贝恩咨询分析方法PenetrationCurve.ppt

Author: Nancy GreenePenetration Curve (S-Curve)March 1998Copyright© 1998 Bain & Company, Inc. 1Penetration CurvesBOS Agenda•Concept•Application•Case example•MethodologyCopyright© 1998 Bain & Company, Inc. 2Penetration CurvesBOS •Products can be thought of as following a life-cycle that has several distinct phases:–early adoption/trial–mass adoption–saturation/substitution–decline•The product life-cycle generates a series of market penetration rates that tend to follow a specific pattern, called an S-curve (or penetration curve)•Given the pattern of penetration curves over time, and given that this relationship holds to some extent in almost every case, penetration curves can be a powerful tool in predicting the growth of new productsWhat is a Penetration Curve? (S-Curve)3Penetration CurvesBOS Product Life-CycleTimeVolumeProducts can be thought of as following a life-cycle that has several distinct phases.Early adoption/trialMass adoptionSaturation/substitutionDeclineDriven by people who always adopt new things early and fuel the product’s initial growthThe product becomes accepted by the general market; characterized by rapid product growthThe product has either saturated the market, or is already being substituted for by another product that is in an earlier phase of the product life cycleThe decline of the product, driven by complete substitution for or replacement with another product4Penetration CurvesBOS Characteristics of Penetration Curves•Penetration can occur at the expense of an existing product, or it can drive new markets to grow•Several factors will influence the end state, or saturation point, of a product:–duration/speed of product life-cycle–competitive forces–innovation–culture/societyWhen considering penetration curves, there are a few characteristics to keep in mind:5Penetration CurvesBOS Agenda•Concept•Application•Case example•Methodology6Penetration CurvesBOS Applications of Penetration Curves•Validating/determining market growth for a new or established product•Showing precedents for success of a product through existing penetration curves•Predicting the saturation level for a product•Determining market entry strategy or timing, given the predicted penetration of a product•Determining product management or phase-out, given the speed of new products’ adoptionPenetration curves can be useful in addressing a series of different strategic questions which arise when new products enter a market.7Penetration CurvesBOS Agenda•Concept•Application•Case example•Methodology8Penetration CurvesBOS •In 1994, the National Institutes of Health (NIH) issued a statement confirming that most ulcers are strongly associated with the presence of a bacteria (H. Pylori). While previously, doctors had been advised to treat ulcer symptoms with anti-secretory drugs, NIH now recommends that ulcer patients with H. Pylori infections also be treated with antibiotics. The new treatment cures not only ulcer symptoms, but the underlying cause of the disease.•In 1997, a new device was granted a patent by the U.S. Patent Office which will test human breath for the presence of H. Pylori–a blood test for H. Pylori is already on the market–doctors may also test for H. Pylori using tissue samples obtained through endoscopy, a relatively invasive and painful medical procedure•Our client is considering buying the patent and believes that the breath test technology holds significant advantages over both blood tests and endoscopy–less invasive than endoscopy–more accurate than blood testsSituation:•How large is the market for H. Pylori tests over time?Key Question:A Bain case team used penetration curve analysis to examine how quickly medical practices change.Disguised Case Situation (1 of 2)9Penetration CurvesBOS •A search of the existing literature revealed that in 1995, only 5% of doctors were testing patients suspected of having an ulcer for H. Pylori–in 1991, the figure was close to 1%•In order to determine how quickly the NIH report would impact doctors’ behavior, our case team surveyed doctors in the fall of 1997:–60% said they were testing their suspected ulcer patients for H. Pylori–15% said they treated all suspected ulcer patients with a course of antibiotics --- without first testing for H. Pylori8the team agreed that this represented a “ceiling” on testing: 15% of doctors would always treat H. Pylori without testing for itResearch:•Historic penetration:–1991 = 1%–1995 = 5%–1997 = 60%•Saturation point = 85%Penetration Curve Data:Combining secondary research with surveys, the team obtained a few historic penetration points, as well as the projected saturation point.Disguised Case Situation (2 of 2)10Penetration CurvesBOS H. Pylori Testing Market= data point= predicted penetration curve (based on regression)Saturation point = 85%Penetration curve analysis smoothed the known data points into an S-curve that showed an aggressive schedule for the adoption of H. Pylori tests.11Penetration CurvesBOS Agenda•Concept•Application•Case example•Methodology12Penetration CurvesBOS •Gather historic data and set up spreadsheet•Pick saturation point–this should be the logical ceiling on a product’s penetration (i.e. not every home will have a computer, even in the most optimistic of scenarios)–saturation point = 100% in this example•Calculate penetration ratio–(historic percent) / (saturation point - historic percent)198919901991199225.0%29.0%36.3%42.5%YearsHistoric Penetration198919901991199225.0%29.0%36.3%42.5%YearsHistoric Penetration0.3330.4080.5710.740Penetration RatioMethodology (1 of 3)13Penetration CurvesBOS 198919901991199225.0%29.0%36.3%42.5%Years Historic Penetration0.3330.4080.5710.740Penetration Ratio7.5957.5967.5977.597ln(years)(1.099)(0.895)(0.561)(0.301)ln(penetration ratio)198919901991199219931994etc.25.0%29.0%36.3%42.5%Years Historic Penetration0.3330.4080.5710.740Penetration Ratio7.5957.5967.5967.5977.5977.598etc.ln(years)(1.099)(0.895)(0.561)(0.301)ln(penetration ratio)Predicted Penetration Ratio0.9661.268etc.•Take the natural log of the years and the penetration ratio•Regress the log of years vs. the log of the penetration ratio•Use results to calculate predicted penetration ratio–use formula for a line, and take anti-log–antilog of ( ln(year) x x-coefficient +b )Methodology (2 of 3)14Penetration CurvesBOS •Calculate predicted penetration using the predicted penetration ratio and saturation point–((predicted penetration ratio x saturation point) / (1 + predicted penetration ratio))19891990199119921993199419951996199719981999200025.0%29.0%36.3%42.5%YearsHistoric Penetration0.3330.4080.5710.7407.5957.5967.5967.5977.5977.5987.5987.5997.5997.6007.6007.601ln(years)(1.099)(0.895)(0.561)(0.301)ln(penetration ratio)Predicted Penetration Ratio0.9661.2681.6642.1822.8633.7554.9246.457Penetration Ratio49.1%55.9%62.5%68.6%74.1%79.0%83.1%86.6%Predicted PenetrationMethodology (3 of 3)15。