a french american conglomerate their product was the powers machine despite having access to all of ibm's superior products r d and manufacturing capacities btm could not sell as well as its american counterpart ibm's american sales territory was only twice as large as btms yet it generated 20 times the revenue from that territory ibm easily crushed its rival in the united states they had 90 percent share there but the market in the british empire was evenly split between power samas and btm legendary ibm executive thomas watson blamed this underperformance on btm's sales incompetence their selling style can best be described as low-key and typically british as in business came after fishing hunting and shooting btm on the other hand again blamed the onerous 25 royalty rate a millstone around their neck the truth was probably somewhere in between but there is no denying that the two companies leaders were not all that fond of one another in 1937 btm chairman phil potts asked watson if ibm would reduce its royalty payment terms so that btm could sell more of its stock watson said well if you have any difficulty in placing any of the stock ibm would be delighted to pick it up phil potts shot back mr watson to sell you stock would be the last thing i would ever do in the 1930s both power samas and btm began expanding beyond simple import and sales a bi british movement was in effect and that motivated btm to start designing their own machines at the outbreak of world war ii btm had about 1 200 employees not a bad size firm but far smaller than ibm during the war both british punch card machine firms stopped investing in r d for their products their facilities were turned over to the uk government for war purposes they started making precision military equipment like bomber aircraft frames and gun aimers after the war ended the two companies began building up their sales organizations again with demobilized troops however those in the industry knew that things would never be the same again like so often the case in times of war world war ii accelerated the development of electronics and the store program computer such machines were to be the future the punch card industry was headed into obscurity but ibm and the rest of its punch card practitioners did not know it yet computers at the time were seen as expensive niche devices that only the government and the biggest of businesses would ever use perfectly reasonable assumption too since at the time they were about as big as your house they also cost over a million dollars back when a million dollars actually meant something selling million dollar house size computers was not the business ibm btm and power samas wanted they wanted to sell hundreds or perhaps thousands a smaller lower cost calculating machines so the steps they took were slow and cautious seeking to adopt certain aspects into their existing product lines to quote ibm advertising evolution not revolution the ibn 603 electronic multiplier was the company's first electronic calculator announced in september 1946 it could read six thousand cards an hour and quickly do several calculations this product sold very well moving hundreds of units two years later in 1948 the company developed and sold the ibm 604 electronic calculating punch this product also did mathematical steps like division and multiplication very quickly but it now also had the ability to execute some programming steps however this programming ability was limited you inserted a card in the back it then did the calculations via vacuum tubes and plug board and finally at the end out comes a card in the front with the answer thus programs are limited by the time from when you insert the card and when that card leaves the machine regardless the ibm 604 sold extremely well way more than management expected and they worked fine for many people for many years in 1974 over 25 years after its introduction there were still over 400 ibm 604s in use the year 1949 would see the founding of the people's republic of china and the retreat of the oh wait wrong video 1949 forever changed world history but it also happened to be extremely impactful for the british computing industry a number of massive developments in 1949 ibm and btm broke up the british and american punch card giants mutually dissolved their decades-long sales agreement and agreed to go into open competition with one another btm no longer gets access to ibm's product lineup and r d but at the same time they would no longer have to pay ibm that onerous 25 royalty rate this tripled their operating profit which the company could then invest in product development looking back at it the decision seems crazy btm generated 5 million pounds of revenue that year ibm did 400 million usd why on earth would btm try to compete with ibm of all companies and in some ways yes they were a little delusional in believing that british effort and british skill can be matched successfully against any competitor in our business whether national or international but the ibm of 1949 had yet to beat the technology colossus that it would soon become and btm had no idea that the computer revolution would so completely remake the world like it soon would looking back at it the license was something truly special it had no end date could only be ended by mutual consent and gave btm access to everything ibm sold including computers at cost plus 10 percent the agreement lasted half a century and kept ibm out of the british territories for that long i wonder what might have happened to the british computing industry had it continued shortly after the war ended british policymakers became concerned about their country's ability to exploit their own scientific and industrial innovations so in 1949 the labor government founded the national research development corporation or nrdc the goal of this quasi-governmental organization was two-fold first to secure the national development of the uk's public research second to acquire and hold the rights of any inventions from said research the nrdc's first general manager lord visited ibm in the summer of 1949. he left convinced that the company would soon develop and sell electronic data processing computers someone quoted him saying afterwards well i can tell you gentlemen that the only place you'll see punch cards and punch card equipment in 10 years time will be in the british museum housebury was off by a few years the u.s social security administration still used punch cards deep into the 1990s but the idea was on the right track he was also convinced correctly so that the british computing industry as it was then could not compete with the american giant he began visiting the various members of the british business community in an attempt to cajole them into the computing industry since ibm and its punch card affiliates were rather tepid on the computing revolution a few british companies dipped their toes into the space three in particular are worth noting ferranti a british electronics manufacturer founded all the way back in 1883. they had made radar and instruments during the war but with demobilization was concerned about what was next thus they partnered with manchester university to build a commercial computer this was supported by a contract from the ministry of supply thus in february 1951 just