IBM

9082 IBM13 - GERSTNER & CO; TURNED AROUND OR HOLLOWED OUT?

This chapter looks at life after death at IBM. Thus, the new IBM which emerged from the disasters was, in many respects, a totally different organisation. As we will see, it was taken over by outsiders and rebuilt in the image not just of that taught by the Harvard Business School but of that favoured by the leading management consultancies.

Perhaps as a result, it seems - on the surface at least - to have been much more amenable to the application of conventional management theory. It has, indeed, become an ordinary company! But, according to Doug Garr[i] in the opening to his excellent book (IBM Redux) about this period, "Even Gerstner [the new CEO], a man with a considerable appetite for achievement, admits IBM will never be an omnipotent presence again." This nicely sums up the new IBM. The old IBM is dead, long live the new? 

It may be an ordinary company now, and Gerstner reportedly[ii] even wanted to distance it from other high-tech companies, but it is still a giant amongst even the other multinationals. Its revenues at the turn of the Millennium (in 2000) were just short of $90 billion, but still short (at $88.4 billion) by almost $12 billion, of the $100 billion target it had once set itself for 1990! Its own funds flow statement showed it to be more than $3.5 billion in funds in 2000; though this was steadily declining.

The Challenge…even when Gerstner took over in 1993, and despite the recent disasters, IBM still had some strengths:

Indeed, surprisingly, despite its 1991 and 1992 annual accounts showing record-breaking losses these were almost entirely due to the special charges added to the accounts.

It is true that 1991 was not about to be an especially profitable year, only recording earnings of $600 million (less than a tenth of those of the previous year). But IBM was still not loss-making, and $600 million would have been a record high profit for almost any other company!. Much the same would have been the case in 1992 (with a potential operating profit of more than $1.5 billion); though the figure would have been reduced to almost zero (actually still a positive $34 million) in 1993 - Gerstner's first year. The decision to make a bad year (1991) even worse - and get as many future costs out of the way at the same time - was deliberately made by Akers and his board; in much the same way that other companies have traditionally used 'bad' years to get all the bad news out of the way in one go! It is likely, however, that the similar decision the following year (1992) was not made by Akers - since it cost him his job! The remarkable turn-around Gerstner engineered two years after his arrival - in 1995 - was probably just as much a result of the advance costs squirreled away in the previous three years of dramatic 'losses'.

 On the other hand, IBM was facing a number of major problems:

Intrinsic Strengths…the example of IBM, which by 1993 looked to be on the ropes, shows just how much life may be left in such an organisation - if its resources are profitably used, and it has the confidence to use them. On the other hand, at a time when creative accounting is widely applied, the annual reports may not always record the true position - and a degree of cynicism about the figures is often justified. Management theory tends to focus on the superficial as did John Akers - and, as we will see - with one or two important exceptions, so did Lou Gerstner after him!.

Lou Gerstner

The new CEO had previously only had three jobs in his career, but they had been significant ones. Starting as one of the few chosen by McKinsey to be a high-flying consultant, where he shone as a future star, he had then spent eleven years at American Express, where he had risen to be its number two. For the four years prior to IBM he had run RJR Nabisco, following its leveraged buy-out (the largest in US history), where he had led a seemingly brilliant turnaround - including halving its $29 billion debt.

At RJR Nabisco he had ruthlessly cut cost, not least by firing thousands of employees - something which IBM also needed to do! He was not loveable, but Wall Street liked him; and he was lucky!

The outside world worried about his lack of technological knowledge but - as I have already hinted - that was just about the last thing IBM then needed. As it turned out, more than anything it needed a boost to its self-confidence. Thus, seemingly against all the odds (and unimaginable under Akers), the turnaround was well under way less that two years later.

He was also a good negotiator. His initial salary package at IBM was a complex one, but would run into millions of dollars, even if he failed; and IBM guaranteed the gains on his RJR Nabisco stock options, once again luckily for him in view of future developments there! It was later estimated that, if he succeeded over the longer term the package could add up to around $500 million.

Axe Men…a relatively new breed of executives emerged in the 1990s, whose reputation for turnarounds was largely based on their ability to cut costs in the short-term, usually by deep cuts in the workforce (including, almost for the first time, in management). It was an unpalatable job, but even so probably did not justify the reward packages they received. Indeed, the long-term impact of their efforts typically had yet to be seen when they were given their golden bonuses - and often then moved on! The axe may sometimes be needed, but its use is not a panacea, and should be judicious while keeping an eye on the longer term. This is one of the embarrassing topics which is usually hidden behind terms such as 're-engineering' so that its more unpalatable aspects can be ignored.

The Early Gerstner …born in 1942, the second of four sons, Lou Gerstner had a strict, but relatively modest, upbringing. His father was a night traffic manager with a brewery and his mother worked in the local community college. They both worked hard to put all four children though college, starting with the local Catholic high school - the legacy of which remains in his religious inclinations. Despite being considered smart rather than intellectual, he worked hard - a trait he has to this day - to achieve outstanding academic results.

Even so, he had to follow in the footsteps of his older brother, Dick. Not least, Dick later became a lifetime IBMer; joining as an engineer and rising through the ranks to eventually head up its Personal Systems Group in 1988, when he was seen as a serious candidate to replace Akers as overall CEO. Unfortunately he was then plagued by back problems, Later (too late for his career) this was diagnosed as Lyme disease (which was then easily curable) but before this he was, in 1989, effective forced into semi-retirement (which was formally confirmed in 1992).

Different commentators held different views about the relationship between the two brothers. Some expected Dick to return to a high level position under his brother; especially as he was once more working for IBM, albeit on a consultancy contract, when Lou arrived. But, paradoxically in view of the precedents set by the Watson dynasty, Lou apparently thought this might smack of nepotism. In any case, those who knew the two well reported that that they were not especially close; and indeed were very competitive! Some even claimed their dislike for each other was intense; though others claimed there was some brotherly love between them! Robert Slater, in his book 'Saving Big Blue', even went so far as to say that Dick Gerstner was sacrificed in order to very publicly demonstrate the dictum of his brother Lou, that "We had a few public hangings of people who didn't want to get on the new programs. That told everybody we were serious."[iii] Dick was the epitome of the old IBM, and perhaps did pay the price for this. Whatever the reason, Dick never featured again in the IBM story!

