Suggestions on how to optimize data imports for a Real Estate Company with many products
Our company is a real estate developer and I am having a problem dealing with too much data. For example, we consider our condominium units as our products. These units are unique from each other (i.e. they have different unit number for each building and for each project development.) In short, no two units are the same and our total inventory can blow up to over 1,000.
Now. May I please solicit best practices from you on how to deal with this scenario? Is there other possible ways other than importing the whole product list?
Re: Suggestions on how to optimize data imports for a Real Estate Company with many products
1,000 is a low number. For financial models, its common to see record counts in the millions.
Here is the basic approach I take to staging data (95% or more of the time):
Create a numbered list (let's name it Condos)
Add a Display By Property in the list and format it as TEXT. Identify this property as the Display By property.
Add a module named Condo Data
The only list related to this module is our Condos list. This is important because, by only having 1 list, the data you load will be dense (and not blow up the model). You might also include Versions, if applicable.
In Condo Data, add whatever line items you need for attributes related to your Condo Data. Format the line items as appropriate. You may need to create ancillary lists, especially for items you may want to perform lookups or sums on later.
Add a line item to this list to calculate (or store) the description/display by value you want to show in the list. If this is a static value that comes directly from the load file, then you can load this directly into the list (display by) property. Otherwise
In a two step process:
Add your 1,000+ condos to the list (directly to the list). If you have a unique Condo Unit ID, load it into the list as the CODE. A CODE is not required, but it helps make lookups easier later (particularly if you need to fetch details about a single condo).
Load the data for the 1000 condos to the module. Load based on unique code value loaded in step 1. if you didn't have a code, you can load based on a combination of properties, perhaps address & unit number info. The important point is to load in a way that keeps each record unique.
Now you have your Condos defined.
If you need to load monthly fact data, such as, I don't know, revenue & expense, create a second module using the Condos list and Anaplan TIME (in months?) (plus Versions, if applicable). Load your monthly fact data here. At this point, you have two modules, the first contains attributes for reporting, the second contains monthly activity for reporting.