four months ahead of univac one the first verrante mark one computer was installed elliott brothers with two t's a british firm that goes back to the 1800s in 1947 the research laboratory produced a stored program computer which they attempted to refine into a commercial product in 1953 elliott brothers was the second british firm to install a commercial computer the product however was low powered and limited in scope they would later focus on niche low-power process computers and finally english electric the baby of the three founded after the end of world war one a diversified manufacturer that produced aeronautics and locomotives the company expanded into computers working alongside the national physical laboratory slowly ferranti and elliott were the only british firms pursuing the commercial computer space with any sort of speed or effort but ultimately they were in no position to compete with the americans where giants like ge rca and honeywell gently prodded the market to see if something was there however none of the relevant players felt that computers had that much commercial potential until the mid-1950s when ibm would finally fulfill lord housebury's prophecy and enter the computer market in 1955 ibm introduced the ibm 650 the world's first mass-produced computer this early general purpose computer was very popular throughout the 1950s and would be the first to make a profit in the market the 650 was a marvel however it was still early and the american giant had still yet to come to the united kingdom they would only ship two such computers to the british isles throughout the entirety of the 1950s it left a massive door open for a domestic player of all the domestic players who could have entered the british general computer market btm had the most potential they had built up their r d capacity in electronics over the past few years and they had also produced a computer of their own in conjunction with the university of london the hollerith electric computer or hec one yet throughout the decade btm did not show any real interest in entering the computing space in 1956 they did release the hec4 machine but 90 percent of the revenue still came from their very profitable punch card machines marketed mostly to accounting departments to be fair few self-respecting business analysts would tell them otherwise btm made great products and it was a great business model too for instance ibm made 20 of its revenue throughout the 1950s from selling the cards alone but as a 1950s came to a close it became clear that the company was facing a massive challenge from ibm in its major export markets overseas ibm did not have a foothold in the united kingdom but they were winning in australia south africa and india decreasing competition had weakened power samas over the years and by 1958 they were almost dead btm saw an opportunity to consolidate their industry so in 1959 the two merged creating a company called international computers and tabulators limited or ict acquiring powers would create a massive headache for the btm management digesting their mechanically minded r d department lost the company years of progress years that could have been spent upgrading their tabulating machines to electronic computing ict and its 17 000 employees had a big challenge ahead but they felt they had a good chance to come out with a competitive product half of their r d spend would be allocated towards the computing industry however in 1956 ibm introduced the future the ibm 1401 replaced vacuum tubes with transistors and as a result was cheaper faster and more reliable than any computer before it besides being a technical marvel the 1401 brought with it an ecosystem of enterprise hardware peripherals printers mass storage and application software altogether the ibm 1401 served as a complete data processing system that went after the core use case of the entire punch card industry the 1401 sold more units in the first five weeks than ibm management had expected to sell throughout its entire lifetime he would go on to sell over 10 000 installations by 1962 the computer system was the company's single most important product the 1401 with its transistor hardware obsoleted every computer that dared to use vacuum tubes so basically every first generation computer product in britain for each of the companies now dabbling in the computer industry the question was whether or not to bulk up their capacity on both the product and business sides to challenge ibm or drop out english electric and elliott automation stayed in the business though elliot retreated to a niche space a few others did drop out including ferranti and ge english electric and ict took the opportunity to bulk up their operations by acquiring the assets and employees of the dropouts in 1961 ict acquired the computer interests of gec computers 1962 the computer interests of electrical and musical industries or emi the former music company then finally in 1963 the computer interests of ferranti this last acquisition gave ict a key product for its future the canadian made ferrante packard 6000.
english electric for their part purchased the computer interests of leo computers in 1963 and marconi in 1964. by 1964 the british computing industry was basically just english electric and ict but this wave of consolidation left both those remaining players with a form of indigestion mergers can be a good thing for instance i talked about japan's successful intervention in its fragmented automobile industry during the 1960s arranging shotgun mergers and marriages between these failing companies until two large ones emerged toyota and nissan however when done haphazardly and for its own sake the result is just a massive mess especially in the case of the computer industry when each merged company differed in its engineers approach techniques used and systems administration ict was making too many computer products and with the exception of the ferrante packard 6000 none of those products were good and none of their software's or hardwares were compatible with each other it resulted in a large technology gap a gap further demonstrated in 1964. the transistor revolution swept the entire electronics industry computer manufacturers rushed to replace their existing first generation vacuum 2 products with those using transistors but with this emerged a new problem each time a piece of hardware was replaced or upgraded the operating team needed to write brand new software for it as a result software costs were starting to grow and in some cases surpassing the cost of purchasing the hardware analysts called this a software crisis ibm decided that it needed to stabilize the platform so it invested 500 million dollars to develop the system 360. the system 360 was a range of computers 6 processors and 40 peripherals that would replace all of ibm's previous products and offer a single scalable architecture now customers can write and adopt software with no fear of it outgrowing their hardware they simply upgrade to another computer and their programs came along with it the system 360 would be considered the third generation of computer systems and it caught the entire industry by surprise ict had been mulling the product idea of a family of computers for a while but the original road map had them coming out with it in 1968.