Academic Track Record…Lou gained a scholarship to Dartmouth, where he was just as assiduous in his studies, as an engineering science major; something which is often overlooked in view of his later career in consumer goods and services. I followed a similar route myself, first as a physics graduate from Imperial College, the UK's top science university, and then followed a career in marketing and general management - again mainly in consumer goods - before joining IBM. I can report, with some confidence then, that - against all the hype at the time - this background would have made Gerstner an ideal 'trainee', even in the good old days of IBM. It was even noted that, in his initial presentations to the outside world, that - much like any IBMer - he wore a (perfectly pressed) dark suit and white shirt! Mind you, luckily he wasn't just a trainee; for he had to move all the immovable objects in IBM - something you will realise I myself had clearly not been able to even start!

Perhaps more important, in view of the role he was about to undertake, he then took an MBA at Harvard, as indeed had his unfortunate predecessor at IBM; and up to this point his background was almost indistinguishable from Akers. But he followed the MBA, as did a number of other high-flyers, by joining McKinsey; as indeed I myself almost did - since it was then the top of the tree in terms of intellectual achievement in the commercial sector! This immersion in the consultancy perspective was to become a major plank in his later approach to the problems of IBM; as it also was for a number of the other outsiders he brought with him. Despite his reputation for being boastful and arrogant - I identify more and more with him - he worked successfully with a range of clients and was promoted rapidly; before he made his name by initiating and building McKinsey's Finance Group practice.

Consultant MBAs…the high flyers in the 1980s and 1990s had almost invariably taken an MBA - something for which I am grateful (where my own School graduates a quarter of the UK total, and I teach most of these). On the other hand, our students are part-timers and we deliberately teach the 'theory' alongside the experience they are gaining in their employment; and use this 'marriage' to practically consolidate the lessons. This is a luxury the blue-chip business schools do not have available; even though they may simulate it by the Harvard case-study teaching methods. So, their students (frequently straight out of the ivy-league universities) are taught the best theory money can buy. Graduating, these high-flyers then tend to go to the financial institutions or the consultancies; to make their reputation by putting into practice all the theory they have so expensively learned, To these they add the 'specialities of the house', those additional 'theories' (the USPs) the consultancy specialises in! You would hardly expect the business schools, or the consultancies, to do other than promote the lessons they rely on for their income - and they don't!

In 1968 he married Elizabeth Robbins Link, whose character is reportedly the antithesis of his; "…while Gerstner is shy, abrupt, and socially lacking on occasion, Robin is outgoing, warm, gracious and charming[iv]." Now with two grown up children, he reportedly is still a committed family man.

Amex

Following the well trod McKinsey career path, in 1978 he was recruited by American Express - on whose account he was working - to head up its Travel Related Services (TRS)  group; though he was gently introduced to this by first running a (charge) card unit. As the full head of TRS, soon after his arrival, he ran Amex's card offerings across the board. As such he was number five in its hen-pecking order; though he had thought he would be number three! Starting a pattern later seen at IBM, he brought in Rick Thoman; an original member of his Finance Group at McKinsey. Gerstner also immersed himself in marketing activities, becoming an ad agency's ideal client - in that, if he trusted the agency, he was willing to approve campaigns he personally disliked. He also demonstrated a passion for branding; something which was, later, to color his first moves at IBM. Following a number of political battles, which might have prepared him for his baptism of fire at IBM, he eventually moved up to become Amex's number two; as its president.

He also took on, in 1988, responsibility for implementing its Project Genesis; its five year $250 million global programme of upgrading its IT infra-structure. The resulting near-disaster, it actually took nine years with huge cost overruns, was largely laid at IBM's door; which must have coloured Gerstner's views before he arrived at IBM.

RJR Nabisco…despite being the crown-prince at Amex, Gerstner seemingly felt he was in something of a cul-de-sac, for in 1988 he started looking around for alternatives (first campaigning to be CEO of United Airlines[v]). The opportunity came, however, when RJR Nabisco was taken over by a $24.5 billion junk bond buyout; the largest in US history. The company was, though, then saddled with a mountain of debt, around $26 billion, and on arrival Gerstner had to start with something of a fire-sale, selling Del Monte for $1.5 billion along with the candy business which included Planters, Lifesavers and Baby Ruth. At the same time he found that almost two thirds of the two dozen strong management team had bailed out before he arrived. Once again, therefore, he brought in his loyal lieutenants; Rick Thoman and David Kalis - his media gatekeeper from Amex.

Setting a precedent for his arrival at IBM, he clocked up a quarter of a million air miles in his first year - pressing the flesh with customers and employees.

On the other hand, it seems clear that Gerstner was not comfortable with the tobacco business. Perhaps it was that, as much as the financial constraints which he later described as 'traumatic'[vi], which made him susceptible to IBM's approach only four years after joining RJR Nabisco. Even so, he probably left that company in better shape than when he joined; at least he had reduced its debt mountain. Luck was, though, once more on his side. The day after he arrived at IBM a price war erupted in the cigarette market; and when (in 1999) the RJR Nabisco food and tobacco businesses were split apart again the conglomerate was only worth $9.5 billion!

Get Out Ahead of the Receiver…one of the key lessons, which nobody seems to talk about, is that of developing the sixth sense which enables you to move on just before everything comes crashing down about your ears. This is more than luck, which every successful manager needs, it is an acute awareness of the danger which lurks in the accounting figures; something Akers never saw. Some of the most successful, or at least best-paid, executives have left a trail of collapsed businesses behind them; yet they still command vast salaries. Perhaps the boards of directors who appoint them should think twice in future!

