IBM
0213 Demise of IBM Biomedical
In terms of the market segments addressed by IBM Biomedical Group, unfortunately the (5880) ECG products were marketed to totally different users (cardiologists) than the blood products (whose users were haematologists). There was no synergy at all. Even within the blood products, the 2991 and 2997 were, again, bought by different groups of purchasers. In this case, again, the synergy was marginal (even the technology, though employing similar concepts, was not truly shared).
Elsewhere, the constant change syndrome, which was so effective elsewhere in IBM, exacted a price in the case of Biomedical manufacturing. In its seven year life the plant was moved no less than three times, admittedly all within a radius of 5 miles (and keeping much the same workforce), with considerable disruption and associated cost. The biggest cost though was that this ultimately resulted in the decision being taken to move to a building which could cater for all foreseeable future expansion, so that future moves would be unnecessary. This building was of the order of 100,000 sq. ft. and the cost of leasing it, and servicing it, was in itself not negligible. The biggest overhead though was a variant of that identified by C Northcote Parkinson in his observation that work tends to expand to fill the time available; in this case it expanded to fill the space available. It was evident on walking into the building soon after it was occupied that it was already being relatively fully utilised.
One space planning oddity, which surely only an IBM division could commit, was to separate its head office (and also ECG development lab) from the plant. This had an admirable pedigree elsewhere in IBM, and normally generated a number of advantages; it allowed a better, less biased, perspective and improved security on new products. For what was, in effect, a small company however this was an unjustifiable overhead; and resulted in a loss of top management control (where even the sales operation was based at the plant). The two locations were only some 60 miles apart, but were on opposite sides of New York city (the head office was at Mount Kisco in upstate New York), and the travelling time between the two was of the order of 3 hours. The head office was, though, conveniently close to Armonk, which may have entered into the original political calculations of the group general management!
In the case of Biomedical, the sensible decision was eventually taken and the head office was consolidated with all the other functions at Dayton. It was clear however that the ECG development staff would not relocate, and they were accordingly moved to Poughkeepsie; even further from Dayton! The assumption as to their unwillingness to relocate was correct, though, for when this lab too was eventually moved to Dayton almost all the key personnel transferred to other divisions!
This then brings me to the final aspect of Biomedical; its overall structure. The 'anomaly' of this was brought home to me eventually, during my review of the future options for the group. It was as a result of a comparison with my previous company, BTR. Shortly before its demise Biomedical had a world-wide turnover of around $30 million, and a workforce approaching 500. Prior to joining IBM I myself had run a part of BTR, its Burton on Trent (Polymeric) Group, which coincidentally had been of very much the same size. Although BTR was even then a very sophisticated company, the comparison with IBM could not have shown greater differences. The structure at BTR was still that which would be recognisable to a manager in almost any other medium size company or division. It was a structure that was relatively autonomous, with its own separate identity. Indeed the 'hands‑off' management style developed by BTR head office at that time is believed to have been largely responsible for its subsequent success. The structure of communications within my group there typically ran back from its periphery to its own centre. Communication with the rest of BTR was in the main hierarchically via senior management; and typically through myself. For most of its employees the focus was the narrow confines of group itself, not BTR overall. The pyramidal organisation of the group would be instantly recognisable to most other businesses (with departments for manufacturing, engineering, accounts, sales etc.)
Biomedical Group on the other hand looked most like an IBM branch (or staff department, since the environment was much the same throughout the whole IBM marketing operation). It did not have a unique identity, with a clearly definable periphery. It was instead most obviously a part of a larger, relatively uniform, organisation. Communications did not lead from the periphery to the centre, but covered the group in a matrix; which led outside, at all levels, to connect with company-wide systems.
