What is the best way to use ROI (return on investment)? The raw value, or as a rank?
(All fake data. When Castpoints is going, a simple search will produce realtime exact data.)
Say we want to invest some time and money. The raw ROI value is more direct and has clear units;
Small banks = 2.8% per year.
Hotdog stands for a parade = 80% per day.
Cleaning services = 5 to 10% per month depending on who’s hiring.
But its hard to compare the 2 because the percents, and timeframe, are far apart. We need to guess what a hotdog stand ROI might be at a yearly level. And then compare that to the cleaning service. Suddenly we have at least 9 things to compare, and guess at.
Plus, if we do a search for companies with ROI’s > 10% we will miss out on valid banking opportunities.
If we rank ROIs (0-100%) we can then easily compare them. But then banks will always rank low. So we rank within a classification. Which also gives reference points.
Money centers > Small bank = 90%.
Vending > Hotdog stand = 40%.
Services > In home cleaning service = 70%.
Now if we search for XYZ with a ROI rank > 80%, we are going to get the best in each category.
In other words we can now sort through a lot of information with minimal effort.
Instead of saying “80% ROI rank”, abstract another bit to get “80r.” That saves 5 characters, a lot on a cell phone. Also, typically the rank steps are in 5’s: 75r, 80r, 85r, etc. And searches are in steps of 10 until more detail is needed.
Evolution maximizes a scope’s, and all scope’s, future freedom. That is easy to apply when options can have specific values. But when they don’t, when they are open ended word problems, how are options quantified? A resulting high ROI suggests that the decision was rather intelligent. Also, ROI includes both extrinsic and intrinsic returns.
Instead of having to make a new data base for each exiting language when a new language is added, only one data base needs to be added for each new language.