The New Master at IBM

Over the years there had been political battles amongst the otherwise rather anonymous senior executives on IBM's board. The time of TV Learson, in particular, had been noted for these. But they were largely about individual's moves to gain power in parts of IBM. Having arrived at IBM, Gerstner and his team, however, set out to destroy their opponents across the senior management of IBM. The new culture had to force the old into submission. The creation of such a hierarchical autocracy was against the flow in the rest of industry, which was about to replace the remnants of this with collegiate networking; though the latter had not served IBM well in terms of facing up to the changes it had needed to make over the previous decade!

The first group to feel the heat was the IBM Board, which had just hired him! Almost half its members disappeared. Even with some new recruits it was still reduced to a more manageable dozen members.

Unsullied Hands…in a genuine turnaround situation, it is almost inevitable that all the insiders - no matter how inherently competent - will be unsuited to leading the organisation out of its troubles. This is partly because, due to the classic Groupthink symptoms, they will be incapable of any meaningful strategic response; and, even if they were, the other insiders would then shun them as traitors to the group! The search for a suitable outsider will, however, be as difficult as it was for IBM; the few who might be capable probably will refuse! But most of the theorists, and all of the head-hunters, would agree on the desirability of appointing such a new broom!

The New Team…indeed, Gerstner's greatest priority was to bring in his own team; with which to face down IBM's existing management. His first hire was, significantly, a communications officer; David Kalis. He had previously moulded Gerstner's all-important image at both American Express and RJR Nabisco; and was already in place - working the press - at Gerstner's debut conference.

Kalis did not just handle the press, he was also the gatekeeper to Gerstner's office. Like his equivalents in the party political machines, he reportedly relished the cloak and dagger aspects of his work; and his job was to protect Gerstner from the press, at any cost, which he was said to do brilliantly. Some authors have claimed that Kalis was the first PR professional to be hired at such level within IBM, which in terms of outside PR experience is probably true, and that this was a dramatic break with the past, which probably isn't true. Having myself worked in the previously named 'Corporate Affairs' Group, I would merely note that this change was more about the new CEO's relations with the press, something which had not been thought necessary in previous times, where the group had earlier focused more on IBM's political lobbying activities and keeping the frequent indiscretions of its many staff out of the press!

On the other hand, Gerstner soon had them revert to type, by having them protect him from the media rather than use him to promote IBM's image. He was soon limiting interviews with journalists to 45 minutes - and demanded the list of their questions in advance - as had happened in the bad old days at IBM! Worse still, he later got involved in a vendetta with Fortune - because it had the effrontery to suggest he was rude and boastful (which, of course, he was!).

Perhaps sensing, as I had done, that (mass) marketing was one of IBM's greatest weaknesses. An important early appointment - just two months later - was another of his proteges; Abby Kohnstamm, who had previously spent her who career at Amex, ending up as Senior Vice President of its card marketing operations. A strong supporter of Gerstner's views on branding, she initially reported to Kalis but later directly to Gerstner. Indeed, Gerstner's drive on brand awareness did recover some ground; though an Information Week survey[vii] of 200 IT managers, at the end of the 1990s, still cited twice as many who thought Microsoft was trustworthy as thought IBM was!

Spin Doctors…despite their own poor image, spin doctors are essential to a turnaround. What matters most, in the initial stages, is not what is actually being done but what is perceived as being done. The staff, as much as customers and the financial markets, will want to be reassured; and that is a job for the best spin merchant money can buy! On the other hand, even ordinary PR is largely ignored by management theorists!

Jerry York

The first heavyweight to be recruited, though by headhunters rather than from Gerstner's existing clan, was to be its Chief Financial Officer (CFO); Jerome B York. Having just spent fourteen years at Chrysler, and having risen to be its CFO, he had overseen a $3 billion cost-cutting effort there, along with the selling off of $3 billion of assets. Reportedly a rough, results-oriented, manager who - favouring four letter words above the rest of the vocabulary - had a reputation for using fear as a motivator[viii], he was seen as the ideal partner to bring IBM's costs into line.

A week after York's appointment, Gerald Czarnecki completed the team as Senior Vice-President of Human Resources; or, as The Economist[ix] reported, 'Gerstner's Chief of Staff' charged with re-engineering significant cuts in costs! A banker, with a big ego, he had just made his reputation by cutting the staff at Honolulu Federal Bank by thirty five per cent; and he had previously been chief technology officer at a number of banks - so he knew his way around IT. The writing seemed to be on the wall for IBM's cherished personnel policies. As it turned out, despite having no less than eleven taskforces re-engineering IBM, he proved to be too nice a guy and wanted to preserve the jobs! A year later, having been slow off the mark in reducing headcounts, and in particular in firing the poor performers, he was the first of the new team to be parted from IBM!

Indeed, although Aker's had got rid of tens of thousands of staff, more than Gerstner was to do, he had not actually got rid of its 'respect for the individual' promise and had bought off the leavers with up to two years salary. Under Gerstner the regime got much more brutal. 'Respect for the Individual' went out the door, and payouts were limited to less than 26 weeks; and staff knew that lifetime employment was well and truly buried! On the other hand, since the day, more than a decade earlier when John Opel made the fateful decision to recruit 100,000 extra staff, there never really had been any alternative.

The top-team, Gerstner and York, divided their effort so that the latter was responsive for cutting the expenses and Gerstner for boosting sales; almost in a good cop (Gerstner) bad cop (York) routine! York was both to restructure costs, trimming the fat off across IBM where operating margins had declined under Akers from fifty five percent to forty per cent, but also to sell off the inessential businesses - to boost cashflow.

Good Cop, Bad Cop…in turnarounds someone has to do the dirty work, since heads will have to roll and redundancies will need to be announced. Sometimes this is the job of the CEO himself, but in this case it was his side-kick. Having faced this problem myself, you should know that it is a job with a short time horizon; rarely does the hatchet man survive longer than two years. As with many unsavoury aspects of management, theory does not acknowledge this role!

Disposals…in terms of disposals, the most obvious choice was IBM Credit Corporation, which was not merely peripheral but made little money. On the other hand, it was too central to financing IBM's sales to larger customers. So it stayed. But IBM's Federal Systems Company had no such excuse - and it was sold to Loral for $1.58 billion[x].

