So you're ready to get started uploading data to Anna? Great! Before doing so, watch the following video (4 min) (or read below instead) and run through the checklist to make sure your data meets our recommendations and requirements.
Video 1: What data Anna works best on (4min)
Don't feel like watching? Read below instead.
What data Anna works best on?
At least of couple of key measures
Rich dimensions to explore
Transactional/ Time Series Data
Looking at patterns or trends over time
What data does Hyper Anna not work on?
Unstructured data - Hyper Anna does not perform sentiment analysis therefore anything that has a free text field will need to be grouped or excluded
Geographical visualisation - maps (longitude, latitude)
Hyper Anna does not do predictive modelling or forecasting - predictive modelling requires business acumen knowledge, judgement and overlays; However if you have your own forecast scores, you can upload that into Hyper Anna to trend and analyse.
Hyper Anna cannot merge multiple datasets automatically, Anna needs to read from a single table*
*Note if you have a database connection you can create your own custom SQL to merge tables into a single view
What's a good use case?
A good use case is one that solves a particular problem. Below are some examples of typical use cases with the problem statement, how data should be structured and the questions that can be answered.
right click and 'open image in new tab' to expand the view.
What's not an ideal use case?
One that is a data dump of every type of data that business has. There's no specific problem that this type of dataset can solve, and it makes it confusing for business to understand every single data column (especially if there's similar columns i.e. Department Aggregated, Source System Department) and try to make sense which column is best for their analysis.
For every use case you will need to think about:
Data availability - Is there currently data available to perform this type of analysis
Data source - Do you know where the data is stored? Is it in one location? Does it require merging of tables before hand to get it into a single view for Hyper Anna
Data quality - Is the data clean and in a shape that is ready for direct analysis
Data transformation - In the second part of this article, we will go through steps to make your data “Hyper Anna ready” and some data aggregation tips.
Data Preparation Checklist
The requirements and recommendations in the checklist below are applicable to both data upload methods.
Tips and tricks
We've outlined below a few of our tips and tricks that you can use to take your data from good to great.
While the examples are shown in a .csv file, these tricks also apply if you are connecting Hyper Anna to your database.
How to quickly re-format a date
If you have 2 date columns (i.e. Start Date, End Date) that you would like to analyse time taken, we recommend creating a duration column and set it as a segment.
Hyper Anna provides FY analysis automatically, you will need to define the start of the FY month in Step 5 of the data upload process; there's no need to have a separate date column to calculate FY time gran.
Hyper Anna will automatically provide a distinct count of every segment; If you have Employee ID, we are able to provide a calculated field 'Number of Employee ID' so that you don't need to provide another column 'Count of Employee'.
Hyper Anna will automatically aggregate measures and counts based on available time granularity in your data, i.e. if your data is at a daily level, we are able to provide a weekly gran, monthly gran, yearly gran automatically*, there's no need to have separate measure columns for each granularity.
Averages - Hyper Anna will automatically calculate averages between a measure and a segment
e.g. Measure = Sales, Segment = Department
Average = Averages Sales per Department.
There's no need to manually calculate averages in your data file.
* If your data is starts at monthly gran, Hyper Anna will not be able to provide daily/weekly gran.
If your data has coded values (i.e. Is Active Flag = 1/0), we recommend changing the data to natural language for business to understand i.e. Is Active, Is Not Active.
If you data is in the following format, we recommend transposing the data so that the data is combined into a single segment column (i.e. Office Location), instead of splitting it up into multiple measure columns (i.e. Sydney Office Revenue, Melbourne Office Revenue, NZ Office Revenue)
Now, your data is Hyper Anna ready! Read here to find out how to upload your data into Hyper Anna.