Quick Tips

Short videos with quick tips to help you on your analytics journey. All videos can be accessed through Matt's YouTube channel.

Why is a good set of KPI's important?

Understand why it is better to have a dashboard with a well-thought set of KPI's instead of generic information in a business report.


Calculate percentage changes, skipping dates with no values in Power BI

Some people want to calculate sales percentage changes across dates, skipping dates with no sales. So the sales of each date should be compared to the immediately previous date that actually has numbers, skipping dates that have zero sales (or whatever) in between. You may see an approach to do this in the following video:


Three important aspects of demand forecasting

Demand forecasting allows businesses to optimize inventory by predicting future sales. By analyzing historical sales data, demand managers can make informed business decisions about inventory planning.


Intro on Dimensional modeling in Power BI

Some thoughts on why dimensional modelling & star schemas are important on Power BI data modelling.


Surrogate keys and column encoding in Power BI models

It is important to use integer surrogate keys in your Power BI semantic models for many reasons. Even if you can't, you should at least ensure your string based column keys should be encoded appropriately, to increase performance.


Save on Power BI storage costs by disabling MDX availability

You can reduce the size of your Power BI semantic model, by disabling the availability of columns for MDX based tools. This might stop excel users from connecting directly to your semantic model and creating pivot tables that utilize the pre-calculated measures you have in your model (they'll still be able to export report data to excel of course), but if that's not necessary you might see capacity saving of over 30%.


Combine in-budget and out-of-budget views in one Power BI graph

A client wanted to have different prespectives of their 'within budget' and 'out of budget' cases, in one graph. This is easy to achieve through DAX.