Indeed, the team grabbed whatever funds they could find. Over the next four years it sold off 25 million square feet of real estate (including IBM's prestigious 43 story skyscraper on 57^th Street in mid-town Manhattan and its 556 acre complex in Boca Raton), and even 350 paintings from its $25 million art collection. By expanding into lower tax markets (or perhaps by manipulating transfer prices as many other corporations have done?) they also managed to drive down IBM's income tax rate from 41.5% in 1994 to just 32.5% in 1997.

Hidden Assets…one of the few upsides of a turnaround is that it forces the management to confront where it has buried its assets - and, as in the case of IBM - this exercise often unearths investments which could be much better used elsewhere. Hidden assets, being hidden, are rarely theorized about!

One of the first tasks was a benchmarking study, against IBM's main competitors. This showed that IBM's expenses to revenue ratio (at 42%[xi]) needed to be shaved by nine per cent (or $7 billion!); a requirement my own corporate planning team had identified a decade earlier! The old IBM favourite, of tightening up on expenses - literally down to paperclips, was reintroduced; with a vengeance.

Benchmarking…this is a high-sounding title for a very valuable activity; learning lessons from other organisations, and especially those - such as your competitors - who are similarly placed. Management theory teaches this lesson; but focuses too tightly on the numbers against which the benchmarks may be run - where the 'soft' (unmeasurable) issues may be even more important.

Rick Thoman

Following the further PC fiasco in 1993, Gerstner brought in his fourth new lieutenant; Rick Thoman, his most loyal executive who had been following him from job to job for more than a decade. Although his background was in finance, he was put in charge of the PC - along with a number of other peripheral divisions. He acted immediately, to staunch the bleeding; closing down a half dozen labs to concentrate PC development at Raleigh. He effectively killed all the sub-brands with the exception of the one successful one - the ThinkPad. The PS/1, PS/2, Value Point and Ambra were all mothballed. Even so, the PC group staff still had to run more than 3,500 SKUs (Stock Holding Units); an impossible task!

Thoman's quick fix was to replace the monthly meetings with weekly ones; starting, perhaps to make a point, at 8.30 am on Saturdays and lasting most of the morning. These meetings reviewed 40 odd pages of statistics on production and sales[xii]. At the same time thirteen European warehouses were closed. Tightening up operations in this way, he reduced the finished-goods inventory by 65% and procurement and distribution costs overall by 50%. Gross margins improved from 13% to 19% - closing in on Compaq's 25%.[xiii]

But neither Thoman or Gerstner really understood the interactive nature of IBM's R&D intensive businesses and, despite making $75 million in the first half of 1994, the PC was still in deep trouble. It was, indeed, Jerry York who recognised the problems in development and - from the beginning of 1995 - became the de facto head of PC group. But IBM's share of the PC market had already declined from 22% in 1985 to 8% a decade later.

To build share, Jim Canavino (IBM's long serving chief strategist) explored a merger with Apple; but it fell at the final hurdle - reportedly[xiv] due to the lack of commitment by Apple executives who seemed to want to take the money and run!

The one thing IBM then did do right at the time was to call Intel's bluff, when its newly-launched Pentium chip was found to be flawed. IBM insisted on a total recall; though Intel later presented this as a piece of sound disaster recovery on their own behalf. If only IBM had been as tough with Microsoft!

Following another fiasco, this time with the Workplace software development, but also paying the price of a number of similar misadventures, in 1995 Jim Cannavino was retired early. He was replaced by Bruce Harreld, previously president of the Boston Chicken fast-food franchise. He had little experience of technology, but was also a Harvard MBA and management consultant; albeit this time the Boston Consulting Group. Perhaps almost as important, he was a Lotus Notes fanatic; using it previously to run Boston Chicken. The almost inevitable joke in IBM was "Cookie man hires chicken man!"

Revolving Door

Perhaps the most surprising development was that - following a dozen senior jobs being reshuffled by Gerstner in a major restructuring - Jerry York, after just two years (rather than the three years he had apparently planned[xv]) as Chief Operating Officer rather than the CFO his title implied, resigned; going to help Kirk Kerkorian at Tracinda Corporation. He had been central to the cost-reduction programmes which had boosted IBM's profits, had put cash in the bank and - in particular - had repaired IBM's relations with Wall Street. His explanation, that he had completed his job, sounded reasonable; though apparently he commented[xvi] privately that "Lou's charter is, at this point, certainly the opposite [of mine]". But one was also left wondering what else he knew, especially where he had also been involved in IBM's new addiction to creative accounting!

Gerstner appointed Rick Thoman in his place, despite his questionable track record while at IBM. He, too, was a Wall Street favourite - so the loss of York, in terms of PR, was less than might have been expected.

His replacement at PC group - along (later) with responsibility for Storage Solutions division, Networking Hardware and Network Computers - was Bob Stephenson, an IBM veteran. The specific responsibility for the PC Group itself was then delegated to Sam Palmisano; another IBM time-server - who was eventually, in 2002, to be Gerstner's successor as IBM CEO. In less than a year, Palmisano managed to cut the development cycle and get to market faster with IBM's PC products than its competitors. He also boosted its market share by three points; less than Stephenson's goal of ten points[xvii], but the group's first good news for a long time.

In mid 1997 Rick Thoman also left IBM, to become President, a newly created number two spot, at Xerox. It took ten months for Gerstner to find a replacement; Douglas Maine, CFO at MCI Communications. As IBM was by then cash-rich, his role was just to be CFO, spending most of his time courting Wall Street.

Revolving Door…its is inevitable that - when bringing together a turnaround team made up of relatively unknown players - some will not survive. Less obvious, some of the key players who are needed to take the tough decisions at the outset may be unsuitable for the rebuilding tasks later; when their 'axe-men' reputations may have become counter-productive. In both cases their departure should then be rapid; helped along, if needed, by some financial incentives. This aspect has been reported in research undertaken by John Stopford at the London Business School, but has otherwise been largely ignored.
 