ict decided to rework that aforementioned ferranti packard 6000 by adding higher and lower end computers to create the ict 1900 series they then quickly released it taking advantage of an ibm production delay that kept uk deliveries from being made until nineteen sixty six this helped the ict 1900 series sell well despite being a technically inferior product english electric on the other hand decided to license rca's computer architecture which was compatible with ibm system 360 products the resulting product family the system 4 was announced in 1965. equipped with advanced transistors from marconi it sold well prestige customers included the british post office the uk atomic energy authority and more however this deeply linked one of the uk's top computer companies to r d done in the united states and this create a great deal of unrest within the political class in october 1964 the harold wilson government came to power one of the government's first acts was to establish a ministry of technology or mintec and mintec's top focus was the british computer industry if action was not taken quickly the british computer industry would rapidly cease to exist facing as was the case in other european countries the most formidable competition from the american giants in both the uk and france political leadership were increasingly concerned about america's widening technology gap in 1964 general electric purchased a controlling share of france's biggest computer manufacturer bull and then in the mid-1960s the usa denied an export license for several american computers with the goal of preventing the french from perfecting its h-bomb these concerns spurred the two countries to double down on the concord project it also led france to spend 100 million dollars on its plan kalkul to create a french national champion for computers plan calcul ultimately failed like many others of its type across the continent each of the european countries invested considerable resources to start and run their own ibm but none of those companies could compete against the real thing kind of reminds me of google today the uk was one of the last bastions of europe where ibm did not have dominating market share mintec sought to consolidate their industry and make it stand up in competition against ibm that meant merging the two big players english electric and ict the government's first meetings between the two happened in 1964 literally right after mintec's founding but back then the two companies felt that they were in a strong position and thus were cold to the proposed marriage the wilson government was re-elected in 1966 and they doggedly pursued their vision of a british national champion in computers and this time they offered up to some 30 million pounds as a dowry by now things had vastly changed for the two companies ict was doing quite well while english electric was finding the development costs for the system 4 to be onerous to the bottom line the two firms agreed to a merger in 1967 but a delay over the financial arrangements proved costly 1968 would see the worsening of the uk economy ultimately resulting in the devaluation of the british pound as a result mintec went back on its 30 million pound dowry providing only 13.5 million regardless the merger finalized in july 1968 creating icl icl generated 100 million pounds of annual revenue employed 34 000 people and was the world's largest non-american computer manufacturer icl's mission now was to rationalize its range of computers bulk up its r d output and compete with ibm they inherited two mainframe product families from their predecessors the 1900 series from ict and the system 4 from english electric the two were incompatible with one another icl initially wanted to phase out the 1900 which was technically out of date but political pressure from the labor government prevented that from happening the system 4 was not a british made thing and that mattered the company developed a range of computer lines ranging from the most powerful mainframe to an interesting low-powered computer icl eventually can the ladder and sought to sell these high-end mainframe computers to prestige clients like oxford university it took some persuading political interference kept icl in its lane as a mainframe manufacturer ironically this would hasten its eventual marginalization the over emphasis on the expensive mainframe industry caused the company to completely miss out on the mini computer boom of the 1990s the same boom that would overtake the ibm giant 1968 would see the computer industry's first recession which took out general electric and rca from the business both companies sold their computing interests off to competitors icl called for government intervention and aid but 1970 saw the election of edward heath and a new conservative government that did not look kindly on such things his government merged mintec into the department of tech and industry and maintained an anti-interventionist hand the uk government eventually had to rescue rolls-royce in february 1971.
