Microsoft’s Crisis in Perception (Part 3)

In the parts 1 and 2 of this series I made the case that Microsoft is mistaken for a consumer products company when, at the core it is really an enterprise software company.

This misalignment between reality and perception is largely self inflicted and continues to be be dangerously reinforced by Microsoft’s marketing and PR machinery. The net effect of this ‘Crisis in Perception’ has been an undervaluing of the company’s assets and a distraction which has prevented the management team from pursuing a more focused strategy which would unlock that value. In this chapter I will discuss some of the structural challenges facing the company today and how they combine to place limits on the company’s room for maneuver.

When I joined the Microsoft in 1994 the structures and decision making processes at the company were quite straightforward. There were product development groups who all answered to BillG and there were sales, marketing and support organizations who all answered to SteveB. The COO was responsible for making sure the bills got paid and things like manufacturing and distribution were run as efficiently as possible. There was only a small legal team and HR was there to help you identify the best talent to hire from outside and then to keep the talent happy when it joined the company.

Of course the company was far smaller then and with fewer employees operating across a much smaller number of geographies. In 1995 Microsoft was ranked number 250 on the Fortune 500 list with revenues of just over $4.5 Billion. In the intervening period the company has become a behemoth and will check in around $60 Billion dollars in revenue this year (Disclaimer: All the revenue numbers I discuss here are projections based on Microsoft published third quarter results and publicly available historic data available here. I have no knowledge of fourth quarter FY’10 performance.) It is illuminating to compare the P&L performance for the 10 years prior to 2000 and for the ten years since.

In the period fiscal year 1990 to 1999 Microsoft’s revenue grew 1665%. Operating expense grew by less than revenue 1207% contributing to a bottom line net income growth of a massive 2787% over that ten year period. At the same time R&D investment went up by 1641% driven almost entirely by the company’s push to become a credible player in the enterprise. Scarily, the cost of administering all of this growth accelerated faster than revenue at 1833%. Finally, sales and marketing costs grew by 1079% as the company expanded both its global presence and enterprise sales capabilities.

Now compare the period 1990-1999 with the subsequent decade from fiscal 2000 to fiscal 2009. If you want to understand why the stock price has been depressed for so long these numbers tell the story. The company’s top line revenue grew by only 255% during this period (Obviously the law of large numbers starts to kick in and a 255% growth on Microsoft 2000 base revenue is a huge amount of revenue.) However, operating expenses during the same period outgrew revenue at 319% (Ouch!) and as a result net income grew a paltry 155%, exactly 100% less than revenue. At the same time general and administrative (G&A) costs soared by 352%, sales and marketing expense by 312% and costs associated with getting all the revenue (Support, licensing, distribution, channel etc.) rose a whopping 405%. No wonder the stock hasn’t done anything in 10 years.

G&A costs as a percentage of revenue are one proxy for the ‘Complexity’ of a company. The simpler the business model, product line and organizations structure needed to take it to market the lower your G&A needs to be as a percentage of income. In the period ’90-’99 G&A averaged 3% of revenue. In the period ’00-’09 that ratio had nearly tripled to 8%. G&A costs combined with the ‘Cost of Revenue’ and ‘Sales and Marketing’ lines is also a very good metric for efficiency. The fact that all of those costs grew substantially faster than top line revenue growth points to Microsoft facing its very own ‘Malthusian catastrophe‘. The situation is not sustainable and needs radical action today even more than it did in 2000.

One of SteveB’s first strategic decisions as CEO was the introduction of a GE like P&L structure. Each product group would become an ‘Independent’ division with it’s own P&L meant to represent the true costs associated with building, marketing and selling each of the products. In my opinion, there were only one or two ‘Minor’ problems with this plan.

The health care, lighting and aircraft engines businesses at GE really are independent of each other. There is almost not overlap between the technology in their products or the markets into which they sell. Each of the GE’s divisions has its own independent and dedicated sales and marketing resources. A strategic decision taken by the president of the health care business can be taken without reference to other divisions and with the knowledge that he or she has complete control over the execution of the strategy. The only thing GE Group cares about is whether a division meets it’s revenue growth and profitability targets.

Microsoft’s business is almost the antithesis of GE’s. In 2004 the company came up with the tag line ‘Integrated Innovation‘ which almost perfectly captured Microsoft’s development philosophy in one compact phrase. It also put a dagger in the heart of any plan to run development divisions as truly independent businesses. The strategy of ‘Integrated Innovation’ creates such a dense web of interlocking dependencies between all of Microsoft’s products that one division cannot sneeze without asking the permission of three other divisions first. ‘Integrated Innovation’ is absolutely the right strategy for an enterprise software company bent on meeting the needs of complex customers. However the strategy puts a straitjacket on any attempts to leverage your assets into other markets. Being held hostage to the ‘Integrated Innovation’ strategy is one of the primary reasons why Microsoft’s previous Windows Mobile operating systems compared so badly to iOS and Android. (It remains to been seen whether Windows Phone 7 can break the bonds imposed by the ‘Integrated Innovation’ model.)  The ‘Integrated Innovation’ messaging was also like red meat to the wolves of the Linux community and government regulatory classes who saw it as an attempt to leverage Microsoft’s dominant position in one market to gain advantage in another.