Turnaround from the Front

When he arrived at IBM, once again - following the pattern he set at RJR Nabisco (where he travelled150,000 miles in his first year[xviii]) - almost the first thing Gerstner did was to visit his 'troops'. He visited literally dozens of IBM locations. The form of these 'town meetings' was simple. He would only make a brief speech before spending an hour or more fielding questions. His was the message of an evangelist; stressing the need to remake IBM and 'leverage its size and assets'.

But the commentators who saw this as a new departure for IBM were wrong! It followed a long tradition. A decade and a half earlier, I had led an IBM team charged with doing exactly the same for a new general-manager parachuted in (that time from elsewhere in IBM) to rescue an IBM division which was then facing meltdown. Morale was at rock-bottom, and we too had to take the new general manager to the troops. The team, of half a dozen senior executives, spent the best part of two months building his image. The culmination was a speech to the whole division, about which he was so nervous that - breaking every rule - we had to get ply him with alcohol to relax him for the ordeal!

The Road Show…Gerstner was apparently more laid-back. In one especially impressive meeting he reportedly took off his jacket and sat on the edge of the stage while he extemporised. Relaxed and humorous, he was a wow! Again, though, I had seen the same thing happen previously in IBM. Indeed one IBM Vice President - Frank Cummiskey  - had, a decade or more before, made his reputation on this form of cosy approach. But the ease of extemporisation was an illusion. I had several times watched him rehearse these speeches over and over, for up to four hours (and through the early hours of the morning), so that they looked as if they were off-the-cuff. Although I was not in Gerstner's equivalent team, I suspect that - despite the reports that he spoke without any notes (and from the heart!) - the IBM staffers took as much trouble with him; after all, IBM's future depended on their work!

Gerstner made certain of getting the message to all the other employees by sending - apparently for the first time by an IBM CEO - an email to them; on 6 April 1993. In it he included the comments "Some of you were hurt and angered by being declared 'surplus' after years of loyalty and by some reports in the press about performance ratings. I am acutely aware that I arrived at a painful time when there is a lot of downsizing. I know it is painful for everyone, but we all know it is necessary. I can only assure you that I will do everything I can to get this painful period behind us as quickly as possible so that we can begin looking to our future and to building the business."[xix] Despite the many wordsmiths who must have worked on it, it was hardly uplifting, but - after all the prevarications of the Aker's period - it seemed refreshingly honest.

He was just as honest with the shareholders' annual meeting, just 26 days into the job, when he set out four objectives[xx]. The first two clearly signposted the cut-backs to come:

"Paring IBM to a more efficient size" and

"Developing a strategy that will make clear which businesses IBM will focus on"

The other two were more positive, if less direct:

"Decentralising decision making" and

"Taking more care of IBM customers."

On the other hand, his 'care of customers' was never to become as powerful as IBM's 'customer service' philosophy had once been, but in his own way he did continue to push the theme. Thus, Business Week[xxi] in 1996 stated that under him "…IBM has gone back to the most basic notion of how to succeed in business: talking to customers, learning their needs, and figuring out how to satisfy them…making that customer connection in Gerstner's theory."

Visible Leadership…the most important contribution to the initial 'spin' will be what the new CEO does. This is, in many respects, a theatrical performance - and one which (in a multinational) must take to the road in order to press the flesh around the world - but one where the impact on the audience has to very carefully judged. The best primer for this may well be Konstantin Stanislavky's classic, 'An Actor Prepares'; but don't expect any consultant to give it to you!

Culture…Gerstner's next task was to gain control of the culture. With more time available, my own 1970s team supporting the divisional general manager, whose entry was described earlier, achieved this by the expedient of putting all the senior management of the division - along with most of the field force - through the equivalent of the first year of the MBA at the London Business School. The resulting 'brain-washing' contributed significantly to the resulting turnaround - which saw the division rapidly quadruple its revenue!

Gerstner didn't have the luxury of time. As a result, his early moves were meant to be deeply symbolic, though they probably seemed trivial to outside observers. He demanded that business reports were kept below ten pages; a major change when you remember the difficulty I had in cutting back its marketing plan to below two hundred pages! He was even more rigorous with his two dozen top managers, who had to write a five page report giving an honest (and very concise) appraisal of their business, answering the questions:

What business are you in?

Who are your customers?

What's your marketplace?

What are your strengths and weaknesses?

Who are your main competitors?

These were, though, all classical questions which would be familiar to any Harvard MBA.

Even more radical, in terms of the prevailing culture, the use of acetate foils was not merely banned, so Gerstner could have meetings 'eyeball to eyeball', but the overhead projectors were removed from conference rooms! You will appreciate, remembering my earlier comments on management style, the impact this had on IBM's managers. It probably was even more cathartic, not least in Gerstner's own eyes, because over the previous years a pre-occupation with the style of presentation over the content had reportedly[xxii] dominated management style; and had come to militate against any change emerging from the ranks of senior management.

Mills & Friesen[xxiii] make the point of the previous regime that "…presentations took on a life of their own. Management Committee meetings became a theatre in which operating unit executives tried to outdo one another in a contest of who had the best ideas. This was the opposite of management-by-walk-around. Instead of managers going to visit people in the company to discover what was afoot, top managers were all brought to the mountain - to headquarters."

Cultural Change…transforming a strong culture, such as that of IBM, from the inside takes a very long time; typically of the order of 15 years. Thus, Opel's cultural changes, started at the end of the 1970s, only came to (bitter) fruition in the early 1990s. The pace can be accelerated by:
· the clear prospect of 'death' - such as happened to IBM in the early 1990s - which makes the stark choices open to employees brutally obvious
· bringing in charismatic, but untainted, 'outsiders' - such as Gerstner - who will form the focus for the new culture
· developing unique new policies - such as the abandonment of 'Respect for the Individual' - which clearly separate the new from the old
· promoting distinctive signs and symbols which very publicly show the changes being implemented - as the ban on the use of acetates did. For once, these very practical factors have been well documented by the academics working in the field of turnaround.

