This section gives an initial example of the extrapolation of statistical results to business values and decisions, based on simple averages.

Say you are doing an analysis for a potential new drug. The following is information you currently hold:

As a business person, perhaps you would like to answer the following with regard to this scenario:

The statistical analysis works on the principle discussed in this chapter that the profitability will equal:

Profitability = Value * number of customers – fixed cost – variable cost

Based on the fact that the value has a low, medium and high value (mean and confidence interval), this leads to three profitability values:

High profitability: $2,550*15,000 – $8million – $500*15,000 = $22,750,000

Mean profitability: $1,800*15,000 – $8million – $500*15,000 = $11,500,000

Low profitability: $1,000*15,000 – $8million – $500*15,000 = -$500,000 (loss)

This does mean that the confidence interval of values includes the possibility of a loss. The breakeven value can tell us how far along the spectrum of CLVs a positive profit starts to occur. Breakeven occurs at profit = 0, i.e.

(Lifetime value * Number of customers) – Fixed cost – (Cost * Number of customers) = 0

Breakeven CLV = (Fixed cost + Variable cost) / Number of customers

= ($8million+$500*15,000) / 15,000

= $1,033

This breakeven CLV is quite close to the lowest end of the confidence interval. In fact, if we were to loosen our confidence to 90%, the lowest end of the confidence would probably be in positive profit territory (because less confidence makes for a narrower confidence interval). This suggests that the risk of loss is relatively low, and the vast bulk of the statistically likely range of profits is positive.

Of course, as discussed earlier, many organizations aim for a positive profit breakeven to achieve a minimum cost of capital. The above equations could easily be adjusted to reflect a minimum cost of capital.

Say that a regression which pertains to your industry and company size shows an unstandardized slope for the independent variable “Strikes” (number of days of trade union strikes annually in companies) on the dependent variable “Share Price” (share prices of the companies in dollars) of B = -0.0045, p < .01.

Say your company has 4.2 strike days in the year, and apart from any impact on share prices your management accountants estimate you lost $12,059,100 in profits per strike day. If you have 56,261,000 shares in issue and count a dollar of loss in share value as a loss of $1 in value, what is the total cost of strikes for this year?

In order to answer this, remember that an unstandardized regression slope has the meaning “the change in the dependent variable for a one-unit increase in the independent variable” i.e. in this case “the change in the share price for a one-unit increase in strike days.” Revisit Chapter 13 to review these concepts if necessary.

Therefore, the total loss estimation is:

Loss = (Daily loss * Strike days) – (Strike days * slope for days on shares * Number of shares)

Loss = -$12,059,100*4.2 + 4.2*-.0045*56,261,000 = -$51,711,553