Using numbered transactional list vs. normal list with unique id (concatenated)

gplacona

I came across a model where all the transactional files are loaded into a numbered list, using the combination of properties (like period, cost center/profit center, company) for uniqueness. I understand the best practice would be to load these files into a flat list, using a unique id which, in this case, would have to be a concatenation of some of these "properties" and then, of course, load the data into a data module.

In which situation the former solution would be more advantageous than the latter? I hear something around handling errors in data load through lists is more efficient, but would love to hear some different point of views.

Misbah

@gplacona

Here are some of my observations. It goes against PLANS framework

1. Combination of properties is very detrimental to the performance of your model because you are asking Anaplan to create a unique key by merging multiple properties. While Anaplan will successfully create unique keys but it is going to consume a good amount of time based on the numbers of transactions that are going into the list. All of the data has to go to the list properties and you can not load/import the data into module. From my personal experience, I have experienced Anaplan time out and reject the load when I tried loading 10 million items into the list (by combining properties)
2. Since it is a numbered list with no Unique code, it will be almost impossible to refer the data from within the list. You will be forced to use aggregation function and take the data out of your properties into module line items.
3. You will consume more space since you are creating both the properties for storing the data and module line items to pull the data out of the list into modules.
4. Numbered list is unnecessary being used when there was no need to do so.

I do not see any advantage of using "combination of properties" approach.

Hope that helps

Misbah

Miz Logix