The New Court…then he started the management take-over in earnest! On 13 September 1993 he created an 'advisory group' (the CEC) of himself and ten others; meeting monthly. The existing unwieldy WMC (Worldwide Management Council), with thirty odd members, was relegated to meeting every two to three months. Even more radically he formally abandoned the practice of non-concurrence which I described in the earlier chapter. Though Akers had in theory abolished this in 1990, it still had lingered on in the culture. Now nothing, and especially nobody, was to get in the way of Gerstner's plans!

Even so, all but two (Gerry Czarnecki and Jerry York) members of the CEC were long-serving IBMers. But, within the next two and a half years, six of these departed. The new regime was brutal. It was no longer sufficient to make the numbers - you had to publicly sign up to the new policies as well! This sent a very clear message to the rest of IBM's executives.

One significant difference of the whole king-making process, which didn't seem to occur to any of the commentators, was the degree to which it personalised the issues. Like Bill Gates at Microsoft, Gerstner became IBM! Previously even the most controversial books about IBM had had talked about the corporation (with titles which often incorporated phrases such as 'Big Blue'), whereas those of the 1990s focused on Gerstner almost to the exclusion of IBM (as they also did on Bill Gates). This personality focus had never previously existed. Tom Watson Sr. hid his talents for a number of years, and even then he only promoted himself as the symbol of IBM; the reverse of Gerstner's approach. The new CEO, whilst seemingly 'modest' in his dealings with the press (indeed, at times, to the desperation of his team of spin-doctors, being positively antagonistic to journalists) and despite his 'decentralisation' promise to the stockholders, centralised IBM's processes so they all focused on his personal needs. This was in keeping with the times, but perhaps risky for IBM.

Royal Progress…the task facing the new incumbent is very like that facing the usurper in a royal court, the first task is removing potential opponents - fortunately not here by assassination - and ensuring that the key posts are in the hands of supporters, so that changes may be made more easily and faster. This matter is readily discussed by historians, from whom too few lessons in general are taken by management theorists, and occasionally by political scientists - but rarely by management academics. 

Advertising

When she took over, in 1993, as Senior Vice President heading up worldwide advertising, Abby Kohnstamm was faced with an IBM which was viewed with derision by much of the outside world. Not least, it was rated[xxiv] high on the scales of 'arrogant', 'bureaucratic', inflexible/rigid' and 'poor service'. In May 1994, therefore, her first major move was to consolidate all its advertising with one agency; Ogilvy. It had previously undertaken informal soundings with a number of other agencies - when Ammirati Puris Lintas[xxv] famously mistook Kohnstamm for a secretary - but the final choice was made without a single pitch having been delivered. On the other hand, IBM - with upwards of $500 million per annum to spend - negotiated a financially very favorable deal; which later was regretted by some of Ogilvy's management.

Starting with the intention to conduct more than a hundred focus groups, the outcome became obvious after the first couple of these. Although they still rated its products highly, even IBM's customers hated it! IBM had let them down, it didn't care about its customers, it had become too arrogant.

I myself had observed something like this happening when John Akers was taking over. The emphasis had then been on 'shifting iron' - on quick sales of hardware. I well remember a call I made with a large systems salesman, on an account which was planning to place orders for several tens of millions of dollars worth of networking equipment within four years; which was why I was there, and so interested. To my horror, after just a few minutes and having established that the business would not appear before year's end, the salesman thrust a business card into the customer's hand and told him to phone when they had some real orders to place. Outside the call the IBM salesman dressed me down for wasting time on business which wouldn't help him of his branch achieve business for the current year!

After a great many false starts, Ogilvy eventually came up with the now almost legendary campaign of 'ordinary folk' talking (in their own language, with subtitles) about IBM products. Most famously it had Czech nuns discussing OS/2Warp! The advertising was memorable, but - as was often the case in IBM - the specific product was wrong; though the impact on IBM's overall image may still have been positive - something that product advertising often ignores. The tag line, 'Solutions for a Small Planet', was just as creative and much more generally applicable.

Effective Advertising…this is one area where marketing becomes an art rather than a science. The creativity needed to generate successful campaigns is an unpredictable resource which, when found, needs to be husbanded; but it must be harnessed to the strategy, rather than the artistic needs of the copywriters. The best advice still is that given, several decades ago, by a practitioner: David Ogilvy.

The Atlanta Olympics…it was then not surprising that Gerstner, with his experience of using major events for promotional purposes - at Amex he sponsored the cleaning of the Statue of Liberty and supported causes, from orchestras to zoos, around the world - should want to provide the IT support for the 1996 Olympic Games. The challenge, with 271 medal events and 130,000 participants, eventually needed no less than 2,000 IBM staff; backed by four S/390s, eighty AS/400s and thirty RS/6000s, along with literally hundreds of LANs linked to seven thousand PCs!

It was to be foolproof, for IBM's (brand) reputation was riding on it; and half the world's population would be watching. Estimates[xxvi] put the total costs to IBM as high as $300 million; but, in return, it was put at the focus of IBM's promotional efforts.

Unfortunately, it was something of a disaster. The usual data entry errors, causing some comical slip-ups, might have been overlooked. But the all-important data feeds to the fifteen thousand journalists attending the event all but collapsed. Results had to be moved by hand, and the journalists complained not just to IBM but to their millions of viewers and readers. It was nearly a PR meltdown, but fortunately the viewers - concentrating on the pictures of the events - didn't notice the problem; and the post-event focus groups were positive.

IBM, for once, learned its lesson. Two years later, at the Nagano Winter Olympics, and four years later, at the Sydney Olympics in the summer of 2000, the system worked flawlessly; but by then Gerstner had withdrawn from the technology sponsorship of later events!