My interpretation of the comparison with BTR, and other companies, was quite simply that the IBM which then existed could not easily, if at all, avoid its bureaucratic structure. No matter how hard it wished otherwise, the structure and the culture were far too strong. The bureaucracy would always strike back. The series of IBM senior managers drafted into Biomedical to resolve the problems (and there was a new General Manager every two years or so, much as there was later with PC group) certainly could not handle it. They were not in a position to even recognise that the structure should be otherwise, for it was, after all, exactly the same structure they had been previously immersed in elsewhere in IBM. Even an outsider (the standard answer in many other companies, but then an unthinkable option in view of IBM's internal promotion philosophy), who might perhaps have recognised the anomalies, would have been ineffective because he or she would still have been unable to cope with IBM's unique communications with the rest of the bureaucracy outside Biomedical Group!
The main reason that I eventually persuaded IBM to close down the Biomedical IBU, in 1984 after seven years of existence, was that there appeared to be no simple way that IBM (and probably other companies operating on a similar scale) could avoid the impact of its bureaucracy; no matter how hard Armonk (and especially John Opel) tried to isolate, and shield, them - as key developing businesses. The problem of the bureaucracy had to be faced squarely, and in the case of developing businesses this meant (at least in the case of IBM) they were unlikely to be profitable (when saddled with all the overheads of the bureaucracy) on a gross revenue of less than $200 million per annum. IBM's investment strategies needed to take this into account. Though John Opel must have read my supporting case when he signed the IBU's death warrant, which was incidentally a brave act where such IBUs were his personal initiative, he showed no evidence of having understood the key message as he wrestled with the problem on the wider stage over the next months!
At the time, I calculated that, as a very approximate rule of thumb, a revenue of $200 million per annum was the minimum required for an IBM group to be able to carry the overheads of the bureaucracy. In the case of PC group, which quickly followed Biomedical in being formed as an IBU, it was indeed fortunate that within the honeymoon period it easily surpassed the $200 million barrier; and initially became a tremendous success, despite the overheads. This was, in all probability, largely possible because the Armonk bureaucracy fortunately did lose control, as John Opel wanted, for a matter of a few crucial months; when it could not initially cope with the new policy directives of no bureaucracy!
It was clear to me that Biomedical Group would never reach this $200 million level before the weight of the bureaucracy pulled it under; so I accordingly recommended, for this reason alone, that the group be discontinued. It is to the credit of EHQ (European Headquarters) that, after some initial reservations, senior management there backed this recommendation. But most of all my admiration goes to John Opel who, presented with this recommendation in competition with one from the US for further investment, took the decision to close down the IBU. This was a brave decision. Biomedical was, after all, the first of his IBU's; launched in a fanfare of publicity. To kill it could be seen as an admission of defeat; just before he was due to retire. It is a measure of his, and Armonk's, maturity at that time that there was no hesitation.
It is my observation that the quality, and courage (or security), of senior management is best shown by how it handles its failures. How it handles the close down of such a venture says more than how it copes with success. Armonk then passed with flying colours; though such decisions took longer than necessary, as even within the security of IBM there were few management teams willing to present Armonk with such a brutally clear decision as did Biomedical. The pity is that John Akers later did not seem to have the same sort of courage.
The conclusion that I then drew from this experience was that IBM was locked into its existing business, except where new ventures could be rapidly grown past the $200 million limit; not necessarily that onerous a condition for a company with IBM's financial muscle at the time. Survival would have been helped by a recognisable link to the existing business; which made the mismatch with the bureaucracy that much less severe. Even then the success of the new venture was likely to be very dependent upon having a critical mass of enthusiasts who could 'manage' the traditional IBM bureaucracy to the new group's advantage. Small inexperienced groups, such as Biomedical Group, might be very successful in relation to other companies in their chosen marketplaces (as indeed Biomedical Group was; it achieved technological and marketing leadership in its field) but within the IBM framework they simply could not be profitable.
Some of the possible difficulties of the smaller units were illustrated by the launch of the Business Development Division, which was to become the home of the IBU's. It was also the home for something of a rag‑bag of IBM's other ventures, including typewriters and bureau services, which were declining businesses rather than growth potential. For me it was most graphically illustrated at the management 'kick‑off' meeting in the UK. As a succession of speakers extolled the glowing future for the growth business (the PC in particular), I looked around and noted that less than 5% of the management present belonged to these new 'growth' groups; 95% were concerned instead with halting the decline in the other, dying, businesses!
hits