CHAPTER 10 Business Strategy as Applied Social ScienceRead more at location 1896
It is hard to exaggerate the strength of America’s competitive position in the world economy in September 1945.Read more at location 1898
by the 1970s, Europe and Japan had started to compete effectively again,Read more at location 1905
In 1963 Bruce Henderson, a farsighted purchasing executive who had recently lost his job at Westinghouse, convinced the Boston Safe Deposit & Trust Company to give him one room and a salary to form a consulting firm that within a couple of years was known as the Boston Consulting Group (BCG). He turned out to be one of the most original and influential business thinkers of the twentieth century.Read more at location 1910
By illustrative example, an auto manufacturer that had built 10,000 units of a specific type might observe that the 1,000th of these cars had cost $10,000 to manufacture, the 2,000th had cost $8,000, the 4,000th had cost $6,500, and the 8,000th had cost $5,100.Read more at location 1914
company could use this model to price cars at, say, $4,000 per unit now and lose money on the next few thousand, but seize market share from competitors who priced their cars with the aim of making money at current production costs.Read more at location 1921
what if a competitor develops a new production technology that is vastly more efficient? Or what if a new kind of product is introduced that is superior in cost or functionality? Or what if other competitors have access to lower-cost capital?Read more at location 1927
The Texas Instruments calculator business, in fact, imploded after several years of amazing growth when other competitors refused to play along.Read more at location 1930
Near the end of his professional life, Henderson wrote The Logic of Business Strategy (1984),Read more at location 1939
Henderson argued that strategic competition offers immense time compression versus natural competition.Read more at location 1947
by figuring out where natural competition is headed over many future trial-and-error steps, and jumping there in one big step, the strategic competitor compresses many evolutionary steps into one premeditated leap.Read more at location 1947
Henderson characterized natural competition as “evolutionary” and strategic competition as “revolutionary.”Read more at location 1951
vision of what must be known to compete strategically was incredibly demanding.Read more at location 1957
We are more than twenty-five years on from this judgment, and I see no danger of our developing the kind of comprehensive knowledge that Henderson said true strategic competition required. In retrospect, his prediction seems hubristic to the point of outlandishness.Read more at location 1966
I had become increasingly fascinated by applying mathematics to predict human behavior.Read more at location 1971
I started work at SPA in 1987, at age twenty-three, and immediately loved it.Read more at location 1982
My first assignment was as the junior member of a team charged with developing a strategy for the leading competitor in a mature industry that made commoditized glass-based products.Read more at location 1986
Evaluating competing claims for program effectiveness in a business usually is not simple,Read more at location 1998
the economy as a whole started to grow faster,Read more at location 2001
a new technology from an adjacent industry began making significant inroads into this industry,Read more at location 2002
Executives use experience, observation, and data to form intuitive judgmentsRead more at location 2004
the informed judgment of an experienced professional can be reliable.Read more at location 2007
after several years, changes in the competitive environment made the modeling clearly obsolete—muchRead more at location 2012
a new competitor entered the market that was part of a larger, integrated enterprise, and was making decisions that violated the economic assumptions of our framework, because they apparently were less concerned with making money in this market than in serving some larger corporate objectives.Read more at location 2016
They were really manifestations of one underlying problem: the analytical model of the business was always incomplete.Read more at location 2022
the route to success lay in superior execution of natural competition.Read more at location 2025
what really mattered was motivating and empowering the peopleRead more at location 2027
Tom Peters and Robert Waterman’s epochal business best seller, In Search of Excellence (1982),Read more at location 2028
The emotional energy behind this movement was a cri de coeur of the middle manager: I matter! I’m not just some piece on your chessboard.Read more at location 2029
senior executive who refused to accept an analytically derived strategyRead more at location 2034
a practical version of the observation that the analytical models the strategists used were incomplete,Read more at location 2035
Typically the most compelling of these objections would be linked to arguments about human behavior:Read more at location 2037
potential creative technological or business process innovations.Read more at location 2039
A classic example for American consumer products companies was what we came to call “the Walmart bomb.” A senior sales executive often would react to some strategy he didn’t support—say, eliminating some products or changing prices—by saying something like, “Sure, that might make us an extra $20 million, but it will put the whole Walmart account at risk, and if we lose them, we go out of business.” It’s plausible, terrifying, and usually not analyzable.Read more at location 2040
Two alternatives to strategic nihilism were attempted by those who saw that the strategy models were incomplete but wanted to find a wayRead more at location 2045
more general frameworks that could incorporate things like technological change, human motivation,Read more at location 2046