The other difference between GE and Microsoft is how you get your products to market. Microsoft has one sales force which sells through partners but has a direct relationship with small, medium and large enterprises. Each of the ‘Independent’ product divisions has to negotiate access to the sales resources required to get a product in front of customers. The life of a front line sales person at Microsoft has become unbelievably complex. They are the one point in the whole messy business where Microsoft’s product portfolio comes together and has to be integrated into a consistent conversation with the customer. Far from simplifying the business the new P&L model likely added complexity and cost due to the overhead in resources now needed to manage all the internal negotiations. No wonder all those operational expense items on the P&L have risen faster than income.

Then there’s the issue of leadership talent. When Steve introduced the P&L model he made it clear that he was expecting the the divisional leaders to step up to the role of CEO for their businesses. That of course requires that the company had the skills in its existing management team to act and lead like CEOs. Its important to put the size of these businesses in context. The Windows and Office divisions will generate $18-20 Billion revenue in 2010. If those were independent companies they would rank no lower than 120 on the Fortune 500 list. The server and tools division is a $13-15 Billion revenue business which would rank no lower than 177 on the Fortune list if it was a separate company.

To be the CEO of any company of this size requires: The ability to define and articulate a clear and exciting vision of the future, excellence in personal communications, an outstanding ability to select and nurture talent, the confidence to admit your mistakes and learn from them and a deft ‘political’ sensibility amongst many others. Not quite the skill mix you find in most great engineers and yet most of the presidents running these mega-businesses today rose through Microsoft’s engineering ranks. There are a few true superstars in this role like Stephen Elop who is a breath of fresh air but was hired in from outside where he already held a CEO position. As a shareholder I really do hope he can continue to thrive and succeed in the Microsoft culture.

As things were spiraling out of control in the early 2000s Steve made another management selection which proved controversial at the time. When it came time to pick a new Chief Operating Officer for Microsoft Steve reached out to an individual and an organization her knew well. His pick was Kevin Turner, then CEO of Wal-Mart’s Sam’s Club division.

Turner’s selection was unorthodox. Here was someone whose entire mindset and expertise was defined by a lifelong career in a company with a completely unique world view; Sell simple products in huge volume to consumers whose primary value driver is low cost, and squeeze every last cent out of your operations no matter what the cost in terms of morale or relationship with your suppliers (If you are the largest retailer in the world then your suppliers depend on you, not the other way round.) Of course if you believe in the fallacy that Microsoft is or can become a consumer products company then Turner might be the perfect COO. On the other hand, if the reality is that your business depends on selling very complex enterprise software solutions to complex customers who expect care and feeding for the millions of dollars they spend with you, then perhaps you need a COO with a broader professional skill set.

Ironically, I defended the hiring of Turner at the time to my organization and anyone else that would listen. My view was that the sales organization in particular was in desperate need of some discipline and structure if we were ever to get cost of sales under control. To his credit, Turner did just that in his first two or three years, forcing everyone to adopt rigid new scorecard and revenue forecasting processes. However, as with all things there is a law of diminishing returns and ultimately other strategic issues which need to be balanced. Unfortunately if you have a ‘Wal-Markian’ world view then there is no such thing as a diminishing return. In the consumer retail business where margins run at 5-6% one more penny squeezed from the cost line makes a difference. In the world of complex enterprise software sales, squeezing another penny out of the cost line just demoralizes the talent in your sales team who you depend on to explain the complexity of your product offering to customers who spend millions of dollars with you every year.

In my view its a pity that Steve did not take the opportunity to shake Turner’s hand, thank him for his great contributions and then let him go when his contract expired several months ago. If cost cutting becomes so endemic that the quality of your sales talent begins to decline and the ‘Tyranny of the scorecard’ becomes so stultifying that you leave no room for innovation and initiative, then your enterprise customer relationships and revenue will follow. If that happens Microsoft will enter a death spiral because there is no other business, not consumer, not on-line which can cover a shrinking enterprise business. In my opinion, we’re entering a very dangerous phase where the rigidity enforced by the network of dependencies between products, and the ever decreasing freedom to innovate in the field, is severely impeding the company from finding any creative way out of it’s very serious problems.

In the next and final chapter I will discuss some bright spots on the horizon and give my personal take on what strategy the company needs to pursue in order to achieve it’s full potential.