Webclass 67: Top 5 tips for data modelling – how not to make your Data Analytics campaign die a slow death

Q1. What are the 3 main problems that can occur when your data model is not good?

Q2. If you have more than one million records in the general ledger table, generally speaking, can you do your data analytics in Excel?

Q3. If you want your end-users to be able to discover issues for themselves, rather than being presented with exceptions, would you need more or less disk space/ RAM?

Q4. Why should we join the general ledger header table to the general ledger detail table in the scripting phase/ preparatory phase, rather than behind the data visualizations in the data model?

Q5. What would happen in ACL, if you joined the general ledger detail table to the general ledger header table, rather than joining the general ledger header table to the detail table?

Q6. Why do we ensure that key fields that we use to add information to our main table are unique?

Q7. Why should we keep long tables narrow?

Q8. Which information should we take out of the detailed table and put in the outer tables of the star?

Q9. Why should we use summary tables under objects (graphs, KPIs, etc) on the visualization?

Q10. Why should we only have fields from one table within one object on the visualization where possible?

Q11. If your management does not want to provide you with a lot of disk-space or RAM, what should you do?