macro approach probably reached its intellectual apogee with the publication of the massive tomes Competitive Strategy (1980) and Competitive Advantage (1985) by Harvard Business School professor Michael Porter.Read more at location 2048
this framework is really just a very detailed and intelligent list of issues,Read more at location 2056
This is far from useless but is also pretty far from Henderson’s visionRead more at location 2064
to “go micro” by tackling somewhat more bounded problems.Read more at location 2065
not become so strategic that non-analyzable factors would overwhelm the benefits achieved through careful analysis and modeling.Read more at location 2071
What was innovative yesterday becomes routine tomorrow.Read more at location 2088
this is a good example of the overall process by which high-wage jobs are created and then destroyed in the information economy.Read more at location 2089
I ended up trying to build models for pricing, product introductions, and other consumer-oriented decisions. I discovered that I could build analytically sophisticated theories all day long, but it was very difficult to know whether they were correct, because by making slightly different assumptions in the analysis, I could get very different answers for the best predicted course of action.Read more at location 2096
as long as we were trying to predict human behavior, the problem was too complicatedRead more at location 2101
trying to predict the effect of changing the name of a convenience store.Read more at location 2104
executive of a company that operated 10,000 convenience stores, of which 8,000 were named QwikMart, and 2,000 were named FastMart.Read more at location 2106
annual revenue per store was $1 million in the QwikMart stores and $1.1 million in FastMart stores.Read more at location 2108
She wanted to know whether the company would increase sales by changing the names of all the QwikMart stores to FastMart.Read more at location 2109
the first logical question to ask was whether there were systematic differences between the QwikMart and FastMart storesRead more at location 2112
physical size of the store, how long the store had been open, number of people who lived near the store, average income of people who lived near the store, average number of children per family living near the store, number of nearby competitor locations by brand, relative quality of merchandise at each competitor store, number of parking places, traffic count on the road in front of the store, ease of access from the road, distance to nearest highway, visibility of store and signage, number and quality of other complementary nearby retailers, exact interior store layout, number of open hours per week, number of in-store employees, tenure and background of store manager and employees, mix of employees by skill level, match of employee demographics to customer demographics, amount of shelf space allocated to each department, number of individual products by department, exact position of each product on each shelf, total inventory on hand and inventory mix by department, number of stock-outs by department by day of week and time of day, number of checkout positions or cash registers, deployment of anti-theft technology, cleanliness of the store, quality and maintenance of interior lighting, presence of an ATM in the store, level of TV, radio, print, and other channel of advertising we had done for the market in which the store operated, level of competitive advertising in the same market by channel, relative quality of advertising copy we and competitors had executed for each market, and so on, in practical terms, ad infinitum.Read more at location 2114
The standard method for doing these adjustments is to create a regression equationRead more at location 2137
The conclusion is typically couched as “$50,000 is the estimated impact of store brand after controlling for other factors.”Read more at location 2139
we can never know we have identified and collected data on all the potential causal drivers of sales.Read more at location 2142
Adjusting for some but not all of the other potential control variables often does more harm than good in estimatingRead more at location 2144
But in our regression equation, we can have only one coefficient for the variableRead more at location 2152
In a complex system driven by human behavior, interaction effects are not peripheral issues, but usually are centralRead more at location 2159
direction of causality between control variables and the outcome of interest is often unclear.Read more at location 2162
Well-known examples include decision trees, case-based reasoning engines, neural networks, modern implementations of Bayesian statistics, clustering, and support vector machines, as well as various hybrids and extensions of these methods.Read more at location 2183
none of these can resolve the three core problems of omitted variable bias, interaction effects, and intercorrelationRead more at location 2187
the rebranded stores might have experienced a change in sales even if we had not rebranded them.Read more at location 2198
the first step is to ascertain the bias in selecting the case group versus the control group.Read more at location 2221
The only generally reliable way to test our theory is the approach that C. S. Peirce, Jerzy Neyman, and R. A. Fisher discovered many decades ago: roughly speaking, pick a random sample of QwikMart stores, rebrand them as FastMart, and compare what happens in them to a control group of stores that we do not rebrand.Read more at location 2240
A company can earn a lot of money by making experiments a central element of how it makes decisions—specificallyRead more at location 2249
But experiments must be integrated with other nonexperimental methods of analysis, as well as fully nonanalyti-cal judgments,Read more at location 2251