this led people to point out that heath's anti-interventionist stand did not make sense in light of france and germany spending hundreds of millions of pounds in government support to support their own computer industries icl on the other hand was laying off 3 000 workers and cutting r d spend in 1971. thus icl did ultimately receive a 40 million pound loan from the government this helped maintain r d from 1972 to 1977. however at this point it became clear that the company could not afford the heavy r d cost of staying competitive in the british mainframe computing industry annual r d spend rose by 50 from 1974 to 1978. two costs in particular were very heavy the first was the cost of writing new software over a third of icl's entire r d budget was dedicated to writing software every time a new architecture came out new operating systems had to be written for it and second the tightening cycles of semiconductor production you no longer had five or seven years to come out with a new product semiconductor manufacturing technology was accelerating and icl could no longer compete one might point out that as a percentage of revenue about seven percent r d costs were not all that bad this is due to the company's successful entry into the small business space with the 1976 acquisition of singer business machines and its very successful system 10 computers however everyone focused in on the mainframe business those were the quote unquote real computers after all and in there ico was lagging they started looking for a collaborator capable of bringing real technical chops to the table i'm going to skip the long history of japanese computing and get to the good stuff fujitsu was one of many japanese computer companies seeking to challenge ibm but they all fell on their back foot when the system 360 came out fujitsu looking to respond partnered with the amdahl corporation to release a series of very advanced computers fujitsu's more advanced semiconductor technologies gave those amdahl plug compatible mainframes technical parity with ibm's highest end frames and sold very well encouraged by this success fujitsu directly entered the computer mainframe market and started a war with ibm by 1980 the japanese company had become one of ibm's very few real competitors and the only company to outsell the giant in japan ibm responded to fujitsu's challenge with extreme prejudice and tighten its semiconductor production cycle they replaced their entire line of system 360 and 370 machines with the ibm-4300 series the 4300 offered a four times improvement in cost performance over their predecessors and kaboom the market icl its technology at one point just three years behind ibm found itself caught in the crossfire initial estimates in 1978 indicated that annual r d spend would have to nearly triple to 89 million pounds by 1983 in order to maintain market competitiveness 1980 hearkened another economic recession the exchange rate between the british pound and the us dollar ballooned to a peak of 2.35 ico projected a 50 million pound loss as sales both abroad and at home cratered icl management decided that this trend could not be sustained they approached the government for more help but the newly elected thatcher government was loath to provide it they eventually did agree to a 200 million pound loan guarantee but demanded that the management team first be replaced even with the loan icl cut 10 000 jobs from 1980 to 1985.
most of all icl had to manage its skyrocketing r d spend and that meant cutting down on that semiconductor burden they went to fairchild lsi logic and a variety of other companies in search of a partnership the best the americans can do was 3 000 logic gates on a chip far better than icl's 1000 gates but fujitsu blew them away with 5 000 gates furthermore fujitsu said if you're our partner we'll tell you what our roadmap looks like for the next five years so you can design a computer to use chips that you know don't even exist yet knowing that they're coming thus in 1981 they struck a technology transfer deal with fujitsu ico would transfer to the japanese company some of its leading peripheral technologies like disks tapes and the such they would also sell fujitsu's biggest ibm compatible mainframes icl would in turn gain access to fujitsu's semiconductor technology which as it turned out was even more advanced than they had at first showed eight thousand gates rather than five thousand two processors one medium and another small entered the market has the series 39 in 1985. but by then ico was already taken over by british telecommunications firm standard telephone and cables or stc in 1984. ibm had already released the home computer in 1981 and the mainframe market had turned into a dinosaur the 1981 rescue and the 52 partnership ended the company's hopes of being a national champion the company slashed its r d spent and began striking partnerships with outside vendors to create its home office and enterprise products ico closed their manufacturing plants and focused on its sales and systems integration functions at this point the company evolved into a simple assembler and distributor profits wise ico was doing fine making 50 gross margins sdc soon ran into financial crisis after acquiring icl so ironically enough the computer maker contributed the majority of their profits thereafter but those profits should not distract you from the fact that the company simply sourced other people's cheapest components put them together and sold them to its existing roster of clients icl would never be the computing leader it's thought to have been in 1990 fujitsu acquired icl and that was a minor shock to the nation the japanese company promised that it would eventually float the brand but after 2000 shut it down for good looking back on the downfall of icl you see a variety of common trends first the value of having technical primacy the british were always following ibm and the japanese giants and could not gain business leadership of its geographic domain second the splintering of the computer industry people predicted that the computer industry would become like telecom massive vertically integrated combines ibm sought to be that combine but in the end the competitive intensity and sheer size of the market caused it to splinter apart individual companies emerged focus on certain previously niche products like software and they raised the table stakes for everyone else to the point where the combines could not keep up how can icl invest as much into operating systems and applications as microsoft can and finally the hollowness of national prestige icl and its political sponsors wanted to beat ibm where it was at its strongest big bulky mainframe computers so they bulked up and invested in order to be like ibm but this national champion could never muster the same amount of resources and in the end they too ran head first into the same market circumstances that caused ibm's decline into irrelevancy in the 1990s all right that's it for tonight thanks for watching if you enjoyed the video consider subscribing check out the newsletter or follow the twitter want to send me an email drop me a line at john asianometry.com i love reading your
emails introduce yourself suggest a topic or more until next time i'll see you guys later