Forecasting analysis in Power BI can be completed in quite an effective manner if you utilise the time intelligence functions in DAX effectively. In this example I wanted to run through the many options you have available to start automatically creating forecasts based on historical information.
Your forecasts in a lot of cases will be derived from your historical results, so why not look at ways to just automatically generate them if you can. What isn’t probably sensible though it to just take a like for like replica of any previous year, you may want to smooth it somewhat to account for some seasonality in your datasets or maybe even some anomalies that may not occur again.
I run through all a number of techniques here which are very easily repeated across a number of datasets, so definitely think while you’re checking out the video how you many apply this within your own environment. Sometimes I do see some pretty complicated forecasting processes around, but I think what you’ll find is that this is such and easy and effective way to complete a comprehensive forecast that can also be built upon, say by adding in economic factors, unique business factors and much more. You can use these results as an initial baseline number and then adjust it where you feel appropriate.
It’s all about the business applications with this. I actually cover a lot of similar applications around forecasting and much more in my Solving Analytical Scenarios virtual course at Enterprise DNA Online. This is worth checking out if you want to explore forecasting, budgeting, grouping techniques as well as scenario analysis and others.
One quick fire way you could even extend this is by averaging multiple prior years data. I haven’t done that here, but I may well do that in the future to show you exactly what I mean, but definitely go for it if you think you can.
All the best with this one