Pertains to
SUCCESS Rules

2.2.5 and 2.2.6

Analyses within the framework of the SUCCESS Rules are explained here in greater detail since it is not easy to distinguish them from the key data associated with measures.

What do analyses mean in this instance? The idea here is to summarize those functions which a high-performance software solution either provides or should provide to analyze existing measures (including key data) in various dimensions.  

Clearly, one could say that measures are a component of the data model; analyses are a component of practical application.

In measures, we make a distinction between basic dimensions, on one hand and key data, on the other. We break analyses down into two groups: (A) comparisons and deviations on one hand and (B) analyses on the other. This differentiation is somewhat arbitrary, yet justified for the following practical reasons: Comparisons and differences play a special role in many management reports; they already have a high level of complexity and, to a certain extent, belong together more than the other forms of analysis.

Both groups are explained in further detail in the following:

(A) COMPARISONS AND DEVIATIONS

For the afore-mentioned reasons, both of these forms of analysis are treated separately under the evaluations grouped together under analyses.

Looking at the typical analyses used in financial and controlling reports under Comparisons and Deviations, it becomes clear that they are classified very well using the overview shown on the left (Rule 3.2.6):

+Comparisons place measures next to one another, for example the actual values for a year next to values for either the previous year or the budget.

+Deviations in absolute or percentage values are differences in measures, for example the difference of the previous year as a percentage of the previous year's value for a measure.

The system used in the picture to the left has been proven in actual practice, it can be easily extended to incorporate other types of budget, planning, or forecasts smoothly.

Not only is a uniform label for the different types of data such as ACT and PY as well as types of deviations such as ΔACT and ΔPY%, particularly important, primarily uniform visualization is as well.

Rule 2.2.2 illustrates our suggestion for standard dimension “period types.” Rule 2.2.3 illustrates our suggestion for the standard dimension of data categories.

(B) ANALYSES

This term refers to all other forms of analysis that serve to evaluate measures.

(1)Averaging: The term average usually refers to the mean of different values, but it can also refer to the median. Both flow and stock figures are suitable to calculate averages. Averages are formed for both temporal analyses (for example the months of a year) as well as structural analyses (for example the sales of several subsidiaries).

Typical averages for periods of time are:
+ Average daily value (within a month, quarter or year)
+Average monthly value (within a year or a quarter)
+Average quarterly value (within a year)

etc.

In this instance, specific measures are calculated for an average period

Typical structural averages are:
+ Average by product
+ Average by country
+ Average by customer
etc.

In this instance, specific measures are calculated for an average element of a dimension.

We do not include the following measures in analyses, but rather in key data:
+Result per employee
+Sales per square meter
+Costs per unit
etc.

In this instance, ratios are formed which should preferably be treated as key data.

(2) Accumulating: In this instance, accumulation (accumulation, year to date, ytd, annual progress value) pertains to the sum of successive time period values beginning at a set point in time. Usually the monthly values within either a calendar or business year are added up in order to see whether the figure for the year reached to date – usually in relation to the same periods of the previous year or budget – are within the anticipated ratio or not.

Accumulation in the stricter sense applies only to fluxes, such as sales or costs. We can also then speak of accumulation for stock figures if an “accumulated average” is formed here – for example the average number of employees in the first four, five, or six months of one year (this function is mentioned under Average because this refers to a progressive average over time in this instance).

Of course, measures can be summed up in forms other than periods, for example according to the valid structure of a dimension such as regions or products – however, these analyses should not be referred to as an accumulation, but rather as a summation, for example.

(3) Rolling: In an analysis, rolling (rolling, moving total, moving average) refers to the sum or average of a certain constant number of periods. Usually the last twelve monthly values of a calendar or business year are added up to determine whether or not the annual value reached thus far – usually in relation to the same periods of the previous year or budget – is within the ratio expected.

In this instance, rolling refers to both flow and stock figures – the designations MAT (moving annual total) and MAA (moving annual average) are common.

(4) Adjusting: Various forms for adjusting numerical series are used for business analyses:
+ season adjusted
+ inflation adjusted
+ currency adjusted
etc.

These analyses are significant for both flow as well as stock measures. Even if an adjustment for time (for example neutralization of the currency effect over several years) is usually used, there are also adjustments for structure (for example varying purchase prices for important raw materials at different locations).

(5) Ranking: This analysis refers to descending or ascending rankings of elements. Usually this involves forming sequences of measures in structural analyses. Application of ranking functions for temporal analyses is rarely useful. Of course, flow and stock measures can be ranked usefully in the same manner.

When ranking sequences, one can differentiate between characters (alphabetical) and figures (numerical).

(6) Selecting: In an analysis, selection is related to the ranking function: Usually this is used to determine either the best (fastest, most expensive) elements or the worst (slowest, cheapest) elements: Top ten, last ten, first quartile, last percentile, and so on, are the most common forms of selection, which are also used for statistical observations.

(7) Summing: In an analysis, summing (sum, add), refers to the addition of selected elements. An addition already specified in the hierarchy of a dimension – for example the addition of individual countries to one continent – should not be counted here.

Typical applications of the analysis function, Summing, are sales by customer or products with specific properties. 

(8) Indexing: The analysis function indexing summarizes all of the different functions which involve analyzing the deviation of a reference value of 1 or 100%:
+Deviation from the mean value
+Deviation from the maximum value
+Deviation from the value of a given dimension
etc.

In this instance, there are both temporal indexing (development of an exchange rate in relation to a reference date) as well as structural indexing (comparison of sales at one branch with average sales).
Flow measures and stock measures are both equally suited for this analysis function. 

(9) Standardizing: In this analysis function, standardizing (norm) refers to a temporal or also structural comparison of several shares of the whole expressed in a percentage of the sum.
Examples of this are the monthly export share of several products and the percentage of sales in export to several countries over the period of one year.

The index and standardize functions are similar, though indexing usually concerns the reference (usually expressed in percent) to one value, while standarzing concerns a comparison of several parts (usually expressed in percent) of one whole.

(10) Special analyses: As previously mentioned, the distinction of these analyses from other forms of analysis is derived from pragmatic considerations and the afore-mentioned list is by no means complete.

Further functions include:
+ Juxtapositions (which are not considered to be comparisons)
+Differences (which are not considered to be deviations)
+The last day of a period
+The first day of a period
+etc.

As examples of comparisons and differences in this sense, we refer to all ad-hoc analyses which serve a unique purpose – for example comparison of two outliers.

In regular analyses of this type, on the other hand, we refer to as comparisons and deviations (see above).

(Practical experience will show whether this somewhat complex distinction can be upheld...)