A bottoms-up approach mandates that we look at the investment required to reach the market we are after. Here, we can view some investments by AT&T. After a disappointing attempt to capture "last mile customers" with an "almost broadband" service called "Project Angel," AT&T determined that it had to compete with the Incumbent LECs (ILECs) via cable—that is, via a physical connection versus a virtual one. This choice also proved imprudent from an investment perspective, despite what the "Wall Street" pundits were claiming at the time, and serves to illustrate the bottoms-up reconciliation approach. Let us look at an estimate of AT&T's investment and market capture requirements (Table 9).
Cable company acquisitions estimate
Two-way broadband upgrades estimate
Power solution cost estimate
Total infrastructure cost estimate
A market-acceptable rate of return might be taken to be 10%. This being the case, AT&T would have to earn $13 billion in net income annually, beginning in year one, to return an acceptable rate of return. Moreover, AT&T's investment is basically in the current period, while revenue streams would be in future periods; hence, net income streams would have to be discounted by the appropriate present value discount for the respective annual period. Nevertheless, if we accept a general estimate of $13 billion in required net income per year to simplify the calculations, we may ask how many customers AT&T would need to generate this level of net income.
Assuming AT&T's sales staff could generate revenue per customer of $125 per month, or $1500 per annum, for bundled services with a net income after tax per subscriber of 12%, or $180, we can quickly see that AT&T would need 72+ million subscribers ($13 billion divided by $180; well in excess of FCC market limits). Moreover, there are only approximately 280 million people in the United States, represented by approximately 115.9 million U.S. households.  Add to this mix a significant number of large well-funded competitors, and we see there was little chance for AT&T to capture sufficient market share to justify such an investment. Typically, market leaders capture just over 30% of any given market. As an aside, studies at General Electric also indicated that to have influence in a market, a player needs 13% to 15% market share.
Here is a wireless case in point. In early 1999, Vodafone PLC announced its acquisition of Airtouch Communications for approximately $62 billion. This was reported to the marketplace as an acquisition at a "good" value, as Vodafone only paid approximately $262per PoP, Airtouch having licenses that covered 236 million PoPs. However, drilling down, we would have noted that Airtouch only had 1998 revenues of $7.2 billion, and net income for that same year of $560 million. Moreover, its customer base was comprised of 5.4 million wireless phone subscribers and 4.1 million paging customers.
Looking at Airtouch's $560 million net income after tax in 1998, versus what was paid in January 1999, namely $62 billion, we see less than a 1% annual return—far below what could be earned on a risk-free Certificate of Deposit or U.S. Treasury Bill/Note at the time. Additionally, as the investment was up-front, and net income streams have to be present valued over time, we see an investment that appears to return significantly less than risk-free market-acceptable rates of return. Moreover, if we look at what was paid per actual customer versus per PoP, we see that Vodafone paid $6526 per existing customer but only $262 per PoP.
Such excesses in the wireless valuation arena continued into 2000 and 2001, with a number of European countries selling 3G wireless licenses for over $130 billion—just for the licenses! Estimates are as high as another $130 billion for European 3G infrastructure. With an estimated sunk cost of $260 billion (licenses and infrastructure), it is hard to imagine how carriers could ever earn a market-acceptable rate of return on this large of an investment base relative to the anticipated revenue and net income flows.
Undertaking a bottoms-up/top-down analysis, we can look at two of these countries, the United Kingdom and Germany. (For this example, we will treat them as one.) Asking ourselves how many customers they need to justify an investment of approximately $70 billion each, for a total of $140 billion (approximately $35 billion each for licenses, with a similar amount for infrastructure), for a 10% return on the investment (ROI), we can discern the following:
Germany in 2000 had an estimated population of 83 million and the United Kingdom 59.5 million. A return of $14 billion annually would be required for a 10% ROI.
Assuming subscribers will pay an average of $70 per month or $840 per annum, with a net of 10% or $84 per subscriber, we see that Germany and the United Kingdom would require approximately 167 million subscribers ($14B/ $84 per subscriber). This requirement is more than the estimated total population of both countries. Carriers in neither country arelikely to achieve market-acceptable rates of return based on such costs and population bases.
Looking at United Nations' population figures, Europe's population is estimated to decline between 2000 and 2050, and Maslov's Laws of Hierarchical Needs are at work here, too.  That is, people will not spend on mobile communications until other basic needs are met (food, shelter, security, etc.). So, based on human needs fulfillment and socioeconomic factors, some significant percent of the population will either be too young or not be able to afford 3G services, even if they would like to have them. Hence, a bottoms-up analysis provides fundamental but important information regarding potential investments.
Compounding the 3G dilemma and ultimately affecting values is the introduction of newer, more robust technologies: 4G and 5G. These newer technologies are already entering the market, before 3G has been implemented.
Another reason to temper third-party studies with supplemental assessments is that they often miss important fundamentals when making market prognostications. Here, we only have to look back a few years, when many study houses were projecting stellar growth in the LMDS markets. Well, as we all know now, these markets imploded, and the principal LMDS service providers have filed for bankruptcy protection. So why did these prognosticators miss the mark? Well, there are several reasons.
The first reason is "technology elasticity." Technology elasticity is the ability to substitute one solution with numerous substitute technology solutions. That is, many analysts do not understand or adequately address the dynamics of technology development and substitutability and the role that Schumpeter's "Creative Destruction" model plays. As many of the studies often virtually ignore disruptive technologies, it signifies the importance of maintaining a rearview mirror perspective in order to watch for disruptive technologies that may be gaining market acceptance.  In fact, a 360 view today is imperative to be, become, and remain a market leader and winner.
2000 U.S. Census data: http://quickfacts.census.gov/hunits.
Maslow, A. (1971). The farther reaches of human nature. New York: Viking Press; (1943). A theory of human motivation. Psychological Review, 50, 370–396.
For a great examination of disruptive technologies, see Christensen, C. M. (1997). The innovator's dilemma: When new technologies cause great firms to fail. Harvard Business School Press. Please note that the author believes this book is misnamed. It is not the Innovator's Dilemma, but everyone else's.