Networking Hardware

In terms of technology, one of IBM's blind-spots had become networking. This was to become a major problem; where IT (Information Technology) had transformed itself into ICT (Information Communications Technology) - with the emphasis on communications, and that was now where the money was to be made. IBM had, since the 1970s, ruled the roost with its SNA (Systems Network Architecture), which linked its massive mainframes. Although it had, in the early 1980s, lost some ground in Local Area Networks (LANs) to the Ethernet protocol, it was confident that its own answer - 'token-ring' - was more powerful. But it was to find that its very profitable offerings were starting to get out of step with the market and when TCP/IP (Transmission Control Protocol/Internet Protocol) appeared - to drive the Internet, IBM was bypassed! Once again it effectively handed most of this business to another corporation; this time Cisco systems, a start-up run in part by ex-IBMers. Cisco gratefully accepted the gift of the new multi-billion dollar market, and went on to become - at one point in 2000, before the dot.bomb crash - the world's most valuable corporation, then worth (yet again) several times as much as IBM.

Storage Systems…a similar crash afflicted IBM's (Adstar) storage systems division. This, too, had been one of IBM's traditional areas of great strength. But its products, too, had become too expensive and , worse still, under-performed their competitors. Using RAID (Redundant Array of Independent Disks) technology, EMC took more than half the market.

At first Gerstner railed against the hype of the industry, and especially against its 'vaporware' products which never appeared. He, in true IBM mode, favoured solid progress. Unlike Microsoft, IBM had always focused on the reliability of its products. But, by the second half of the 1990s, he was forced to recognise that cutting the development cycle time, and getting new products to market much faster, was what markets wanted. Accordingly, IBM products started to sacrifice reliability for speed of launch.

CMOS…perhaps his biggest technical decision, though, was to authorise the switch from bi-polar to CMOS technology for IBM's larger mainframes. With this move he effectively wrote off the tens of billions of dollars that IBM had invested during the 1980s in the plants needed to make this technology. But, CMOS was the technology used by its competitors, as well as on its own smaller machines. It was cheaper, by some 25%, so IBM profits could be boosted; at least in the shorter term. Although less powerful, it was much less cumbersome. I fondly remember frosty mornings at IBM, looking out of my office window at the steam shrouded cooling towers which needed for its water-cooled mainframes! Compared with a bi-polar machine which weighed in (with more than 6,000 parts) at more than 31,000 pounds in the early 1990s, the CMOS equivalent later in the decade weighed in at only 2,100 pounds and had just 92 parts. But, once more, the move from bi-polar technology removed an IBM unique feature, and its mainframes started to look like those of anyone else - and could be much more easily compared on a price per MIP basis.

Software…even this area - which, driven by IBM's mainframe business (which was steadily slipping), had seen IBM as the global leader for decades - eventually slipped into second place behind Microsoft; whose 1997 revenues, at $13.1 billion, just beat IBM's $12.8 billion. In the same year Oracle (with $1.82 billion in revenues) also overtook IBM (with $1.8 billion) in the database market. In quick succession IBM had lost leadership in most of its markets.

OS/2…We have seen, in the chapter on the PC and that on Microsoft, how IBM gave away the PC software market.. From IBM's point of view the final 'divorce' of their joint development effort probably came at the COMDEX show in 1989. In effect, out-negotiated by Bill Gates and Steve Ballmer, IBM conceded the small machine market to Microsoft, while it was to continue support of OS/2 on larger machines. MS Windows was already a major threat, but IBM's representatives (Jim Cannavino on-site and Jack Kuehler back at Armonk) thought their own developers could win the war. As we all now know, they were very wrong! Not least they did not expect Bill Gates to copy Apple's ground-breaking use of icons - the now standard GUI, Graphical User Interface which Apple had in turn copied from Xerox - and legally get away with it!

By 1994 the situation facing Gerstner was one where IBM only had about 4% of the PC operating systems market - to the 80% held by Microsoft. Even so, he equivocated about what should have by then become an obvious decision on the future of OS/2. Indeed, although no doubt he privately thought it was a lost cause, Gerstner expended a great deal of energy trying to win the war which was already lost. Jerry York complained that OS/2 was a black hole: "It isn't sustainable, it's a loser, the fucking thing sucks."[xxvii] The final drive revolved around re-branding it as 'Warp'; indicating that it was much faster than Windows. Unfortunately, although it was marginally faster in operation, which users did not notice, it took forever to load, which they did notice! Never was a brand given a more inappropriate name! By then, in any case, Microsoft was just about to move on to Windows 95; almost as silly a name, but a much more successful product. The nail in the coffin for OS/2 was IBM's inexplicable decision to ship both Windows and OS/2 on all its machines; despite what this did to the disk space demanded. Although IBM still supports OS/2, which is more than Microsoft does for the earlier versions of its software, Gerstner finally admitted defeat in mid 1995; after losing more than $1 billion on it.

PC… in this case, production remained an intransigent problem. York, with his background in Chrysler engineering, had rapidly recognised that the core problem was the range of incompatible parts which had to be stocked. At a time when it was estimated that - due to the rapid obsolescence in these fast-moving markets - IBM's PC parts inventory lost value at between two and five per cent per month. IBM's inability to accurately forecast individuals parts requirements either meant that it ran out of stock, and lost profitable sales, or it was left with massive amounts of unsold parts which (due to their incompatible designs) could not be used on other lines.

The problem was compounded by the fact that, when a new 'generation' of machines was launched, the gross margins on competing machines reduced to 20% after just  two months and to 10% two months later. So, being two to three months later to market, as IBM had been, meant it was losing large amounts of money.

The classical example had been seen in the mid-1980s, when IBM more or less ignored the arrival of Intel's 386 chip; and effectively handed leadership of the whole PC market to Compaq which, until then, had only led the small market for portables (or, to be more accurate, that for 'luggables', where these weighed a number of kilograms in weight). Running a loyal IBM PC dealership at the time, I was forced to switch almost all my business to Compaq!

In the case of the PC, at least, the key to this was not the speed of development but delays in its supply chain. The crunch came in 1993, when a series of glitches left IBM at year end with no less than $3 billion in PC inventories.

