Download (pdf 1,4 MB)
MT 3 Semantic notation of basic measures
(by Rolf Hichert and Jürgen Faisst, published on 2017-01-17)
We would like to share some considerations on the semantic notation of basic measures (UN 3.1 Unify measures) in order to start a discussion on this open field. This is a new discussion and it is probably too early to include this topic in IBCS® version 1.1.
A semantic notation concept for basic measures must be the base for a notation concept for ratios which is not defined yet.
IBCS® version 1.0 already mentions three classifications (“cuts”) of measures:
+ Value measures and volume measures
+ Flow measures and stock measures
+ Measures of positive, negative, and neutral impact
Value and volume: If it helps to distinguish value measures (measured in currencies) from volume measures (measured in physical units), IBCS® suggests to use rectangular shapes for values and rounded shapes for volumes. However, the adoption of this suggestion on the side of IBCS® Certified Software vendors and IBCS® users is very low. And here is the reason for this as we think: Distinguishing value measures from volume measures alone is not enough to really “see” specific measures – it needs a more specific notation.
Flow and stock: In his 2016 Warsaw session Rolf suggested to use lines and areas for stock measures if it is helpful to distinguish them from flow measures. Again here: There was no overwhelming support of this concept.
In a comment on UN 3.1 Rolf further developed the idea to using trapezoid shapes for stocks. Also here, the feedback was not positive at all and Rolf now also thinks that this is going in the wrong direction and withdraws this suggestion.
And here our reasoning is the same as with values and volumes: not specific enough to really “see” concrete measures.
Positive, negative, and neutral impact: Andrej Lapajne made the following suggestion in his 2015 Amsterdam session dealing with waterfall charts: “Good” contributions should be shown lighter, “bad” ones darker. However, it has been criticized that gray shades already indicate previous periods and are used in stacked columns and bars. Using this concept we must agree on at least four different shades of gray. And this suggestions lacks a consistent solution for neutral contributions, too.
In summary: There is no accepted semantic notation for a practical classification of basic measures yet.
Need for a semantic notation of basic measures
Will readers of reports, participants in slide presentations, and users of interactive dashboards benefit from a semantic notation for the different measures? Sure they will, but it must be a consistent concept, easy to understand and memorize - and it must be a well accepted standard. There must be one concept only – not only within one organization but hopefully only one “worldwide”…
In music, architecture, engineering, cartography, and many other disciplines specific visual notation concepts are used worldwide for many years – with great benefits for all its users. Nothings exists like this in business communication yet – and this is what IBCS wants to deliver: Not only for scenarios, time periods, etc. but for the business measures which is probably the most important dimension.
Developing a visual language for specific measures
When creating visualizations for specific measures, we could be creative and design different visual expressions for sales, inventory, headcount, and some other important basic measures – similar to the letters of an alphabet. However, the “semantic distance” of some measures is small (e.g. cash inflow vs. cash outflow) whereas the semantic distance of other measures is big (e.g. sales vs. headcount).
Pattern recognition would be much easier if similar measures would look similar. So we suggest to create visualizations for the three cuts mentioned above, combine them, and add further distinguishing features making them looking unique for groups of measures resp. even specific measures. This procedure leads to a systematic development of a visual language for the identification of specific measures.
What is new?
What we have learnt from the status quo: It is not sufficient to use a visual concept for the three separate “cuts” mentioned above. This does not help to understand things much better.
But all of us would benefit dramatically if we could detect directly a group of measures (e.g. sales) or even a specific measure (e.g. export sales) – instead of its classifications only.
However, these cuts are very helpful when it comes to develop a visual language for specific measures. And this is what we have done here: For each “cut” we suggest a consistent visual language which allows to identify not only the twenty or so most important basic measures but leaves it up to the user to build more specific measures for his or her specific needs.
We have prepared a presentation illustrating the concept for download (pdf 1,4 MB).