The accelerating pace of change is perhaps best illustrated by the rapid launch of generation after generation of Intel micro-processors:

'Workplace'…at the upper end of the market Gerstner indulged the development team on one of its blue skies projects, following in the steps of Future Series. Costing more than $500 million per annum, this was an operating system (Workplace) which would work across the whole range of IBM hardware. In one kernel it would combine MVS, VM, AIX and would even work on a desktop PC. Once more, even after more than $1 billion had been spent on it, it never saw the light of day. When Sun released Java it was shelved. This particular fiasco finally put paid to Jim Cannavino's career, and he took early retirement.

Completing IBM's shift away from its domination of the marketplace, 1994 also saw IBM actually producing chips, as a 'commodity' supplier, for other computer companies!

The Internet

Like a number of similarly placed CEOs, Gerstner was not at first a supporter of the Internet. Indeed, in 1995 he said[xxviii] that he hadn't met a single CEO who supported it; though by then Lotus - with its Notes software development well under way - had already become part of IBM. In that year, at the COMDEX show, he also said "We've come to understand that client/server is, in fact, not a full-blown phase of computing. It's really the leading edge of what will be the next phase: network-centric computing." This, rather than the Internet was to be his next flavour of the month!

By 1996 he was starting to see definite practical advantages in this ('net-centric') approach for the IBM hardware business. Led by Larry Ellison at Oracle, but supported by a range of other vendors such as Sun, NCs (Network Computers, cut-down PCs which demanded many of their resources be supplied by the network rather from their own internal hardware). It was promoted as the cheapest solution for desk-top computing (and one which was much easier for the central IT staff to control), which it may well have been, but users wanted independence! IBM's offering was a 'thin client', physically little more than a screen and keyboard. It too was something of a flop - as the price of fully configured PCs collapsed to become almost as cheap as the NCs - and IBM only made a quarter of a million units in total. The subsequent back-tracking was as confusing for its customers as for IBMers. Both were unclear what IBM's (or at least Gerstner's) strategy really was - a problem which his changing strategies often posed for IBM!

By 1998, however, Gerstner - for once moving faster than Bill Gates - had changed his mind and was fully behind the Internet boom; spending $200 million on an advertising blitz asking 'Are you ready for e-business?' His mainstream strategies reflected this, with support for Java-based open systems (including its own software), as well as endorsement for Windows NT and a range of minor alliances to facilitate Internet access. Where he saw Microsoft promoting Windows and its Office software, Intel promoting its microprocessors, Cisco its routers and Oracle its databases, he decided IBM would promote 'solutions'; and especially e-business . It would create networks (using its own and everyone else's hardware and software) to handle critical processes in its client's companies. He was fortunate, with this ambition, for he was then able to ride on the dot.com boom; even though IBM had, in 1998, just sold its 'Global Network' (then the world's largest data network) to AT&T. The justification for this sale had been that keeping it only alienated its competitors amongst the other telcos, who were to become major partners in IBM's new Global Services offerings. Of course the money also helped!

E-commerce…this is not the place for a one sentence answer; where I have elsewhere written hundreds of pages about this subject. But, in general, there may be two particularly valuable markets opening up. One of these, that of Consumer-To-Consumer (C2C) networking will be the largest in the longer term; but - with its emphasis on ownership of large memberships of individuals - is not directly available to IBM. On the other hand, that for facilitating Business-To-Business (B2B) co-operation, which is already huge, is ideal for 'consultancy' organisations - such as IBM is (and has been for many decades). The hype surrounding e-commerce - in both boom and bust - has, so far, hidden most of the realities.

Internet 'Hardware'…to handle the hardware side of its Internet business, IBM had - at the turn of the century - four series of e-servers:

xSeries (including Netfinity and, following its agreement with Sequent, NUMA-Q)

pSeries (including RS/6000)
iSeries (including AS/400)
zSeries (including Systen/390)

pSeries…this series is based on the RISC (Reduced Instruction Set Computing) technology first incorporated in IBM's RT PC in 1986, but launched most successfully in 1990 as the RS/6000 family - supported by a specialised operating system (AIX) which ran a proprietary version of UNIX.

iSeries…these servers can trace their ancestry back though the AS/400 to the S/38 (which started life more than two decades ago, in the 1970s, with full 64-bit addressing capability); which shows how powerful the Future Series (FS), abandoned earlier but which provided the underlying architecture for S/38, might have been.

zSeries…the architecture, introduced in 2000, is still base on that of the S/390 (which in turn came originally from the S/360), but it now has the ability to handle clusters of up to 640 processors, in 20 processor modules !

Post-Prodigy…IBM's own entry, long before AOL had appeared on the scene, had been a long-lasting, some would say long-dying, joint venture with Sears Roebuck. The paradox was that after a decade and a half, this had been killed off in 1996; after racking up massive losses! IBM's new strategy, though, was to concentrate on three solutions: E-commerce payment systems, supply chain services and software, and customer relations solutions (including business intelligence and call-centre technologies).[xxix] Fortunately for IBM, the first two - revolving around EDI (Electronic Data Interchange) were the province of the large multinationals - IBM's preferred customers - as well as at the heart of the fastest growing sector of the e-commerce market-place.

Global Services

Only this division then showed growth; though even this was flattening out. It was loosely based on consultancy and systems integration (turnkey packages in competition with EDS, the previous leader in the sector - whose revenue IBM overtook in 1995), but had a gross margin of less than 27%.

Although it had started to charge for its support services at the end of the 1970s, where previously they had been provided free in support of its hardware sales, IBM was still - at the beginning of the 1990s - a box shifter. But at that time Dennis M Welsh, who had run IBM's turnkey contracts for NASA, managed to start up a unit dedicated to this type of business. Although the turnover was only $7.6 billion in 1993, when Gerstner arrived, he persuaded the new CEO to make it something of a focus for new business - possibly the most important 'proactive' decision the latter took. It was some form of recognition that IBM's traditional business - before the 'Two Johns' switched it to almost pure box shifting - had been in total solutions; when its systems engineers (SEs), providing a (free) consultancy service, had been just as important to the overall sale as its salesmen. Global Services, in effect, stood the box-shifting approach on its head. The hardware was now only to be in support of the sale of the services!

Global Services  is a complex mix of different offerings, which can though be broken down[xxx] under four broad headings across three different types of situation: